The Regulation of Protein Phosphatase 1 by Targeting Proteins

Transcription

The Regulation of Protein Phosphatase 1 by
Targeting Proteins
By
Michael J Ragusa
B.S., Siena College, 2003
A Dissertation Submitted in Partial Fulfillment of the Requirements for the
Degree of Doctor of Philosophy in the
Division of Biology and Medicine at Brown University
Providence, Rhode Island
May 2011
This dissertation by Michael J. Ragusa is accepted in its present form by the
Division of Biology and Medicine as satisfying the dissertation requirement for the
degree of Doctor of Philosophy.
Date___________
_____________________________________
Wolfgang Peti, Ph.D., Advisor
Recommended to the Graduate Council
Date___________
_____________________________________
Sam Beale, Ph.D., Reader
Date___________
_____________________________________
Tricia Serio, Ph.D., Reader
Date___________
_____________________________________
Justin Fallon, Ph.D., Reader
Date___________
_____________________________________
Dagmar Ringe, Ph.D., Reader
Brandeis University
Approved by the Graduate Council
Date___________
_____________________________________
Peter Weber
Dean of the Graduate School
ii
Michael Ragusa
Brown University
Department of Molecular Biology,
Cell Biology and Biochemistry
70 Ship Street
Providence, RI 02906
Phone: 203.948.8680
E-mail: michael_ragusa@brown.edu
Education
9/2005 – present
Ph.D., Biology, Brown University, Providence, RI
Department of Molecular Biology, Cellular Biology and
Biochemsitry
Projected date of completion: August 2010
Advisor: Dr. Wolfgang Peti
Investigated the role of the targeting protein spinophilin on the
specificity of protein phosphatase 1 using a combination of X-ray
crystallography, small angle X-ray scattering and other
biochemical techniques.
9/1999-8/2003
B.S., Biochemistry, Siena College, Loudonville, NY
Graduated magna cum laude.
Honors
2009
2006 – 2007
2003
Brain Science Kaplan Summer Graduate Award, Brown University
RI NSF/EPSCoR Fellowship Recipient
Award for excellence in the field of biochemistry, Siena College
Professional Experience
9/2003 – 8/2005
Research Technician, Laboratory of Dr. John Schwartz
Department of Cardiovascular Research
Albany Medical College, Albany, NY
Investigated the role of the transcription factor MEF-2C in neural
development using a neuronal knockout of MEF-2C in mice.
9/2001 – 5/2003
Undergraduate Research, Laboratory of Dr. Rachel SterneMarr
Biochemistry Department
Siena College, Loudonville, NY
Aided in the expression, purification and mutagenesis analysis of
GRK2 to elucidate the mechanism of interaction between GRK2
and the receptor β2AR. Developed expression protocols for Gα.
6/2001 – 8/2001
Undergraduate Summer Intern, Laboratory of Dr. Kevin
Claffey
Center for Vascular Biology
University of Connecticut Health Center, Farmington, CT
iii
Assayed for increased expression of mRNA binding proteins in
hypoxic breast cancer cell lines compared to non-hypoxic
conditions.
Peer Review Publications
1. Marsh JA, Dancheck B, Ragusa MJ, Forman-Kay JD, Peti W. Structural diversity in
free and bound states of intrinsically disordered protein phosphatase 1 regulators.
Structure 2010, in press.
2. Bollen M, Peti W, Ragusa MJ, Beullens M. The Extended PP1 toolkit: designed to
create specificity. Trends Biochem Sci. 2010; 35: 250-458.
3. Ragusa MJ, Dancheck B, Critton DA, Nairn AC, Page R, Peti W. Spinophilin directs
protein phosphatase 1 specificity by blocking substrates binding sites. Nat Struc Mol Biol
2010; 17: 459-464.
4. Sterne-Marr R, Leahey PA, Bresee JE, Dickson HM, Ho W, Ragusa MJ, Donnelly
RM, Arnie SM, Krywy JA, Brookins-Danz ED, Orakwue SC, Carr MJ, Yoshino-Koh K, Li
Q, Tesmer JJ. GRK2 activation by receptors: role of the kinase large lobe and carboxylterminal tail. Biochemistry 2009; 48: 4285–4293.
5. Li H, Radford JC, Ragusa MJ, Shea KL, Mckercher SR, Zaremba, Soussou W, Nie Z,
Kang YJ, Nakanishi N, Okamoto S, Roberts AJ, Schwarz JJ, Lipton SA, Transcription
factor MEF2C influences neural stem/progenitor cell differentiation and maturation in
vivo. Proc Natl Acad Sci 2008; 105: 9397–9402.
6. Ju T, Ragusa MJ, Hudak J, Nairn AC, Peti W. Structural characterization of the
neurabin sterile alpha motif domain. Proteins 2007; 69: 192–198.
7. Vong LH, Ragusa MJ, Schwarz JJ. Generation of conditional Mef2cloxp/loxp mice
temporal- and tissue-specific analysis. Genesis 2005; 43: 43-48.
In Preparation
Dancheck B*, Ragusa MJ*, Peti W. Molecular investigations of the structure and
function of the protein phosphatase 1:spinophilin:inhibitor-2 heterotrimeric complex.
2010, in preparation.
*These authors contributed equally to the work
Oral Presentations
2009 Department of Molecular Biology, Cell biology and Biochemistry Data Club, Brown
University, Providence, RI
Ragusa MJ. Reshaping the catalytic face of protein phosphatase 1.
2009 Interdisciplinary Graduate Student Seminar Series, Brown University, Division of
Biology and Medicine, Providence, RI
Ragusa MJ. Reshaping the catalytic face of protein phosphatase 1.
2007 Department of Molecular Biology, Cell biology and Biochemistry Data Club, Brown
University, Providence, RI
iv
Ragusa MJ. Regulation of protein phosphatase 1 by the targeting protein spinophilin.
Poster Presentations
2009 Protein Dynamics, Allostery and Function, Keystone, CO
Ragusa MJ, Dancheck B, Page R, Peti W. Spinophilin directs protein Phosphatase 1
substrate specificity.
2008 International Conference on Magnetic Resonance in Biological Systems, San
Diego, CA
Ragusa MJ, Dancheck B, Peti W. The regulation of protein phosphatase 1 by the
targeting protein spinophilin.
2007 North Eastern Structure Symposium, Storrs, CT
Ragusa MJ, Dancheck B, Kelker MS, Peti W. The regulation of protein Phosphatase 1
by the targeting protein spinophilin.
Teaching Experience
2008 – 2010
Sheridan Center Teaching Certificate 1
Brown University, Providence RI
Attended lectures and workshops discussing various teaching
practices and methodologies. Gave practice lectures for critical
review.
2007
Lab Teaching Assistant for Advanced Biochemistry
Brown University, Providence RI
Designed, supervised and graded all laboratories.
Techniques and Instrumentation
Molecular Biology
PCR, cloning, subcloning, mutagenesis
Protein Production
Protein expression in E. coli,
Protein purification (ÄKTA purifier and ÄKTAprime systems)
Biophysics
Protein complex crystallization (Rigaku micromax-007 X-ray
generator/R-AXIS IV++ imaging plate detector, Phoenix
crystallization robot, Cryscam digital crystallographic imaging
system)
Small angle X-ray scattering
Dynamic light scattering (Viscotek 802 DLS instrument)
Isothermal titration calorimetry (MicroCal VP-ITC and ITC200)
Circular dichroism (Jasco J-815 CD spectropolarimeter)
PDB Submissions
v
Critton DA, Ragusa MJ, Page R, Peti W. 2009 Crystal structure of a complex between
Protein Phosphatase 1 alpha (PP1) and the PP1 binding and PDZ domains of Neurabin.
RCSB PDB ID 3HVQ
Ragusa MJ, Page R, Peti W. 2008 Crystal structure of a complex between Protein
Phosphatase 1 alpha (PP1) and the PP1 binding and PDZ domains of spinophilin.
RCSB PDB ID: 3EGG
Ragusa MJ, Page R, Peti W. 2008 Crystal structure of a complex between Protein
Phosphatase 1 alpha (PP1), the PP1 binding and PDZ domains of spinophilin and the
small natural molecular toxin Nodularin-R. RCSB PDB ID: 3EGH
Page R, Critton DA, Ragusa MJ. 2006 Crystal structure of the HePTP catalytic domain
C270S mutant. RCSB PDB ID: 2HVL
vi
Acknowledgements
Graduate school has been an exciting and challenging experience. Throughout my
graduate career I received outstanding support from many friends and colleagues. Dr.
Wolfgang Peti, my advisor, served not only as my mentor but also a role model
throughout graduate school. I truly appreciate his work ethic and ability to think about
tough biological problems. I have tried strongly to adopt these traits myself. I also
appreciate his availability in the laboratory. I understand that being a professor is
incredibly time consuming, but despite this Wolfgang always makes time for his
students. He is constantly stopping into the laboratory to check on how experiments are
going or to discuss results, even if grant deadlines are looming on the horizon.
I would like to thank the two people who are responsible for everything I have
learned in the field of X-ray crystallography, David Critton and Dr. Rebecca Page. Dr.
Page’s understanding and mastery of crystallography is impressive to say the least. I
admire her focus and determination which often were the driving factors behind many
succesful synchrotron trips. David has always been there for me whenever I had a
question about my crystals, or data. I would often ask if I could borrow his eyes for a
moment and he always obliged me.
The members of the Peti laboratory have always made the laboratory an exciting
and entertaining place to work. Everyone in the laboratory is constantly pushing
themselves to be better scientists and it shows. This has motivated me during times
when I felt exausted. I would look at someone else in the laboratory still working hard
even after an all nighter or a couple of fourteen hour days and I thought to myself that I
can do that too. I specifically need to thank Violet who I held many discussion both in
and out of the laboratory about how to improve PP1:regulatory protein complex
purifications. I also need to thank her for her constant friendship and support.
vii
My family has always been a strong support base for me. Whenever I needed to
escape or just hear that I was a good scientist, even when I didn’t feel that way my mom
was there. My dad is a creative and hard working person and someone I have looked up
to for a very long time and someone I strive to be like. My brother and sister have been
excellent sources of support when I needed to talk about music, movies, video games or
just to think about something besides graduate school.
Lastly, I would like to thank my wife Meghan. I would never have been able to
survive graduate school without her. I could always tell that she was proud of me and
never hesitated to mention PP1 whenever anyone asked what I work on. She would
always listen to me when I needed to vent my frustrations when something wasn’t
working with my project. She was always understanding when I said that I hit an
unexpected delay and that I would not be home when I said I would. Her support and
constant understanding kept me going in times of exaustion or frustration and I can not
thank her enough.
viii
Table of Contents
Chapter 1. Introduction. .................................................................. 1
1.1. Phosphorylation.......................................................................................... 2
1.2. Substrate specificity of kinases and phosphatases. ................................... 6
1.3. Protein Phosphatase 1 and substrate specificity. ....................................... 9
1.4. Spinophilin. ............................................................................................... 14
1.5. GM. ........................................................................................................... 16
1.6. Flexibility in PP1:regulatory protein complexes. ....................................... 17
1.7. Research Aims. ........................................................................................ 19
1.8. Outline. ..................................................................................................... 20
Chapter 2. Spinophilin directs protein phosphatase 1 specificity
by blocking substrate binding sites. ............................................ 22
2.1. Abstract. ................................................................................................... 24
2.2. Introduction............................................................................................... 24
2.3. Results. .................................................................................................... 26
2.3.A. The spinophilin PP1 domain is unstructured in solution. ............................................. 26
2.3.B. The crystal structure of the spinophilin–PP1 holoenzyme........................................... 28
2.3.C. Spinophilin binds PP1 at multiple regions. .................................................................. 30
2.3.D. Spinophilin does not affect the PP1 active site. .......................................................... 33
2.3.E. Spinophilin-PP1 interactions define PP1 specificity. ................................................... 33
ix
2.4. Discussion. ............................................................................................... 36
2.5. Methods.................................................................................................... 38
2.5.A. Cloning and expression. .............................................................................................. 38
2.5.B. Protein purification. ...................................................................................................... 39
2.5.C. NMR measurements.................................................................................................... 40
2.5.D. Crystallization. ............................................................................................................. 40
2.5.E. Data collection and structure determination. ............................................................... 41
2.5.F. Dephosphorylation reactions. ...................................................................................... 42
2.5.G. Capture assay. ............................................................................................................ 43
2.5.H. Isothermal titration calorimetry. ................................................................................... 43
Chapter 3. The PP1:GM Holoenzyme. ........................................... 44
3.1. Introduction............................................................................................... 45
3.2. Methods.................................................................................................... 48
3.2.A. Protein expression. ...................................................................................................... 48
3.2.B. Purification of GM and PP1:GM..................................................................................... 49
3.2.C. Dephosphorylation reactions. ...................................................................................... 52
3.2.D. Isothermal titration calorimetry. ....................................................................53
3.2.E. Crystal trials. ................................................................................................................ 53
3.3. Results. .................................................................................................... 54
3.3.A. Sequence analysis of the PP1 binding and CBM domains of GM. .............................. 54
3.3.B. Purification of GM. ........................................................................................................ 56
3.3.C. Substrate specificity of PP1:GM. .................................................................................. 58
iii
3.3.D. Initial structural characterization of GM 2-237. ............................................................. 58
3.3.E. Thermodynamic interaction profile of GM and PP1. ..................................................... 60
3.3.F. Purification of the PP1:GM63-237 complex. ................................................................. 62
3.3.G. Crystallization trials of the PP1:GM63-237 complex. ................................................... 63
3.4. Discussion. ............................................................................................... 65
Chapter 4. The CBM21 domain of GM. .......................................... 67
4.1. Introduction............................................................................................... 68
4.2. Methods.................................................................................................... 69
4.2.A. Expression and Purification of GM102-237. ................................................................. 69
4.2.B. Crystallization of GM 102-237. .................................................................................... 70
4.2.C. NMR spectroscopy of GM 102-237. ............................................................................ 71
4.3. Results. .................................................................................................... 71
4.3.A. Purification and initial characterization of GM102-237. ............................................... 71
4.3.B. Crystallization of GM 102-237. ..................................................................................... 72
4.3.C. NMR spectroscopy of GM 102-237. ............................................................................. 73
4.4. Discussion. ............................................................................................... 77
Chapter 5. Small Angle X-ray Scattering...................................... 78
5.1. Introduction and Background.................................................................... 79
5.2. Beamline design. ...................................................................................... 81
5.3. Acquisition and analysis. .......................................................................... 83
5.3.A. Data acquisition. .......................................................................................................... 83
iv
5.3.B. The Guinier analysis. ................................................................................................... 86
5.3.C. The Kratky plot. ........................................................................................................... 88
5.3.D. The pair distance distribution function. ........................................................................ 89
5.3.E. Ab initio envelope determination. ................................................................................ 94
5.4. SAXS compliments high resolution techniques. ....................................... 98
5.5. Time-resolved SAXS. ............................................................................. 102
5.6. Summary. ............................................................................................... 103
Chapter 6. SAXS analysis of PP1 complexes. ........................... 106
6.1. Structural diversity in free and bound states of intrinsically disordered
protein phosphatase 1 regulators. ................................................................. 107
6.1.A. Abstract. ..................................................................................................................... 108
6.1.B. Introduction. ............................................................................................................... 108
6.1.C. Results. ...................................................................................................................... 110
6.1.C.1. Calculation of unbound intrinsically disordered ensemble models. ..........110
6.1.C.2. Unbound ensembles possess secondary structure of varying similarity to
the PP1-bound complexes. ..................................................................................111
6.1.C.3. Tertiary contacts in the unbound ensembles. ..........................................113
6.1.C.4. Calculation of an ensemble model of the partially folded PP1:I-2 complex.
.............................................................................................................................117
4.1.C.5. The PP1:I-2 complex: assessment using SAXS measurements..............119
6.1.D. Discussion. ................................................................................................................ 121
v
6.1.D.1 Structural properties of the unbound ensembles and their implications for
PP1 binding. .........................................................................................................121
6.1.D.2. Biological implications of the PP1:I-2 complex model..............................123
6.1.D.3. A structural continuum from folded to disordered states. .........................124
6.1.E. Experimental Procedures. ......................................................................................... 125
6.1.F. Acknowledgments. ..................................................................................................... 127
6.2. Molecular investigations of the structure and function of the protein
phosphatase 1:spinophilin:inhibitor-2 heterotrimeric complex. ...................... 129
6.2.A. Introduction. ............................................................................................................... 130
6.2.B. Experimental Procedures. ......................................................................................... 133
6.2.B.1. Protein Expression and Purification. ........................................................133
6.2.B.2. NMR Spectroscopy. ................................................................................135
6.2.B.3. Inductively Coupled Plasma Atomic Emission Spectroscopy. ..................135
6.2.B.4. Capture Assay. .......................................................................................135
6.2.B.5. Small angle X-ray scattering. ..................................................................136
6.2.C. Results. ...................................................................................................................... 137
6.2.C.1. Spinophilin417-583, I-29-164C85S and PP1α7-330 form a stable heterotrimeric
complex in vitro. ...................................................................................................137
6.2.C.2. Two manganese ions are present in the active site of all three PP1
complexes ............................................................................................................138
6.2.C.3. The I-2 RVxF motif is not necessary for stable PP1 binding. ...................138
6.2.C.4. Pull downs of I-2 mutants identify the SILK motif as critical for PSI
formation. .............................................................................................................141
vi
6.2.C.5. NMR spectroscopy of PS and PSI. .........................................................142
6.2.C.6. SAXS analysis of the heterotrimeric PSI complex. ..................................143
6.2.C.7. Ensemble model for PSI. ........................................................................145
6.2.D. Discussion. ................................................................................................................ 147
Chapter 6. Discussion and outlook. ........................................... 151
6.1. The PP1:spinophilin complex. ................................................................ 152
6.2. The regulation of PP1 by GM. ................................................................. 157
6.3. The role of flexibility in the heterotrimeric PP1:spinophilin:I-2 complex. . 158
6.4. Conclusions. ........................................................................................... 161
6.5. Short term outlook. ................................................................................. 163
6.6. Long term outlook................................................................................... 164
Appendix A. The effect of metal binding on the activity of PP1.
...................................................................................................... 166
A.1. Introduction. ........................................................................................... 167
A.2. Methods. ................................................................................................ 168
A.2.A. Protein expression and purification. .......................................................................... 168
A.2.B. Measuring the activity of PP1 against pNPP. ........................................................... 169
A.3. Results. .................................................................................................. 169
A.4. Discussion.............................................................................................. 171
vii
Appendix B. Small angle X-ray scattering analysis of the
unbound forms of Spinophilin and Inhibitor-2. ......................... 172
B.1. Introduction. ........................................................................................... 173
B.2. Methods. ................................................................................................ 173
B.2.A. Protein expression and purification. .......................................................................... 173
B.2.B. SAXS data acquisition. .............................................................................................. 173
B.3. Results. .................................................................................................. 174
Appendix C. Supplementary information for “Spinophilin directs
protein phosphatase 1 specificity by blocking substrate binding
sites.” ........................................................................................... 176
Appendix D. Supplementary information for “Structural diversity
in free and bound states of intrinsically disordered protein
phosphatase 1 regulators.” ........................................................ 184
Appendix E. Supplementary information for “Molecular
investigations of the structure and function of the protein
phosphatase 1:spinophilin:inhibitor-2 heterotrimeric complex.”
...................................................................................................... 192
Appendix F. The extended PP1 toolkit: designed to create
specificity. .................................................................................... 196
F.1. PP1 and the challenge of specificity. ............................................................................ 199
viii
F.2. The substrate specificity of the catalytic subunit. ......................................................... 200
F.3. The PP1 protein interactome. ....................................................................................... 202
F.4. Mutual control of PP1 and PIPs. ................................................................................... 203
F.5. Substrates. .................................................................................................................... 203
F.6. Substrate-targeting proteins. ........................................................................................ 204
F.7. Substrate-specifiers and inhibitors................................................................................ 206
F.8. PP1-docking motifs. ...................................................................................................... 207
F.9. Structural insights into PP1–PIP interactions. .............................................................. 210
F.10. PIPs as intrinsically disordered proteins. ................................................................... 210
F.11. Structural basis of substrate selection. ....................................................................... 212
F.12. Structural basis of PP1 inhibition. ............................................................................... 213
F.12. Enhanced selectivity through regulation of PIPs. ....................................................... 214
F.13. Concluding remarks and future perspectives. ............................................................ 215
Appendix G. Structural characterization of the neurabin sterile
alpha motif domain. ..................................................................... 220
G.1. Introduction. ........................................................................................... 222
G.2. Materials and methods. ......................................................................... 224
G.2.A. Expression and purification of the neurabin SAM domain. ....................................... 224
G.2.B. Chemical shift assignment and structure calculation. ............................................... 226
G.2.C. Zinc titration of the neurabin SAM domain. .............................................................. 227
G.2.D. Neurabin coiled-coil domain (neurabin627-822) expression and purification. .............. 228
G.2.E. Biophysical characterization of neurabin627-822. ........................................................ 229
ix
G.3. Results and discussion. ......................................................................... 229
G.3.A. The neurabin SAM domain has a typical SAM domain fold. .................................... 229
G.3.B. The neurabin SAM domain is a monomer in solution. .............................................. 233
G.3.C. The coiled-coil domain, and not the SAM domain, is the solely responsible
multimerization domain of neurabin. .................................................................................... 234
G.3.D. The shank3 SAM domain is the closest structural neighbor of the neurabin SAM
domain.................................................................................................................................. 235
G.4. Acknowledgements. .............................................................................. 236
List of References........................................................................ 238
x
List of Figures
Figure 1. Phosphorylation. .............................................................................................. 2
Figure 2. The effect of phosphorylation on ERK2. .......................................................... 3
Figure 3. The MAP kinase signaling cascade. ................................................................ 4
Figure 4. Recognition of phosphorylated residues. ......................................................... 5
Figure 5. The substrate specificity of kinases. ................................................................ 7
Figure 6. Tyrosine phosphatase and specificity. ............................................................. 8
Figure 7. The structure of PP1. ....................................................................................... 9
Figure 8. The PP1:MYPT1 crystal structure. ..................................................................13
Figure 9. Domain structure of spinophilin and neurabin. ................................................15
Figure 10. Model for targeting PP1 to AMPA receptors. ................................................16
Figure 11. The PP1:I-2 crystal structure. .......................................................................18
Figure 12. Flowchart of thesis. .......................................................................................21
Figure 13. The unbound spinophilin PP1 binding domain. .............................................27
Figure 14. The spinophilin417-583-PP1α7-330 complex. ..............................................29
Figure 15. Spinophilin’s interaction with the PP1 RVXF binding pocket. ........................31
Figure 16. Spinophilin’s PP1 interaction region II, III, and IV. .........................................32
Figure 17. Spinophilin creates a unique holoenzyme with novel substrate specificity. ...35
Figure 18. Spinophilin affects the PP1 substrate binding surface without changing the
active site. .....................................................................................................................37
Figure 19. The chemical reactions of glycogen synthesis and breakdown. ....................45
Figure 20. Domain comparison of the G subunit family of regulators of PP1..................47
Figure 21. The PP1:GM RVxF peptide crystal structure. .................................................48
Figure 22. IUPRED analysis of GM 2-240.......................................................................55
Figure 23. Attempted purification of GM2-240. ..............................................................56
xi
Figure 24. Purified GM 2-237..........................................................................................57
Figure 25. Substrate specificity of PP1:GM.....................................................................58
Figure 26. 1D 1H NMR spectra. .....................................................................................59
Figure 27. ITC of GM for PP1. ........................................................................................61
Figure 28. Purification of PP1:GM ..................................................................................62
Figure 29. The CBM21 domains from RoCBM21 and GL. ..............................................68
Figure 30. The final GM 102-237 sample. .......................................................................71
Figure 31. Circular dichroism spectrscopy of GM 102-237..............................................72
Figure 32. Crystallization of GM 102-237. .......................................................................73
Figure 33. 2D [1H,15N] HSQC spectrum. ........................................................................74
Figure 34. The backbone assignment of GM 102-237. ...................................................75
Figure 35. Chemical shift index for GM 102-237. ............................................................76
Figure 36. SAXS beamline design. ................................................................................82
Figure 37. Scattering data. ............................................................................................83
Figure 38. Snapshot of PRIMUS....................................................................................85
Figure 39. Real and reciprocal space. ...........................................................................86
Figure 40. Guinier analysis. ...........................................................................................88
Figure 41. Kratky plot. ...................................................................................................89
Figure 42. The pair distance distribution function. ..........................................................90
Figure 43. Different P(r) functions and their respective parameters. ..............................92
Figure 44. Selecting the best output from GNOM. .........................................................93
Figure 45. Structural information contained in scattering curves. ...................................95
Figure 46. Ab initio envelope calculation using GABSOR. .............................................97
Figure 47. Theoretical scattering data..........................................................................100
Figure 48. Investigating flexibility using SAXS. ............................................................101
Figure 49. Overlay of tetraloop receptor homodimer. ...................................................102
xii
Figure 50.Time-resolved SAXS experiments of cytochrome-c. ....................................103
Figure 51. Flowchart for SAXS data acquisition and analysis. .....................................105
Figure 52. Secondary structure content of calculated ensembles. ...............................113
Figure 53. Significantly populated clusters and their fractional contact plots. ...............115
Figure 54. PP1:I-2 NMR study. ....................................................................................116
Figure 55. The PP1:I-2 complex. .................................................................................118
Figure 56. SAXS data and analysis for the PP1:I-2 complex. ......................................120
Figure 57. PP1, spinophilin and I-2 form a stable, hetermotrimeric complex. ...............138
Figure 58. The RVxF motif of I-2 is released from PP1 in the formation of PSI, increasing
the importance of the SILK motif..................................................................................140
Figure 59. Structural rearrangement is seen in spinophilin upon formation of PSI. ......143
Figure 60. SAXS analysis of the PSI complex. ............................................................144
Figure 61. Comparison of the PP1:spinophilin and PP1:MYPT1 structures. ................154
Figure 62. Threonine 72 of I-2 in the PP1:I-2 ensemble model. ...................................159
Figure 63. Overlay of PP1:regulatory protein structures. .............................................162
Figure 64. Metal coordination of PP1. ..........................................................................167
Figure 65. The effect of EDTA on the solubility of PP1. ...............................................169
Figure 66. Michelis-Menten plot for PP1:spinophilin in 10μM CuCl2. ............................170
Figure 67. SAXS data for spinophilin. ..........................................................................174
Figure 68. SAXS data for I-2........................................................................................175
Figure 69. Ragusa, Supplementary Figure 1. ..............................................................177
xiii
Figure 75. Marsh, Supplementary Figure 1. .................................................................185
Figure 83. Dancheck, Supplementary Figure 1. ...........................................................193
Figure 86. The cover of Trends in Biochemical Sciences for August 2010. ..................198
Figure 87. Surface representation of the structure of PP1α (PDB ID 1FJM). ...............201
Figure 88. PP1-interacting proteins with multiple substrate-targeting domains. ...........205
Figure 89. Crystal structures of protein-PP1 complexes. .............................................212
Figure 90. PIP-mediated regulation of PP1 holoenzymes. ...........................................215
Figure 91. Ju, Proteins Figure1....................................................................................223
Figure 92. Ju, Proteins Figure 2...................................................................................232
xiv
List of Tables
Table 1. A short list of PP1 targeting proteins. ...............................................................11
Table 2. Data collection and refinement statistics for the spinophilin417-583-PP1α7-330
and the spinophilin417-583-PP1α7-330-nodularin-R complexes. ..................................30
Table 3. Crystallization conditions of the MR001 screen. ...............................................54
Table 4. Crystallization conditions setup for PP1:GM. ....................................................64
Table 5. GNOM parameters. .........................................................................................91
Table 6. General properties of calculated unbound intrinsically disordered ensembles
and TraDES ‘Coil’ ensembles. .....................................................................................111
Table 7. List of experimental restraints used in ENSEMBLE calculations and agreement
between experimental and predicted values. ...............................................................112
Table 8. Statistics for SAXS analysis of the PP1:I-2 complex. .....................................119
Table 9. Manganese content of PS, PI and PSI complexes as measured by ICP-AES.
....................................................................................................................................139
Table 10. Statistic for SAXS analysis of the PSI complex. ...........................................145
Table 11. The depedence of PP1 catalytic activity on metal ions. ................................170
Table 12. Radius of Gyration for spinophilin and I-2 determined by SAXS...................175
Table 13. Ragusa, Supplemental Table 1. ...................................................................183
Table 14. Marsh, Supplemental Table 1. .....................................................................191
Table 15. Bollen, Table S1. .........................................................................................219
Table 16. Ju, Proteins Table 1. ....................................................................................231
xv
List of Abbreviations
α-amino-3-hydroxyl-5-methyl-4-isoxazole-propionate…………………………..……..AMPA
Adenosine tri-phosphate…………………………………………………………..…….…..ATP
Advanced light source……………………………………………………………………….ALS
All that small angle scattering…………………………………………...………………ATSAS
Angstrom…………………………………………………………………………………………Å
Discrepancy…………………………………………………………………………….……discp
Carbohydrate binding motif…………………………………………………………….…..CBM
Carbohydrate binding motif 21 from Rhizopus oryzae glucoamylase………...…RoCBM21
Chemical shift index…………………………………………………………………….…….CSI
Circular dichroism……………………………………………………………………………..CD
Cryo-electron microscopy……………………………………………………...……….cryo-EM
Cyclin-dependent kinase 5………………………………………………………………..CDK5
Dopamine- and cyclic AMP-regulated phosphoprotein with molecular weight of 32 kDa
………………………………………………………………………………….………DARPP-32
Dissociation constant……………………………………………………………………..……Kd
Electrospray ionization………………………………………………………………………..ESI
Extracellular signal-regulated kinase 2………………………………………..….…......ERK2
Forkhead-associated …………………………………………………………….…..……..FHA
Forward scattering intensity…………………………………………………...……………..I(0)
Glutamate receptor…………………………………………………………………...……..GluR
Glycogen synthase kinase-3………………………………………………………..……GSK-3
Hydrodynamic radii………………………………………………………………………….….Rh
Inductively coupled plasma atomic emission spectroscopy………………..………ICP-AES
Inhibitor-2……………………………………………………………………………………….I-2
xvi
Intrinsically disordered proteins………………………………………………..…………..IDPs
Intrinsically unstructured proteins…………………………………………………….……IUPs
Isothermal titration calorimetry……………………………………………………...……….ITC
Lawrence Berkeley laboratory….………………………………………………………….LBL
Long term depression………………………………………………………………….……LTD
Mitogen-activated protein kinase…………….………………………………………….…MAP
Maximal distance…………………………………………………………………………..…Dmax
Minimal ensemble model……………………………………………………………………MES
Molecular dynamics……………………………………………………………...……………MD
Molecular weight……………………………………………………………………...………MW
Momentum transfer……………………………………………………………………………q
Myosin phosphatase targeting subunit 1…………………….…………………...……MYPT1
National synchrotron light source………………………………………………………..NSLS
Normalized spatial discrepancy……………………………………………………..……..NSD
Nuclear magnetic resonance…………………………………………………...……….…NMR
Nuclear overhauser enhancement………………………………………………………..NOE
Oscilliations………………………………………………………………………………….oscill
p-nitrophenyl phosphate……………………………………………………...……………pNPP
Pair distance distribution function…………………………………………………………p(r)
Phenylmethylsulfonyl fluoride…………………………………………………………….PMSF
Polyethylene glycol…………………………………………………………………………PEG
Polyproline II…………………………………………………………………………………PPII
Positivity…………………………………………………………………………………….positv
Post synaptic density………………………………………………………………………..PSD
Postsynaptic density-95/discs large/ZO1 homology……………………………………..PDZ
Protein data bank……………………………………………………………..……………..PDB
xvii
Protein kinase A……………………………………………………………….……………..PKA
Protein phosphatase 1…………………………………………………………..…...….…. PP1
Protein phosphatase 1:inhibitor 2 complex…………………………………….…....……....PI
Protein phosphatase 1:spinophilin complex…………………………………….…...……..PS
Protein phosphatase 1:spinophilin:inhibitor 2 complex ……………………….……...…..PSI
Radius of gyration………………………………………………………………….……….…..Rg
Secondary structure propensity…………………………………………………….………SSP
Size exclusion chromatography………………………………………………………….…SEC
Stability…………………………………………………………………………………...…..stabil
Striated muscle glycogen-binding subunit ………………………………………………….GM
Small angle X-ray scattering……………………………………………..………..….…..SAXS
Src homology 2…………………………………………………………….………….…......SH2
Systematic deviations…...………………………………………………………………..sysdev
Tobacco etch virus………………………………………………………………….………..TEV
uridine diphosphate……….…………………………………………………………..……..UDP
Validity in the central part of p(r)…………………………………………………………valcen
Wide angle X-ray scattering……………………………………………..………….…....WAXS
xviii
Abstract
Protein
phosphatase-1
(PP1)
is
a
major
serine/threonine
phosphatase
that
dephorphorylates numerous substrates, thereby regulating many cellular events. PP1
contains little substrate specificity on its own, but is able to dephosphorylate its
substrates with high specificity in vivo. This enhancement of specificity is achieved
through the interaction of PP1 with a large number of regulatory proteins. The majority of
these proteins are targeting proteins, which localize PP1 to its point of action, and alter
its substrate specificity. While the structure of PP1 was determined in 1995 in complex
with a small molecule inhibitor, there has been little structural data obtained for PP1
bound to regulatory proteins. As a result, the mechanism for the altered substrate
specificity of PP1 by its targeting proteins remains to be determined. To this end, we
have investigated the structure and function of the targeting protein spinophilin, bound to
PP1. The PP1 binding domain of spinophilin, which is intrinsically unstructured in the
absence of PP1, folds-upon-binding to PP1. When bound to PP1 spinophilin blocks one
of three substrate binding grooves, which results in the altered substrate specificity of
PP1, allowing it to dephosphorylate a subset of its total substrates. We have also begun
to investigate the structure and function of another targeting protein of PP1 called GM.
This protein targets PP1 to substrates that are inhibited by spinophilin and therefore will
provide information about a different mechanism for the altered substrate specificity of
PP1.
Finally, we have investigated the structure of two PP1:regulatory protein
complexes that contain regions of high flexibility when bound to PP1, including the
heterotrimeric complex formed between spinophilin, PP1, and an inhibitory protein,
Inhibitor-2 (I-2). The formation of the heterotrimeric complex creates an inactive enzyme
that is localized to its point of action and is primed to dephosphorylate its targets. The
xix
flexibility within the heterotrimeric complex makes structure determination by X-ray
crystallography difficult. Therefore, to investigate this structure a combination of NMR
spectroscopy, small angle X-ray scattering and molecular dynamics simulations were
used to create the first structural model for a PP1 heterotrimeric complex.
xx
Chapter 1. Introduction.
1
2
1.1. Phosphorylation.
Phosphorylation is the addition of a phosphate moiety to proteins, and occurs on serine,
threonine and tyrosine residues in eukaryotes. In humans, 98.2% of all phosphorylation
sites are serine and threonine residues1. Tyrosine residues represent only 1.8% of all
phosphroylation sites1. Phosphorylation is a ubiquitous and necessary regulatory
mechanism of protein function, which occurs on ~30% of all intracellular proteins2. The
addition of a phosphate moiety to proteins modulates their function, frequently by the
induction of a structural rearrangment
or a change in the charge distribution
of
a
Phosphorylation
process
Figure 1. Phosphorylation.
Phosphorylation occurs on serine, threonine
(shown above) or tyrosine residues. The addition
of a phosphate is carried out by protein kinases
and the removal is carried out by protein
phosphatases.
surface3.
proteins
with
is
the
a
reversible
addition
and
removal of a phosphate carried out
by two distinct enzymes: kinases and
phosphatases (Figure 1)4. Kinases
add a phosphate moiety to proteins provided by adenosine tri-phosphate (ATP), while
phosphatases remove a phosphate from proteins, thus creating inorganic phosphate
(Pi).
Phosphorylation is an essential tool in signal trandusction and regulates the
function of thousands of proteins. Two examples typical for the regulation of a protein’s
function by phosphorylation will be discussed to highlight the diversity of phosphorylation
as a regulatory tool. The first example shows how phosphorylation alters the activity of
an enzyme, the serine/threonine kinase ERK2, and how this is used as an activation
mechanism within kinase signaling cascades5. The second exampe is how
3
phosphorylation is used to regulate protein-protein interactions by either driving or
inhibiting the formation of protein assemblies6-8.
Phosphorylation is a major regulatory mechanism to alter the activity of
enzymes3. A well studied example of an enzyme regulated by phosphorylation is the
serine/threonine kinase ERK2. Phosphorylation of ERK2 on two residues, threonine 183
and tyrosine 185, both located within the enzyme activation loop, results in a structural
rearrangment of the activation loop5. The new conformation of the activation loop in
Figure 2. The effect of phosphorylation on ERK2.
The activity of ERK2 is regulated by phosphorylation. A) Comparison of the dually
phosphorylated (cyan PDB ID:2ERK) and unphopshorylated (red PDB ID:1ERK) structure of
ERK2. The activation loop is highlighted by a black box. B) Close up view of the activation
loop showing the structural rearrangement of the activation loop upon phosphorylation.
Threonine 183 and arginine 68 and 170 are shown as sticks.
ERK2 is stabilized by the formation of electrostatic interactions between the two
phosphate moieties on Thr 183 and Tyr 184 and Arg 68, 146, 170 (Figure 2 A,B)5.
Importantly, the rearrangment of the loop also induces a rotation of the N and C terminal
domains of ERK2 and a repositioning of active site residues. This increases the activity
of ERK2 by ~1000 fold9.
4
ERK2 phosphorylation and hence activation, occurs within a kinase signaling
cascade, refered to as the MAP kinase signaling cascade4. These signaling cascadees
often convey an extracellular
signal to the nucleus ultimately
modulating
gene
expression
(Figure 3). In the MAP kinase
cascade,
an
extracellular
response, often a growth factor,
interacts
Figure 3. The MAP kinase signaling cascade.
A cascade of kinases which ultimately results in the
regulation of gene expression.
with
a
membrane
receptor
resulting
activation
of
a
in
the
MAP-kinase-
kinase-kinase4. This triggers a
phosphorylation
cacade
in
which the initial kinase phosphorylates a second kinase, thereby activating it. The
activated second kinase subsequently phosphorylates a third kinase resulting in its
activation and translocation to the nucleus, where it phosphoryaltes transcription factors
ultimately resulting in the regulation of gene expression. Since phosphorylation is an
integral regulatory mechanism in kinase signaling cascades, it is involved in the
regulation of important cellular events including mitosis, proliferation, apoptosis and
many others.
Another example of the regulation of protein functions by phosphorylation is the
modulation of protein assemblies. Many protein domains specifically recognize
phosphorylated residues, including 14-3-3, WW, src homology 2 (SH2), phosphotyrosine
binding (PTB) and forkhead-associated (FHA) domains10. This allows phosphorylation to
act as a regulatory mechanism of protein-protein interactions. 14-3-3, FHA and WW
5
domains recognize phophorylated serine, threonine and serine/threonine residues
repectively10. SH2 and PTB domains recognize phosphoryalted tyrosine residues11.
Figure 4. Recognition of phosphorylated residues.
A) FHA domain (green) recognition of a phosphorylated peptide (blue). Electrostatic
interactions are shown as dashed lines in all figures. B) SH2 domain (cyan) recognition of
phosphorylated peptide (Pink). PDB IDs: 1G6G and 1P13.
The mode of recognition for FHA and SH2 domains are shown in Figure 4. Currently,
~200 proteins have been demonstrated to contain FHA domains, including kinases,
transcription factors, RNA binding proteins, and metabolic enzymes7. SH2 domains
occcur in 110 proteins including kinases, phosphatases, scaffolding proteins adaptor
proteins and transcription factors6,12. Therefore, phosphorylation is critical for hundreds
of protein assemblies involved in diverse cellular functions.
In addition, phosphorylation can inhibit protein functions through an inhibition of
protein-protein interactions13. One example of this is phosphorylation of the neuronal
scaffolding protein spinophilin. Unphosphorylated spinophilin associates with F-actin to
6
alter the morphology of dendritic spines14. The assembly of spinophilin and F-actin is
ablated upon phosphorylation of spinophilin8.
Since phosphorylation is such a ubiquitous mechanism in signal transduction, it
is not surprising that the dysregulation of the phosphorylation state of many proteins
frequently results in diseases, including cancers, cardiovascular disease and
neurodegeneration,
among
others15-17.
For
example,
the
dysregulation
of
phosphorylation is a critical step in Alzhiemers disease. In Alzhiemers disease, the
protein TAU is hyperphosphorylated, leading to the formation of TAU containing
neurofibrils18. The hyperphosphorylation of TAU is the result of a mutation occuring in
the protein kinase CDK5 creating an overactive form of CDK519. A dysregulation of
phosphorylation has also been reported in a mouse model for Parkinson’s disease20. In
this mouse model, numerous neuronally expressed proteins, including CaMKIIα,
contained abberant phosphorylation patterns, as a result of decreased phosphatase
activity20. These examples demonstrate the necessity for strict regulatory control of the
activity of both kinases and phosphatases.
1.2. Substrate specificity of kinases and phosphatases.
Recent predictions estimate that there are ~700,000 possible phosphorylation sites in
the human proteome2. Therefore, it is clear that both kinases and phosphatases must
recognize their substrates with high specificity to ensure the proper control of cellular
events regulated by phosphorylation. Kinases and phosphatases are characterized by
their preference for either tyrosine or serine/threonine residues2,4. Serine/threonine
kinases, tyrosine kinases and tyrosine phosphatases have high specificity for their
substrates. However, serine/threonine phosphatases do not21. The lack of specificity of
serine/threonine phosphatases is further validated by the number of genes encoding
kinases and phosphatases in the human genome, which contains 428 serine/threonine
7
kinases and only ~40 serine/threonine phosphatases1. Since the majority of
phosphorylation sites are reversible, each serine/threonine phosphatase must recognize
ten times more substrates, on average, than each serine/threonine kinase and therefore
contain less substrate specificity. This is in striking difference to the number of enzymes
encoded in the human genome for tyrosine phosphroylation, which contains 90 kinases
and 107 phosphatases1.
Both serine/threonine and tyrosine kinases recognize their substrates with high
specificity, which is acheived through the recognition of kinase specific consensus motifs
in their substrates2. The serine/threonine kinase, PKA, exclusively phosphorylates the
sequence R-R-X-S/T-Φ, where Φ represents a hydrophobic amino acid and S/T is the
target for phosphorylation2. The surface of PKA is complimentary to the recognition motif
containing pockets of negative charge to support the two arginine residues, as well as a
Figure 5. The substrate specificity of kinases.
The consensus recoginition motif of A) PKA and B) CDK2 support high substrate specificity.
The residues that are important determinants of substrate binding are shown as sticks. PDB
IDs: 1JLU and 1QMZ.
hydrophobic pocket to accommodate the hydrophobic residue at +1 to the
8
phosphorylation target (Figure 5A)22. Another serine/threonine kinase, CDK2, recognizes
the consensus motif S/T-P-X-K/R where the position of the proline at +1 is required for
efficient substrate binding as CDK2 contains a complimentary hydrophobic pocket for
the proline residue in the substrate (Figure 5B)23.
Furthermore,
tyrosine
phosphatases
achieve
high
substrate
specificity
predominantly by two mechanisms. First, tyrosine phophatases contain deep active site
pockets which can only accommodate phosphorylated tyrosine residues (Figure 6 A).
Second, the majority of tyrosine phophatases (79 out of 107) are multidomain proteins
that contain secondary protein interaction domains or motifs24. Many of these protein
interaction regions, including the kinase interaction motif (KIM), CH2 and PDZ domains
form secondary interaction sites that aid in substrate recognition (Figure 6 B).
Figure 6. Tyrosine phosphatase and specificity.
Tyrosine phosphatases achieve high substrate specificity through A) deep active site pockets
which specifically select phosphorylated tyrosine residues and B) secondary protein
interaction domains. Domain representations of different tyrosine phosphatases are shown
with the catalytic domain labeled PTP. Adapted from Alonso A. et al. 2004.
In
stark
contrast
to
other
phosphorylation
enzymes,
serine/threonine
phosphatases do not recognize specific consensus motifs and with the exception of
PP2B, do not contain additional protein interaction domains21. Instead, they contain
broad substrate specificity, which is exemplified by the fact that in vitro purified
9
serine/threonine phosphatases are even capable of dephosphorylating phosphroylated
tyrosine residues in addition to phosphorylated serine and threonine residues.
1.3. Protein Phosphatase 1 and substrate specificity.
The serine/threonine protein phosphatase 1 (PP1) dephosphorylates approximately 1/3
of all serine/threonine phosphroyaltion sites in the cell. PP1 is a single domain protein
with 3 isoforms, α, β (also referred to as δ in the literature) and γ, ranging in length from
323 to 330 amino acids25. The catalytic domain, residues 7-300 is the most highly
conserved region in PP1 with ~92% sequence identity between isoforms. Since apo-PP1
is only stable at concentrations less than 5 μM, no apo-PP1 structure has been
determined. Instead, the structure of PP1 has been determined in complex with various
small molecule inhibitors26-30. Examination of the first PP1:small molecule inhibitor
structure, the PP1:microcystin-LR structure, begins to explain why PP1 has little
Figure 7. The structure of PP1.
The structure of PP1 from the PP1:microcystin complex is shown as a surface representation.
The active site of PP1 is colored green, with two metals bound at the active site shown as
pink spheres. Three substrate binding grooves lead directly to the active site of PP1 and are
shown in red, orange, and blue. The RVxF binding pocket is shown in magenta. PDB ID:
1FJM.
substrate specificity26. In contrast to tyrosine phophatases, which have deep active sites
that aids in specificity, PP1 has a shallow active site (Figure 7). Two metal ions are
10
located at the active site of PP1, which in the crystal structure are Mn2+ because of the
presence of Mn2+ in the expression meduim and purification buffers, but are likely Fe2+
and Zn2+ in vivo based on structural similarity to the purple acid phosphatase, which
contains one molecule of Fe2+ and Zn2+
26
. The metal ions catalyze a single step
dephosphorylation reaction by the activation of a water molecule, which subsequently
performs a nucleophilic attack on the phosphate. The active site of PP1 is surrounded by
three substrate binding grooves termed the C-terminal, acidic and hydrophobic grooves
(Figure 7)26. This suggests that there are multiple distinct modes for substrate binding,
which would increase the number of possible substrates recognized by PP1 and
decrease substrate specificity. In addition, it has been demonstrated that PP1
dephosphorylates short phosphorylated peptides ineffeciently, in clear contrast to
kinases, which recognize short consensus motifs, solidifying the evidence for the
necessity of additional substrate interaction sites in PP121. Despite this requirement, PP1
does not contain any additional protien interaction domains that would provide the
necessary secondary substrate interaction sites. However, PP1 dephosphorylate its
substrates with high specificity in vivo, therefore PP1 must achieve substrate specificity
via a different mechanism than other phosphorylation enzymes31.
In contrast to serine/threonine kinases, tyrosine kinases, and tyrosine
phosphatases, serine/threonine phosphatases achieve specifcity by interacting with a
diverse set of regulatory proteins32. The number of regulatory proteins varies depending
on the serine/threonine phosphatase, but PP1 interacts with ~200 different regulatory
proteins that modulate the location, activity, and substrate specificity of the enzyme21. It
has been predicted that the total number of PP1 regulatory proteins exceeds 60021. The
majority, 90%, of PP1 regulatory proteins interact with PP1 via a common RVxF motif21.
This motif is commonly refered to as [R/K] [R/K] [V/I] [x] [F/W], where x is any amino acid
11
excepto for Phe, Ile, Met, Tyr, Asp or Pro, although it has been recognized that the motif
can be even more degenerate33,34. The RVxF motif interacts within a hydrophobic
binding pocket that is more than 20 Å away from the active site and is directly adjacent
to the acidic substrate binding groove (Figure 7)35. The RVxF interaction site is currently
believed to be the primary interaction site for many PP1 regulators.
There are two types of PP1 regulatory proteins: targeting proteins and inhibitory
proteins34. Targeting proteins, which represent the majority of PP1 regulatory proteins,
allow PP1 to dephosphorylate its substrates with high specificity. Inhibhitory proteins
block the activity of PP1 for its substrates. While PP1 is ubiquitously expressed, many
targeting proteins are often expressed in specific cell types (Table 1)34,36. These
regulators bind to PP1, localizing it to its point of action within the cell and induce
Tissue/subcellular Location
Physiological Function
Targeting Proteins
Skeletal Muscle, Heart, Liver
Glycogen Metabolism
Gm, Gl, R5, R6
Smooth Muscle, Myofibrils
Smooth Muscle Relaxation
MYPT1
Ubiquitous, Nucleus
mRNA splicing and processing
NIPP1, PSF1, p99
Ubiquitous, Plasma Membrane
Glutamatergic transmission
Spinophilin
Ubiquitous
Protein Synthesis
GADD34
Hematopoeitic, Nucleus
Cell Cycle Checkpoint
Hox11
Brain, Cytosol
IP3-mediated Ca
Ubiquitous, Centrosome
Centrosomal Functions
AKAP350, Nek2
Ubiquitous, Nucleus
Nuclear Envelope Reassembly
AKAP 149
Neuronal, Plasma Membrane
Neurite Outgrowth and Morphology
Neurabin
Neuronal, Post-synaptic Density
Synaptic Tranmission
Yotiao
Neuronal, Microtubule
Microtubule Stability?
Tau
2+
signaling
PRIP-1
Table 1. A short list of PP1 targeting proteins.
A complete list of PP1 regulatory proteins including all identified targeting proteins is
summarized in Appendix C. Adapted from Cohen, P.T.W. 2002.
substrate specificity. Localization by PP1 targeting proteins often occurs within
subcellular compartments, including the nucleus, and dendritic spine, but PP1 is also
12
targeted to diverse cellular structures including the plasma membrane, glycogen and
ribosomes34. The targeting of PP1 increases the relative concentration of substrates
within the proximity of PP1, which aids in driving the specificity of its dephosphoryalation
reactions. The combination of targeting and substrate specificity induced by its targeting
proteins allows PP1 to dephosphorylate its substrates with high specificity. The
localization of PP1 by its targeting proteins has been extensively investigated37-39, but
the mechanism by which targeting proteins alter the substrate preferences of PP1 is not
known.
The majority of targeting proteins most likely alter substrate specificity by
changing the substrate binding surface of PP1 upon complex formation. Some possible
mechanisms for this include 1) inducing a conformational rearrangement of PP1, 2)
extending the existing PP1 substrate binding grooves 3) adding novel substrate binding
sites and 4) blocking existing substrate binding grooves. Additionally, it is unclear how
altering the substrate specificity of PP1 affects the dephosphorylation of many PP1
substrates. The central reason for this is that PP1 activity is most frequently measured
against only a single substrate, phosphorylase a35,40-42. As a result, only two major types
of PP1 targeting proteins have been described thus far: those that inhibit the activity of
PP1 for phosphorylase a and those that do not.
Myosin phosphatase targeting subunit 1 (MYPT1) is a targeting protein that
inhibits the activity of PP1 for phosphorylase a and prior to this thesis was the the only
available structure of a PP1:targeting protein holoenzyme (Figure 8)43. MYPT1 is
specifically expressed within smooth muscle cells and tagets PP1 to the myosin
regulatory light chain, where dephosphorylation of serine 19 results in muscle
relaxation44. MYPT1 interacts with PP1 via the conserved RVxF motif and forms
additional interactions with the C-terminal tail of PP1. Upon binding, MYPT1 extends
13
both the acidic and C-terminal substrate binding grooves of PP143. It has been
hypothesized that the extended substrate binding grooves are important for substrate
binding in the PP1:MYPT1 holoenzyme43. However, there is currently no threedimensional structure available for PP1 bound to any substrate.
The primary reason for the lack of any PP1:substrate structure is that it is difficult
to create a catalytically inactive form of PP1, as the metal ions at the active site are
required both for the enzymatic activity and proper three-dimensional fold of PP1 (See
Appendix A). As a result, it has not yet been experimentally confirmed that the
extenstion of PP1 substrate binding grooves by MYPT1 is important for the altered
Figure 8. The PP1:MYPT1 crystal structure.
PP1 is shown as a surface representation in blue. Two manganese ions located at the active
site are represented as pink spheres. MYPT1 is shown in green as a cartoon representation.
PDB ID: 1S70
substrate specificity of PP1. In addition, MYPT1 has been shown to inhibit the activity of
PP1 for phosphorylase a41. However, it is unclear how this inhibition is achieved, as
MYPT1 does not block or induce a structural rearrangment of any substrate binding
grooves on PP1. Therefore, while the PP1:MYPT1 complex structure does provide
information about the interaction between PP1 and its regulatory proteins, it does not
14
provide a clear mechanism for the altered substrate specificity of PP1, which includes
the mechanism of PP1 inhibition for phosphorylase a. Therefore, the determination of
additional PP1:regulatory protein complexes in combination with a characterization of
the altered enzymatic activity of PP1 by its targeting proteins, will likely reveal a
mechanism for the altered substrate specificity of PP1 by its targeting proteins.
To gain insight into the mechanism of altered substrate specificity of PP1 we
have investigated PP1 holoenzyme complexes with two different targeting proteins,
spinophilin and the striated muscle glycogen-binding subunit (GM). These proteins
represent targeting proteins that inhibit the activity of PP1 for phosphorylase a
(spinophilin), and those that do not (GM) 42,45. Indeed, GM targets PP1 to phosphorylase a
in vivo25. From a biochemical standpoint, these two proteins are the best studied
regulators of PP1 with clear biological functions25,46. Therefore, spinophilin and GM are
excellent candidates for investigating the mechanism of the altered substrate specificity
of PP1 by its targeting proteins.
1.4. Spinophilin.
Spinophilin was identified in 1997 in the laboratory of Dr. Paul Greengard as a targeting
protein that localizes PP1 to the dendritic spines of neurons42. Spinophilin and its
neuronal specific isoform, neurabin, are multidomain scaffolding proteins containing 817
and 1095 amino acids, respectively. Both spinophilin and neurabin contain an F-actin
binding domain, a PP1 binding domain, a PDZ domain, and a coiled-coil domain (Figure
9)25. Additionally, only neurabin contains a SAM domain C-terminal to the coiled-coil
domain25. Both proteins interact with numerous protein partners (>30) including PP1,
guanine nucleotide exchange factors, F-actin, regulators of g-protein signaling, and the
tumor supressor ARF46. Due to the large number of interacting proteins, spinophilin is
able to perform many unique functions, which are both dependent and independent of
15
PP146. The most well understood functions of spinophilin are related to the function of
Figure 9. Domain structure of spinophilin and neurabin.
Domain maps of spinophilin (A) and neurabin (B) are shown with each domain labeled and
color coded. The starting and ending residues for each domain are also labeled.
dendritic spines and include regulation of spine morphology, density and synaptic
plasticity47. Additional functions have been identified in neurons, including the regulation
of neuronal migration, as well as in other cell types including cardiac muscle46,48. In
cardiac muscle, spinophilin targets PP1 to the type 2 ryanodine receptor where it
regulates the activity of the receptor and therefore cadiac muscle contraction49.
Within dendritic spines, spinophilin interacts with F-actin to localize PP1 to the
postsynaptic density, where it dephosphorylates serine 845 on the GluR1 subunit of
AMPA receptors50. The phosphorylation state of this residue regulates the probability of
the channel opening in response to the binding of glutamate. Spinophilin knockout mice
have significantly decreased long term depression (LTD), a critical mechanism for
synaptic plasticity, demonstrating the importance of the dephosphorylation of this site by
the PP1:spinophilin holoenzyme47. The importance of the PP1:spinophilin holoenzyme in
the regulation of LTD was further confirmed by monitoring synaptic activity of CA1
pyramidal cells in the presence or absence of a 5 amino acid RVxF peptide derived from
GM51. This peptide competed for spinophilin binding to PP1 resulting in an inhibition of
LTD.
16
Recently, a model has been proposed for the targeting of PP1 to AMPA
receptors52. This model is based largely on two findings. First, It has been shown that
the PDZ domain of spinophilin interacts with the GluR2 and GluR3 subunits of AMPA
receptors, but not the GluR1 subunit52. Second, spinophilin forms homodimers through
its coiled-coil domain, and it has been suggested that this leads to a 2:1 ratio of
spinophilin:PP153. One molecule of the spinophilin dimer interacts with PP1 and is
localized to the PSD through its interaction with F-actin. Once localized, the second
spinophilin molecule interacts with the AMPA receptor by binding to the GluR2/3
subunits via the spinophilin PDZ domain (Figure 10). However, spinophilin has been
shown to inhibit the activity of PP142.
Therefore, while this model explains
targeting of PP1 to the receptor it
does not explain how the receptor
becomes
dephosphorylated.
This
has led to the hypothesis that PP1
must
first
spinophilin
be
released
from
to
dephosphorylate
GluR1. Currently, the mechanism by
Figure 10. Model for targeting PP1 to AMPA
receptors.
Spinophilin homodimers form through interactions
within the coiled-coil domain to allow for PP1 (red
oval) targeting to the GluR1 subunit (blue cylinder)
of AMPA receptors. Adapted from Kelker et al.
2007.
which spinophilin is able to bind to
PP1 and inhibit its activity, while
allowing for the dephosphorylation of
AMPA receptors, is not understood.
1.5. GM.
GM was the first identified targeting protein of PP154. GM targets PP1 to glycogen
granules in skeletal muscle, where it dephosphorylates enzymes involved in glycogen
17
metabolism, including glycogen synthase and phosphorylase a25. Unphosphorylated
glycogen synthase drives the formation of glycogen from glucose, while phosphorylated
phosphorylase a causes the breakdown of glycogen into glucose. As a result,
dephosphorylation of glycogen synthase and phosphorylase a by PP1 induces the
synthesis of glycogen. GM may also play a role in adult onset diabetes as GM knockout
mice display similarities to type 2 diabetes patients, including insulin insensitivity, and
obesity55.
While the altered substrate specificity of PP1 is different between spinophilin and
GM, the mechanism of PP1 targeting by both proteins appears similar. In spinophilin, the
F-actin and PDZ domains, which are N-terminal and C-terminal to the PP1 binding
domain, respectively, target PP1 towards its substrate by localization within a specific
subcellular compartment. Similarly, GM targets PP1 to glycogen granules through the
interaction of a carbohydrate binding module (CBM) domain, which is located C-terminal
to the RVxF motif of GM. However, while the PDZ domain of spinophilin does not interact
directly with the PP1 substrate GluR1, the CBM domain of GM has been shown to
interact directly with PP1 substrates, including glycogen synthase52,56. This difference
may begin to explain how these targeting proteins modify the substrate specificity of PP1
in different ways. Nevertheless, without any structural information it is difficult to fully
understand the regulatory mechanism. Therefore, structure determination of both the
PP1:spinophilin and PP1:GM complexes will provide the necessary information towards
understanding the mechanism of altered substrate specificity of PP1 by its targeting
proteins.
1.6. Flexibility in PP1:regulatory protein complexes.
It has recently been demonstrated that several PP1 regulatory proteins are intrinsically
unstructured proteins (IUPs)57,58. This newly discovered class of proteins does not
18
contain the typical three dimensional fold that is associated with structured proteins59.
Instead, these proteins are highly dynamic in solution and therefore cannot be accurately
described as a single structure, but instead must be described as an ensemble of protein
structures60. IUPs often fold upon binding to their binding partners, with many becoming
restricted to a single conformation58. However, some IUPs retain a high degree of
flexibility when bound, which makes structure determination of these dynamic protein
complexes difficult by X-ray crystallography, as flexibility can inhibit protein
crystallization. Additionally, if crystals are produced, flexible regions often do not give
rise to discrete electron density and therefore cannot be accurately built into the crystal
structure61. This has already been observed in the PP1:regulatory protein Inhibitor-2 (I-2)
complex structure, which has recently been determined by X-ray crystallography62. In the
crystal structure, electron density was only observed for ~30% of I-2, demonstrating that
Figure 11. The PP1:I-2 crystal structure.
PP1 is shown as a surface representation with I-2 shown as a cartoon representation in cyan.
Several regions of I-2 were not visible in the crystal structure and are represented as black
lines. PDB ID: 2O8A.
~70% of I-2 remains flexible when bound to PP1. The regions that were not observed in
the crystal structure include residues that are important for the regulation of I-2 function,
such as the phosphorylation site threonine 7263. Upon phosphorylation of this site, the
19
PP1:I-2 complex regains phosphatase activity without releasing the inhibitor I-264.
Therefore, the crystal structure provides no structural or functional information about the
role of this phosphorylation site in the reactivation of PP1. Instead, investigation into the
flexible regions of this protein complex, requires additional structural biology techniques
including small angle X-ray scattering (SAXS) which can obtain information on the shape
of proteins, independent of their flexibility. Structural characterization of these flexible
regions may reveal possible mechanisms for the phosphorylation induced reactivation of
PP1.
Interestingly, it has been also been shown that specific pairs of targeting and
inhibitory proteins can form heterotrimeric complexes with PP165-67. One such
heterotrimeric complex is the PP1:spinophilin:I-2 complex, which localizes to dendritic
spines65. Currently, it is unclear what the detailed biological role of the heterotrimeric
complex is but it has been suggested to play a role cytoskeletal function and
morphology65. One possible advantage of the heterotrimeric complex is, that it creates a
holoenzyme that is localized to dendritic spines and poised for immediate activation in
response to a cellular signal. It is also unclear how both regulatory proteins are able to
interact with PP1 simultaneously, as both interact with PP1 through the RVxF motif,
which is an essential interaction site for many PP1 regulators41,68. However, it is likely
that either spinophilin or I-2 does not require the RVxF motif for binding and instead
releases from the RVxF motif binding site to allow for the heterotrimeric complex to form.
This release would enhance the flexibility of the heterotrimeric complex relative to the
heterodimeric complexes, indicating that the PP1:spinophilin:I-2 heterotrimeric complex
will likely be highly dynamic. Obtaining structural information may provide much needed
insight into the role of the PP1:spinophilin:I-2 heterotrimeric complex.
1.7. Research Aims.
20
This thesis aims to gain much needed insights into the mechanism by which regulatory
proteins bind PP1 and affect its substrate specificity. Both spinophilin and GM have been
extensively studied, yet the only available structure is that of a 7 amino acid RVxF GM
peptide bound to PP135. This structure provided initial insight into how regulatory
proteins bind to PP1, but provided no information about how they alter the substrate
specificity of the enzyme. Novel structural information is necessary to elucidate the
mechanisms for the altered substrate specificity of PP1 by its targeting proteins.
Additionally, it has been demonstrated that PP1:regulatory proteins contain regions of
flexibility. However, the biological role of these flexible regions in the regulation of PP1 is
unclear. Therefore, we will also pursue structural studies of PP1:regulatory protein
complexes that are flexible, including the PP1:I-2 and PP1:spinophilin:I-2 complex. This
will enable us to elucidate the role of flexibility within these complexes. Lastly, no
structural information has been obtained for any heterotrimeric PP1 complexes.
Therefore, structural information of the PP1:spinophilin:I-2 heterotrimeric complex will
provide insight into heterotrimeric PP1 complex formation, as well as will explain their
role in the regulation of PP1. Specifically, this work aims to address the following points:
1. The mechanism of altered substrate specificity of PP1 by the targeting protein
spinophilin.
2. The mechanism of altered substrate specificity and targeting of PP1 by the targeting
protein GM.
3. The role of residual flexibility in the PP1:I-2 and PP1:spinophilin:I-2 protein
complexes.
1.8. Outline.
21
This thesis is subdivided into two major projects related to the regulation of PP1. The
first project, involved a detailed investigation of the altered substrate specificity of PP1
by its targeting proteins spinophilin and GM. Chapter 2 describes all of the spinophilin
experiments. Structural investigation of the regulation of PP1 by GM required two
separate but related studies. In Chapter 3, the PP1:GM holoenzyme was studied and
progress towards the PP1:GM structure is described. In Chapter 4 we have investigated
the structure of the CBM21 domain of GM, which localizes PP1 to glycogen and
potentially plays a role in the altered substrate specificity of PP1 by GM.
In the second project, we investigated the role of flexibility within the PP1:I-2 and
PP1:spinophilin:I-2 protein complexes. My specific role in these studies was to
investigate their structure using small angle X-ray scattering (SAXS). Since, SAXS is
currently an expanding field with many applications related to structural biology, an
introduction to SAXS theory and data analysis is presented in Chapter 5. This
introduction will aid in the understanding of the SAXS data for the PP1:I-2 and
PP1:spinophilin:I-2 protein complexes which are described in Chapter 6.
Figure 12. Flowchart of thesis.
Two projects are discussed in this thesis: the regulation of PP1 by targeting proteins (blue)
and flexibility in PP1:regulatory protein complexes (orange).
Chapter 2. Spinophilin directs protein phosphatase 1
specificity by blocking substrate binding sites.
22
23
Reproduced from the journal Nature Structural & Molecular Biology, Volume 17, Number
4, April 2010 pages 459-464.
All of the experiments for this paper were performed by M.J.R. except for the NMR
studies of unbound spinophilin which were performed by Barbara Dancheck, and the
model building of the neurabin:PP1 structure which was performed by David Critton.
Michael J Ragusa1,2, Barbara Dancheck1, David A Critton2, Angus C Nairn3, Rebecca
Page2 & Wolfgang Peti1
1
Department of Molecular Pharmacology, Physiology and Biotechnology and
2
Department of Molecular Biology, Cell Biology and Biochemistry, Brown University,
Providence, Rhode Island, USA. 3Department of Psychiatry, Yale University School of
Medicine, New Haven, Connecticut, USA.
Correspondence should be addressed to W.P. (wolfgang_peti@brown.edu).
24
2.1. Abstract.
The serine/threonine protein phosphatase 1 (PP1) dephosphorylates hundreds of key
biological targets. PP1 associates with ≥200 regulatory proteins to form highly specific
holoenzymes. These regulatory proteins target PP1 to its point of action within the cell
and prime its enzymatic specificity for particular substrates. However, how they direct
PP1’s specificity is not understood. Here we show that spinophilin, a neuronal PP1
regulator, is entirely unstructured in its unbound form, and it binds PP1 through a foldingupon-binding mechanism in an elongated fashion, blocking one of PP1’s three putative
substrate binding sites without altering its active site. This mode of binding is sufficient
for spinophilin to restrict PP1’s activity toward a model substrate in vitro without affecting
its ability to dephosphorylate its neuronal substrate, glutamate receptor 1 (GluR1). Thus,
our work provides the molecular basis for the ability of spinophilin to dictate PP1
substrate specificity.
2.2. Introduction.
Transient protein modification is an essential regulatory mechanism for many biological
processes. One type of transient modification, phosphorylation, is predicted to occur on
approximately one-third of proteins encoded in the human genome2,69. In most
eukaryotes, the number of enzymes responsible for the phosphorylation (kinases) and
dephosphorylation (phosphatases) of tyrosine residues is nearly identical, ensuring high
specificity of their critical functions24. However, the ratio of human serine/threonine
kinases to phosphatases is well in favor of the kinases (428:~40). As the majority of
serine/threonine phosphorylation sites are reversible, the specificity of the phosphatases
appears to be low compared to that of serine/threonine kinases.
Protein phosphatase 1 (PP1) is the most widely expressed and abundant
serine/threonine phosphatase. PP1 is a single-domain protein that is exceptionally well
25
conserved from fungi to humans, in both sequence and function. Dephosphorylation
events by PP1 regulate cell-cycle progression, protein synthesis, muscle contraction,
carbohydrate metabolism, transcription and, of specific interest for this work, neuronal
signaling34. The structure of apo PP1 is not known, largely because of its high instability
in solution
27
. However, the structure of PP1 bound to tungstate70 or various PP1
inhibitors27,26,29 shows that the catalytic site of PP1, which contains two metal ions, is at
the intersection of three putative substrate binding regions, referred to as the
hydrophobic, acidic and C-terminal grooves.
Although the specificity of serine/threonine phosphatases appears low, PP1 is
nevertheless able to dephosphorylate its numerous targets with high specificity. To
achieve this specificity, PP1 interacts with a large number of regulatory proteins (~200
confirmed interactors)33. PP1 regulatory proteins include inhibitory proteins that keep
PP1 in an inactive state and targeting proteins that form highly specific holoenzymes25.
Targeting proteins direct the specificity of PP1 by localizing it to its point of action within
the cell as well as by directly altering its substrate preferences34.
Most PP1 regulatory proteins≥95%)
(
contain a primary PP1 bind
ing motif
(R/K)(R/K)(V/I)X(F/W), commonly referred to as the RVXF motif, that interacts with a
binding region more than 20 Å away from the active site of PP134,71. A structure of PP1
with an RVXF peptide has been described35; however, it provides only limited insights
into the regulation of PP1, as this interaction seems to be identical for most if not all PP1
complexes. There is a limited number of holoenzyme structures currently available,
resulting from the instability of apo PP1 in solution27 and the high flexibility of most PP1
regulatory proteins57, which makes crystallography exceedingly challenging. Only a
single structure each of an inhibitor–PP1 (inhibitor-2–PP162) complex and a targeting
protein–PP1 (MYPT1 regulatory subunit–PP143) complex has been reported. These
structures provide the first insights into the regulation of PP1, but a detailed
26
understanding of the molecular basis for the ability of targeting proteins to direct the
substrate specificity of PP1 is still lacking.
To this end, we used a combination of NMR spectroscopy, X-ray crystallography
and biochemistry to elucidate how spinophilin, the most extensively studied neuronal
PP1 targeting protein, binds and directs PP1 substrate specificity. Spinophilin, an 817residue neuronal regulatory protein (Figure 13a), targets PP1 to neuronal synapses42,
where it controls essential neuronal processes such as AMPA receptor activation50 and
cytoskeletal reorganization72,73 and therefore plays a decisive role in learning and
memory formation. Furthermore, spinophilin has been shown to have an important role
in the actions of drugs of abuse73,74, and changes in PP1 function have been associated
with Parkinson’s disease20. Here we show that the PP1 binding domain of spinophilin is
highly dynamic in its unbound state. We also show that this flexibility enables spinophilin
to interact with PP1 over an extensive surface, forming many unexpected interactions.
These results reveal a previously unknown mechanism for the regulation of PP1
substrate specificity, whereby spinophilin binds to PP1 and blocks one of three potential
PP1 substrate binding grooves without altering its active site. This work provides
fundamental new insights into the regulation of the substrate specificity of PP1 at the
molecular level.
2.3. Results.
2.3.A. The spinophilin PP1 domain is unstructured in solution.
It was previously shown that the PP1 binding domain of spinophilin, residues 417–494,
is necessary and sufficient for its complete interaction with PP168. Using NMR
spectroscopy, we investigated the three-dimensional (3D) structure of unbound
spinophilin417–494 polypeptide and showed that it is highly dynamic in solution. A twodimensional [1H,15N] heteronuclear single-quantum coherence (2D [1H,15N] HSQC)
27
Figure 13. The unbound spinophilin PP1 binding domain.
(a) Domain-structure map of spinophilin with the conserved RVXF motif highlighted in green,
the PP1 binding domain in magenta and the PDZ domain in purple (color codes are constant
throughout all figures). (b) 2D [1H,15N] HSQC spectrum of spinophilin417–494. The blue bar
highlights the lack of dispersion for the chemical shifts in the 1HN dimension. (c) An overlay
of the 2D [1H,15N] HSQC spectra of spinophilin417–602 (purple) and spinophilin417–494
(magenta). The blue bar highlights the larger dispersion for the chemical shifts in the 1HN
dimension resulting from the well-ordered PDZ domain, whereas the chemical shifts of
spinophilin417–494 remain unchanged. Red asterisks indicate side chain NH2 groups of
asparagine and glutamine. (d,e) secondary-structure propensity scores (d) and 15N[1H]-NOE
(hetNOE) (e) data are plotted against spinophilin residue numbers. These data indicate that
there are no regions of transient structure nor reduced backbone motions in spinophilin417–
494. Also, no areas of considerably populated transient secondary structure were detected.
Regions of considerably populated transient structure were considered five residues or more
with secondary-structure propensity score >0.2, indicating a transient α-helix, or −0.2,
<
indicating a region with extended structure (not considerably populated regions are indicted
by a gray box).
28
spectrum of spinophilin417–494 contained little chemical-shift dispersion in the 1HN
dimension and little amino acid–specific clustering in the 15N dimension, indicative of a
highly flexible unstructured domain commonly referred to as an intrinsically disordered
protein75 (Figure 13b). In addition, this flexibility was not altered when the PP1 binding
domain was associated with its C-terminal structured PDZ domain (spinophilin417–602;
Figure 13c).
We
measured
carbon
chemical
shifts,
15N
autocorrelated
relaxation
measurements and paramagnetic relaxation enhancements to test for potential preferred
secondary structures or a preferred 3D topology. However, in marked contrast to our
studies of PP1 inhibitors57, the other class of PP1 regulatory molecules, it was not
possible to identify any preferred secondary structures or a preferred 3D topology for
unbound spinophilin417–494 (Figure 13d,e). Taken together, these data directly show
that spinophilin417–494 resembles a random-coil polypeptide in solution.
Nevertheless, the unstructured spinophilin PP1-binding domain is functional and
interacts strongly with PP1 (we used the α-isoform for all studies here). Isothermal
titration calorimetry (ITC) measurements using spinophilin417–583 and our recombinant
PP1 indicated a dissociation constant (Kd) of 8.7 nM (Figure 14a), which agrees well
with data available for the interaction of spinophilin with PP1 from natural source68. Thus,
our recombinant spinophilin and PP1 preparations behave similarly to endogenous
proteins, confirming that the data reported here reflect the biologically relevant
holoenzyme (spinophilin417–583–PP1 complex).
2.3.B. The crystal structure of the spinophilin–PP1 holoenzyme.
29
To elucidate how spinophilin directs PP1 substrate specificity at a molecular level, we
determined the 1.85-Å crystal structure of the complex between spinophilin417–583 and
PP1α7–330 (Figure 14b, Figure 69 and Table 2). The overall structure of PP1
Figure 14. The spinophilin417-583-PP1α7-330 complex.
(a) Isothermal titration calorimetry of purified spinophilin417–583 and PP1α7–330,
confirming that the intrinsically unstructured spinophilin PP1 binding domain is active. (b)
Cartoon representation of the spinophilin417–583 PP1 binding (magenta) and PDZ domains
(purple) and PP1α7–330 (gray, surface representation) complex; centered on the active site
of PP1α7–330. Two Mn2+ ions (pink spheres) mark the active site of PP1. The newly formed
secondary-structure elements of spinophilin417–494 are labeled β1, β2 and α1. The PDZ
domain is C-terminal to α1. Spinophilin residues interacting with the RVXF binding pocket
are represented as sticks in green and those interacting with the PP1 C-terminal groove are
represented as sticks in cyan. (c) The spinophilin PP1 binding domain before (unstructured
model, left) and after binding to PP1 (crystal structure without PP1, right), illustrating the
unfolded-to-folded transition. The four unique regions that characterize the interaction of
spinophilin (magenta) with PP1 are numbered; RVXF, green; C-terminal groove binding
motif, cyan.
is essentially identical to that of reported PP1 structures26,28. The catalytic site of PP1 is
at the intersection of three potential substrate binding regions, which are referred to as
the hydrophobic, acidic and C-terminal grooves. Two metal ions (Mn2+ in the
recombinant protein, as the protein was expressed in LB medium supplemented with
MnCl2 but likely Fe2+ and Zn2+ in native PP126) that are essential for the catalytic activity
of PP1 are present in the active site of the spinophilin–PP1 holoenzyme. The PP1 binding domain of spinophilin, which is unstructured when not bound to PP1, becomes
restricted to a single conformation with residues 424–489 visible in both molecules of the
asymmetric unit (Figure 14c and Figure 69). Thus, the spinophilin PP1 binding domain
30
entirely folds upon binding to PP1. We also determined the 3D structure of the neurabin–
PP1 holoenzyme (Figure 70 and Table 13). Neurabin is the neuron-specific isoform of
spinophilin (neurabin-II), and spinophilin and neurabin have ~80% primary sequence
identity. We observed no major differences between the two 3D structures.
Strikingly,
Data collection
Space group
Cell dimensions
a, b, c (Å)
α, β, γ (°)
Resolution (Å)
Rmerge
I/σI
Completeness (%)
Redundancy
Refinement
Resolution (Å)
No. reflections
Rwork/Rfree
No. atoms
Protein
Ligand/ion
Waters
B-factors
Protein
Ligand/ion
Water
R.m.s. deviations
Bond lengths (Å)
Bond angles (°)
Spinophilin417–
583–PP1α7–330
Spinophilin417–583–
PP1α7–330–
nodularin-R
C2
C2
119.7, 84.4, 109.2
90.0, 93.5, 90.0
50.0–1.85 (1.88–
1.85)
0.078 (0.596)
10.9 (2.2)
99.3 (97.9)
3.7 (3.1)
119.4, 84.4, 109.3
90.0, 93.6, 90.0
50.0–2.0 (2.07–2.00)
27.5–1.85
91,551/4,548
0.179/0.211
40.0–2.00
70,101/3,528
0.192/0.234
6,430
4
521
6,372
113
481
25.6
22.0
33.0
29.6
29.3
35.9
0.011
1.26
0.015
1.54
0.066 (0.281)
13.6 (4.1)
98.1 (82.5)
3.6 (2.6)
spinophilin417–583
interacts
with PP1 in an unexpected and
unique
manner,
occupying
the
not
only
RVXF-motif
binding pocket (Figure 15a) but
also
forming
interactions
multiple
with
different
regions of PP1, including a
substantial part of the PP1Cterminal groove. The highly
dynamic nature of spinophilin in
its
unbound
is
forming
likely
Table 2. Data collection and refinement statistics
for the spinophilin417-583-PP1α7-330 and the
spinophilin417-583-PP1α7-330-nodularin-R
complexes.
essential
One crystal was used for each structure. Values in parentheses are
for highest-resolution shell.
interface of 3,926 Å2 between
substantial
for
state
the
intermolecular
spinophilin and PP176, ~2.5 times larger than the average interface for protein–protein
complexes77 and with ~16% of the surface of the globular catalytic domain of PP1
buried.
2.3.C. Spinophilin binds PP1 at multiple regions.
31
The interaction of spinophilin with
PP1 can be divided into four unique
regions. Region I is formed by
spinophilin
residues
447–451
(RKIHF, the RVXF motif), which
interact in the PP1 RVXF binding
groove (Figure 15a,b); this is the
only expected interaction, based on
previous work. As seen in the
MYPT1
and
regulatory
subunit–PP143
inhibitor-2–PP1
structures,
residues
62
complex
of
the
Figure 15. Spinophilin’s interaction with the
PP1 RVXF binding pocket.
(a) 90° rotation of Figure 8b to highlight the
interaction of Ile449 and Phe451 in the RVXF
binding pocket. (b) Spinophilin residues (RVXF
sequence, green) and PP1 (gray, surface and stick
representation for interacting residues) are shown
with oxygen atoms in orange and nitrogen atoms in
dark blue. Hydrogen bonds are represented as
black dashes. The spinophilin RVXF sequence
(447RKIHF451) is preceded by a long loop that
folds back on itself (Figure 8b,c) to form a strong
hydrogen bond network. Some residues in the loop
have been omitted for clarity.
spinophilin RVXF motif bind in an extended conformation, with spinophilin residues
Ile449 and Phe451 buried in the PP1 RVXF binding groove. However, in contrast to all
previously described interactions, the histidine residue at position X of the spinophilin
RVXF motif forms a hydrogen bond with Thr288 in PP1. In neurabin, the residue
occupying the X position is a lysine, which also forms an identical polar contact with PP1
(Figure 71a). This indicates that, for both the spinophilin–PP1 and neurabin–PP1
holoenzymes, the residue at position X unexpectedly plays a role in the RVXF motif
binding interaction.
Interaction regions II and III of spinophilin undergo what is commonly referred to
as ‘coupled folding and binding’78 (Figure 14c). In region II, spinophilin residues 430–434
and 456–460 fold to form two β-strands that extend the PP1 β-sheet 1 into an extensive
7-strand β-sheet (Figure 14b,c and Figure 16a). In region III, spinophilin residues 476–
492 fold into a four-turn α-helix that forms electrostatic and hydrophobic interactions with
32
the surface of PP1 and is adjacent to both the C-terminal and hydrophobic grooves
(Figure 14b,c and Figure 16b). We did not observe the β-strands or α-helix in unbound
spinophilin; thus, they require PP1 to fold.
Region IV is formed by spinophilin residues 462–469, which bind into a
substantial part of the PP1 C-terminal groove (Figure 14b,c and Figure 16c; Figure 71b
for neurabin). Here spinophilin Arg469, which has the second-highest buried surface
area of any spinophilin residue (only Ile449 of the RVXF motif has more), forms the core
Figure 16. Spinophilin’s PP1 interaction region II, III, and IV.
(a) Region II: spinophilin forms two novel β-strands upon binding PP1. Stick representation of
spinophilin residues 428–435 (β1 includes residues 430–434) and 454–461 (β2 includes
residues 456–460), which form an extended β-sheet with the strands β13 and β14 of PP1. (b)
Region III: spinophilin forms a four-turn α-helix upon interaction with PP1. View of the helix
highlighting electrostatic and hydrophobic interaction clusters between the helix and the
surface of PP1. (c) Region IV: spinophilin residues 462–469 interact directly with the Cterminal groove of PP1. These residues protrude directly into the binding pocket, forming a
complex hydrogen bonding network both with PP1 and within spinophilin. Spinophilin residues
465 and 468 have been omitted for clarity.
of this interaction. It does so by establishing strong backbone and side chain hydrogen
bond interactions both intramolecularly, with Tyr462 and Asn464 in spinophilin, and
intermolecularly, with Asp71 in PP1. These interactions prime spinophilin residues
Tyr462 and Tyr467 to form extensive hydrophobic interactions with PP1.
To understand the influence of each of the four spinophilin–PP1 interaction
regions, we created multiple single-residue mutants of the spinophilin PP1 binding
domain and measured their ability to influence binding to PP1. None of the introduced
33
mutations disrupted the overall fold of spinophilin, as indicated by circular dichroism
spectrum analysis (Figure 72). The mutations that most negatively impacted PP1 binding
affected residues in the spinophilin RVXF motif (interaction region I, F451A) and the Cterminal groove binding motif (interaction region IV, Y467A, R469A, R469D), with a
smaller negative effect observed for residues in the β-sheet area (interaction region II,
F459A) (Figure 17a,b). Notably, the hydrophobic phenylalanine residue in the RVXF
motif (Phe451) had the most negative effect on PP1 binding, in excellent agreement with
the reported RVXF-motif mutational analysis of other regulator–PP1 complexes33,35.
Single point mutations in the α-helix had the least influence on PP1 binding.
2.3.D. Spinophilin does not affect the PP1 active site.
Despite its extensive interaction surface, spinophilin does not bind PP1 near its active
site and appears to leave the active site unchanged. We verified this by determining the
2.0-Å crystal structure of the spinophilin–PP1–nodularin-R triple complex. Nodularin-R is
a molecular toxin that inhibits PP1 by direct active-site interaction27. Comparison of this
structure with the PP1–nodularin-R cocomplex structure27 (PDB 3E7A) reveals that
spinophilin does not alter the binding interactions of molecules directly at the active site
of PP1, nor does it alter the PP1 hydrophobic substrate binding groove (Figure 17c and
Figure 73).
2.3.E. Spinophilin-PP1 interactions define PP1 specificity.
Despite the fact that the interaction of molecules at the active site of PP1 is unchanged,
spinophilin directs PP1 substrate specificity in a manner independent of its ability to
target PP1 to distinct subcellular locations, a phenomenon that has also been described
for other PP1 targeting proteins35,79. This is shown in Figure 17d, which illustrates the
decisively altered dephosphorylation efficiency of PP1 in the presence and absence of
spinophilin417–583. We used the following PP1 substrates: (i) Ser845 protein kinase A
34
(PKA)-phosphorylated GluR1809–889, a specific substrate of the spinophilin–PP1
holoenzyme17 (GluR1809–889); (ii) phosphorylase a, a specific substrate of the GM–
PP1 holoenzyme and not the spinophilin–PP1 holoenzyme80; and (iii) p-nitrophenyl
phosphate (pNPP), a commonly used model substrate. Spinophilin inhibited the activity
of PP1 for phosphorylase a (red), whereas it had no effect on the activity of PP1 for
GluR1 (blue) and pNPP (green). Notably, this inhibition could be completely ablated by a
mutation of either spinophilin Phe451 or Arg469 (Figure 17d), clearly showing that both
interactions are necessary for the formation of the functional holoenzyme. The unaltered
dephosphorylation efficiency of PP1 for the small-molecule substrate pNPP lends further
support to the observation that spinophilin does not alter the binding of molecules at the
active site of PP1. Therefore, the striking change in substrate specificity of the
spinophilin–PP1 holoenzyme must result from a modification of the PP1 substrate
binding surface. This can be achieved by an alteration of the surface electrostatics, by
steric blocking of essential binding sites or both.
Although the interaction surface between PP1 and spinophilin is extensive,
spinophilin binding does not substantially alter the charge distribution of the enzyme
(Figure 17e). Instead, spinophilin directs PP1 specificity by a different mechanism, via
steric inhibition of alternative substrate binding sites. Spinophilin binds a large part of the
PP1 C-terminal groove, blocking access to substrates that require this groove for binding
to PP1. Indeed, mutation of Asp71 (Figure 16c), a PP1 residue located in the center of
the C-terminal substrate binding pocket, led to a substantial increase in the KM for
dephosphorylation of phosphorylase a81, highlighting its role in substrate binding. In the
spinophilin–PP1 holoenzyme, Asp71 forms a hydrogen bond with spinophilin Arg469
and is completely inaccessible to solvent. To further investigate the role of Asp71 in
substrate selectivity, we tested the activity of PP1 D71N against phosphorylated
GluR1809–889 and phosphorylase a (Figure 17f). We found the expected decrease by a
35
Figure 17. Spinophilin creates a unique holoenzyme with novel substrate specificity.
(a) The binding of spinophilin WT and mutants (region I: I449A, H450A, F451A; region II:
Q457A, F459D, T461A; region III: E482A, E486A; region IV: Y467A, R469A, R469D) to
PP1α7–330 was determined using a capture assay. Elution fractions and all loaded
spinophilin samples were subjected to SDS-PAGE and Coomassie blue staining. (b)
Densitometry analysis of gels as shown in a, to determine the amount of spinophilin coeluted
with PP1α7–330, normalized to WT. The experiment was repeated twice. Error bars, s.d. (c)
Overlay of the active site of spinophilin417–583–PP1α7–330–nodularin-R (orange) with
PP1α7–330–nodularin-R (cyan; PDB 3E7A). (d) PP1α7–330 dephosphorylation of substrates
in vitro: pNPP (green) GST-GluR1809–889 (blue) and phosphorylase a (red). Error bars, s.d.
(n = 4). (e) Electrostatic surface representation (red, negative charge; blue, positive charge;
gray, hydrophobic) of PP1α7–330 and spinophilin417–583–PP1α7–330 centered on the
active site. (f) Mutation of PP1 Asp71 inhibits the dephosphorylation of phosphorylase a.
PP1α7–330 WT and D71N dephosphorylation of GST-GluR1809–889 (blue) and
phosphorylase a (red) in vitro. Error bars, s.d. (n = 4). (g) The PP1 D71N mutant reduces
spinophilin binding. WT spinophilin binding to WT PP1 (blue) and PP1 D71N mutant (red) was
tested. Error bars, s.d. (n = 2).
36
factor of 3.0 in the activity of PP1 for phosphorylase a, nearly identical to the decrease
by a factor of 2.9 that we identified when using the spinophilin–PP1 holoenzyme.
However, mutation of Asp71 had no effect on the dephosphorylation of GluR1809–889.
Further corroborating these results, spinophilin shows ~60% reduced binding compared
to that of PP1 D71N (Figure 17g). Lastly, the only interaction region with 100% primary
sequence identity between the neurabin and spinophilin isoforms is the PP1 C-terminal
groove binding sequence, further highlighting the importance of this interaction.
2.4. Discussion.
Our biochemical, dynamic and structural data of the spinophilin PP1 binding domain
alone and the sphinophilin–PP1 holoenzyme show that spinophilin is an intrinsically
unstructured protein that undergoes a folding-upon-binding transition upon interaction
with PP1. The high intrinsic flexibility of unbound spinophilin allows for unique
interactions with PP1, resulting in an unusually large protein-protein interface.
Spinophilin binds, as anticipated, into the PP1 RVXF binding groove. However,
unexpectedly, spinophilin also binds PP1 in numerous additional surface grooves,
including the C-terminal substrate binding groove. The spinophilin RVXF motif is indeed
critical for PP1 binding, but residues that interact with the PP1 β-strands, as well as the
PP1 C-terminal groove, are also important, with the latter interaction region being
necessary for directing the substrate specificity of PP1.
Taken together, these structural and biochemical data provide a novel explanation for
the reduced dephosphorylation efficiency of the spinophilin–PP1 holoenzyme for
phosphorylase a. The results clearly show the importance of the interaction of
spinophilin with the PP1 C-terminal groove for altering the substrate specificity of PP1.
Based on these data, we propose that apo PP1 is able to bind substrates using any
combination of its three potential substrate binding grooves (Figure 18a). Notably,
spinophilin binds into only one of these three alternative sites, so the acidic and
37
hydrophobic substrate binding grooves are still accessible. However, in the presence of
spinophilin, PP1 can only bind (and, in turn, select) substrates that interact with the
hydrophobic and/or acidic grooves (Figure 18b); substrates that require the C-terminal
groove for substrate binding, such as phosphorylase a, are blocked. This mechanism for
altered substrate specificity of PP1 is different from that reported for the MYPT1–PP1
holoenzyme, which was proposed to be driven by altered electrostatics43. However,
when MYPT1 binds to PP1, a modified substrate binding surface with extended
Figure 18. Spinophilin affects the PP1 substrate binding surface without changing the
active site.
(a) Model for substrate recognition by apo PP1. (b) Model for substrate recognition by the
spinophilin–PP1 holoenzyme. The three PP1 substrate binding grooves are colored yellow.
Potential substrates are shown as green, purple and black lines. Spinophilin achieves
substrate specificity by steric exclusions of substrate binding sites.
substrate grooves is formed (Figure 74). Taking the data from these two structures, it is
likely that PP1 can achieve substrate specificity via a substrate exclusion process, which
is exemplified by spinophilin–PP1, as well as by a combined substrate inclusionexclusion process, which is exemplified by MYPT1–PP1.
In addition, this is the first time, to our knowledge, that an intrinsically disordered
protein (spinophilin) has been shown to regulate enzymatic activity by directing the
substrate specificity of an enzyme instead of activating or inhibiting it directly. Whereas
the PP1 binding domain of spinophilin behaves as a random-coil peptide in solution, it
38
completely folds upon binding to PP1 and becomes rigid. The results obtained here,
taken together with the targeting of spinophilin–PP1 to the postsynaptic aspect of the
synapse, will ensure that PP1 is placed in an active state close to its selected substrate,
GluR1.
Further comparison of the spinophilin–PP1 holoenzyme structure with the two
previously reported regulatory protein–PP1 holoenzymes (inhibitor-2–PP1 and Mypt1–
PP1) shows that the binding mode for all three proteins is unique and that the only
shared interaction is the conserved RVXF binding groove43,62. Thus, this comparison,
combined with other biochemical data82, suggests that nearly every accessible surface
on the catalytic face of PP1 is a potential protein-interaction site, allowing for numerous
unique interactions and therefore abundant unique holoenzymes, each with individual
substrate preferences. These observations begin to explain the striking diversity of PP1
holoenzymes, each of which may form a truly unique enzyme with distinctive properties.
This is made even more intriguing by the fact that the number (~200) of identified PP1targeting proteins is still increasing33. If the diversity of interactions observed for PP1 is
conserved
across
serine/threonine
phosphatases,
it
would
allow
the
~40
serine/threonine phosphatases to form hundreds of unique holoenzymes, ensuring that
they are as specific as the 428 known serine/threonine kinases.
2.5. Methods.
2.5.A. Cloning and expression.
We subcloned rat spinophilin417–602, spinophilin417–583, neurabin426–592 and
PP1α7–330 into a modified pET-28a vector, which encodes an N-terminal expression
and hexahistidine tag followed by a tobacco etch virus (TEV) protease cleavage site. We
subcloned spinophilin417–494 into a modified pGEX-based vector, encoding a GST-tag
and a TEV protease cleavage site. We cloned human GluR1809–889 into a modified
39
pETM-30 vector that encodes a hexahistidine-GST tag and TEV protease cleavage site.
We generated mutants using the Stratagene Quikchange mutagenesis kit. We
overexpressed spinophilin, neurabin and GluR1 constructs in Escherichia coli
BL21(DE3) CodonPlus cells. For NMR measurements, we carried out the expression of
uniformly 15N- and/or 13C-labeled protein by growing cells in M9 minimal media
containing 1 g −
l 1 [15N]H4Cl and/or 4 g l−1 [13C]d -glucose (CIL) as the sole nitrogen
and carbon sources. We produced PP1α7–330 and the catalytic subunit α of protein
kinase A (PKA) as previously described27,57. We lysed all cells by high-pressure cell
homogenization (Avestin C3 Emulsiflex).
2.5.B. Protein purification.
We loaded spinophilin417–583, spinophilin417–602 and neurabin426–592 onto a
HisTrap HP column (GE Healthcare), eluted them with buffer containing imidazole,
dialyzed them and incubated them at 4 °C with TEV. We loaded samples onto nickel
nitrilotriacetic acid (Ni-NTA) beads (Qiagen); we further purified the flowthrough using
size-exclusion chromatography (SEC). We performed all SEC purifications using a
Superdex 75 (26/60) column (GE Healthcare). For NMR experiments, we concentrated
spinophilin417–602 to 1 mM in 20 mM NaPO4 (pH 6.5), 50 mM NaCl, 0.02% (w/v)
NaN3, 10% (v/v) D2O and 0.25 mM phenylmethylsulfonyl fluoride (PMSF). We
concentrated spinophilin417–583 to 250 μM, flash froze it in liquid nitrogen and stored it
at −80 °C for subsequent experiments.
We bound spinophilin417–494 to GST resin (Novagen) eluted with buffer containing glutathione, dialyzed it and incubated it at 4 °C with TEV. We removed GST
using a second GST column. We heated protein to 90 °C for 10 min and then removed
precipitated protein by centrifugation at 18,000g. NMR spectra measured before and
after the heat purification confirmed that the topology of spinophilin417–494 was not
affected. We exchanged samples into 20 mM NaPO4 (pH 6.5), 50 mM NaCl, 0.02%
40
(w/v) NaN3, 10% (v/v) D2O buffer and concentrated to 1 mM for [15N]spinophilin417–
494 and 0.67 mM for [15N,13C]spinophilin417–494.
For complex production, we bound PP1α7–330 to Ni-NTA beads. We washed the
column with low-salt buffer and then incubated it with spinophilin417–583 or
neurabin426–592. We eluted complexes with buffer containing imidazole and purified
them by SEC (20 mM Tris, pH 7.5, 50 mM NaCl, 0.5 mM TCEP). We pooled fractions
containing complex, incubated them with TEV at 4 °C overnight (18 h) and further
purified them using SEC. We purified the spinophilin417–583–PP1α7–330–nodularin-R
complex using the identical protocol, with the addition of nodularin-R (Alexis) in a 1:1.25
molar
ratio
immediately
following
TEV
cleavage.
For
ITC
experiments
and
dephosphorylation assays, we purified PP1α7–330 using a single-step Ni-NTA affinity
purification at 4 °C and kept it at concentrations below 5 μM.
We purified GST-GluR1809–889 using a HisTrap HP column. We dialyzed the
eluate overnight (18 h) at 4 °C and purified it using SEC (20 mM Tris, pH 8.0, 50 mM
NaCl, 0.5 mM TCEP). We phosphorylated GST-GluR1809–889 using PKA in 20 mM
Tris, pH 8.0, 50 mM NaCl, 0.5 mM TCEP, 10 mM MgCl2, 200 μM ATP, 1 μM EDTA at
30 °C for 3 h and confirmed phosphorylation using Pro-Q Diamond phosphoprotein gel
stain (Invitrogen).
2.5.C. NMR measurements.
We performed NMR measurements at 298 K on a Bruker Avance II 500 MHz
spectrometer equipped with a TCI HCN z-gradient cryoprobe. We performed chemicalshift referencing, NMR spectra processing, chemical-shift assignments, 15N longitudinal
(R1) and transverse (R2) relaxation rate, {15N}-heteronuclear NOE measurement and
secondary structure propensity calculations as described57.
2.5.D. Crystallization.
41
We concentrated the purified complexes to 4.9 mg ml
−1 (spinophi lin417–583–PP1α7–
330), 4.6 mg ml−1 (spinophilin417–583–PP1α7–330–nodularin-R) and 3.0 mg ml−1
(neurabin426–592–PP1α7–330). We crystallized complexes using the sitting-drop vapor
diffusion method at 4 °C, using a 2:1 ratio of protein to crystallization solution except for
spinophilin417–583–PP1α7–330, for which a 1:1 ratio was optimal. Crystals grew in 0.2
M NaCl, 0.1 M MES (pH 6.5), 10% (w/v) PEG 4000 (spinophilin417–583–PP1α7–330);
0.1 M MES (pH 6.5), 15% (w/v) PEG 550 MME (spinophilin417–583–PP1α7–330–
nodularin-R) or 0.2 M diammonium hydrogen phosphate, 20% (w/v) PEG 3350
(neurabin426–592–PP1α7–330). We cryoprotected crystals using crystallization solution
supplemented
with
25%
(v/v)
glycerol
(spinophilin417–583–PP1α7–330
and
neurabin426–592–PP1α7–330) or 20% (v/v) glycerol (spinophilin417–583–PP1α7–330–
nodularin-R).
2.5.E. Data collection and structure determination.
We collected data for all complexes at the Brookhaven National Laboratories National
Synchrotron Light Source Beamline X6A at 100 K and a wavelength of 1 Å using an
ADSC QUANTUM 210 CCD detector. We indexed, scaled and merged data using
HKL2000 0.98.692i83.
We phased the spinophilin417–583–PP1α7–330 1.85-Å data using molecular
replacement (Phaser 1.3.2 84) with the coordinates of PP1α7–330 27 (PDB 3E7A) and the
PDZ domain of spinophilin52 (PDB 2G5M) as search models. We obtained a solution for
two copies of PP1 and one copy of the spinophilin PDZ domain. We observed clear
electron density for both spinophilin PP1 binding domains.
We phased the spinophilin417–583–PP1α7–330–nodularin-R 2.00-Å data using
difference Fourier synthesis using the coordinates of the spinophilin417–583–PP1α7–
330 complex (without ligand and water molecules). We observed clear electron density
42
for nodularin-R. The REFMAC5 library for nodularin-R was identical to that used for
structure determination of the PP1α7–300–nodularin-R complex5 (3E7A).
We phased the neurabin426–592–PP1α7–330 2.2-Å data using molecular
replacement (Phaser 2.1.1) with the coordinates of PP1α7–330 and the PDZ domain of
spinophilin (3EGG) as search models. We obtained a solution for two copies of PP1 and
one copy of the spinophilin PDZ domain. We observed clear electron density for both
neurabin PP1 binding domains and for the residue changes in the neurabin PDZ
domain.
We completed all models by cycles of manual building using the program Coot85
coupled with refinement using REFMAC5.2.001986. We added water and ligand
molecules during the final stages of model building; we performed the final rounds of
refinement using REFMAC5.2.0019 with TLS87, using a TLS model identified with the
TLSMD server88. We produced all structural figures using PyMOL (DeLano Scientific).
We analyzed the stereochemical quality of the model using the programs MolProbity89
and PROCHECK90.
2.5.F. Dephosphorylation reactions.
We prepared phosphorylase a (Sigma) as described91. Before the dephosphorylation
reactions, we incubated PP1α7–330 alone or in the presence of spinophilin417–583 at
room temperature 22 °C for 15 min. We initiated dephosphorylation by the addition of
either GST-GluR1809–889 or phosphorylase a. We incubated reactions for 30 min at 30
°C and quenched them by the addition of SDS-PAGE loading buffer and boiling for 10
min at 100 °C. We separated samples by SDS-PAGE, and we measured
phosphorylation using Pro-Q Diamond phosphoprotein gel stain (Invitrogen) and
visualized it using a Typhoon 9410 (Amersham Bioscience) with excitation at 532 nm
and emission at 560 nm. We measured total protein using Sypro Ruby protein gel stain
43
(Invitrogen). We quantified band densities using ImageQuant TL (GE Healthcare). We
corrected Pro-Q results for total protein as previously described92.
We initiated dephosphorylation of pNPP by addition of pNPP to a final concentration of 2 mM. We incubated reactions for 30 min at 30 °C and quenched them by
addition of 1 M NaOH. We determined the amount of p-nitrophenolate product from the
absorbance at 405 nm using a molar extinction coefficient of 18,000−1Mcm−1. We
normalized all data to no PP1 controls and subtracted from 1 to give relative
dephosphorylation.
2.5.G. Capture assay.
We bound hexahistidine-tagged PP1α7–330 to Ni-NTA beads (13 parallel 1-ml columns
with fresh Ni-NTA beads (Qiagen)) and extensively washed them. We incubated
spinophilin417–583 with Ni-NTA–bound PP1α7–330. We washed the column to remove
all unbound spinophilin and eluted spinophilin–PP1 complexes with buffer containing
imidazole. We subjected samples to SDS-PAGE gel electrophoresis analysis using
Coomassie blue staining.
2.5.H. Isothermal titration calorimetry.
We dialyzed spinophilin417–583 and PP1α7–330 overnight (18 h) into 20 mM
Tris, pH 7.5, 500 mM NaCl, 0.5 mM TCEP at 4 °C. We titrated spinophilin417–583 (15
μM) into PP1α7–330 (1.25 μM) using a VP-ITC microcalorimeter at 23 °C (Microcal,
Inc.). We analyzed data using Origin 7.0 (Originlab).
Chapter 3. The PP1:GM Holoenzyme.
All of the experiments in this section were performed by M.J.R.
44
45
3.1. Introduction.
PP1 plays a central role in the regulation of glycogen metabolism25. Glucose provides
the primary energy source for the body and excess glucose is stored as glycogen, a
branching chain of glucose molecules that can be broken down as needed4. The balance
between glycogen synthesis and breakdown is an integral process that is regulated by
complex signaling networks. Extracellular hormones, such as insulin and glucagon,
respond to blood glucose levels and activate glycogen regulatory proteins93. These
proteins lead to the essential activation or inactivation of metabolic enzymes, which
control the synthesis and breakdown of glycogen. Improper regulation of this pathway
results in many diseases, including diabetes and hypoglycemia4. Many of the metabolic
Figure 19. The chemical reactions of glycogen synthesis and breakdown.
A) Glycogen synthesis is catalyzed by glycogen synthase. B) Phosphorylase catalyzes
glycogen breakdown.
enzymes involved in this process are directly tethered to glycogen, which forms large
protein glycogen complexes, ~20-30 nm in size, that are referred to as either
glycosomes, glycogen granules or glycogen particles94. Many of these enzymes,
including the rate limiting enzymes for glycogen synthesis, glycogen synthase, and for
glycogen breakdown, phosphorylase, are regulated by changes in their phosphorylation
state4.
46
Glycogen synthase catalyzes the formation of glycogen by adding glucose from
uridine diphosphate(UDP)-glucose to glycogen (Figure 19 A)4. Phosphorylase catalyzes
the breakdown of glycogen by phosphorylsis, cleavage by the addition of a phosphate
moiety, producing glucose-1-phosphate (Figure 19 B)4. Interestingly, phosphorylation of
these two enzymes has an inverse effect on their activity. Glycogen synthase becomes
phosphorylated on up to 9 different serine or threonine residues by the kinases PKA,
GSK3, CaMK, CK1 and CK293,95. Each additional phosphorylation event results in
progressive inactivation of the enzyme. In contrast, phosphorylase contains only a single
phosphorylation site at serine 14 that is phosphorylated by phosphorylase kinase93.
Phosphorylation of this residue results in the activation of the enzyme. Therefore, in
general, phosphorylation of metabolic enzymes drives glycogen breakdown and
dephosphorylation drives glycogen synthesis. The major phosphatase responsible for
the dephosphorylation of these enzymes is PP1 which becomes directly targeted to
glycogen granules by a family of targeting proteins called the G subunits34.
The G subunit family of PP1 targeting proteins includes seven genes, but the
protein product of only four of these genes have been investigated: GM, GL, PTG (GC),
and R6 (GD)25. GL, PTG, and R6 are similar in size, ranging from 33-36 kDa, but GM is
significantly larger at 124 kDa25. All four proteins contain a PP1 interaction region and a
carbohydrate binding motif (CBM) domain, which interacts with glycogen to localize PP1
to glycogen granules (Figure 20)25. The best characterized proteins from this family are
GM, which targets PP1 to glycogen in skeletal muscle, and GL, which targets PP1 to
glycogen in liver96,97. In contrast to spinophilin, which inhibits the activity of PP1 for
specific substrates, including glycogen metabolism enzymes, the G subunits either
target towards or enhance the activity of PP1 for glycogen metabolic enzymes56. The
enhancement of PP1 activity is believed to occur through the interaction of PP1
47
substrates with specific protein interaction regions within the CBM domain56.
Interestingly, these interaction regions are not fully conserved across all G subunit family
members. Therefore, different G subunit family members interact with different PP1
substrates. While, GM and GL interact directly only with glycogen synthase, PTG
interacts directly with glycogen synthase, phosphorylase and phosphorylsae kinase56,98.
The differences in substrate binding, likely allow the G subunit family members to
differentially alter the substrate specifiity of PP1.
Figure 20. Domain comparison of the G subunit family of regulators of PP1.
Domain maps are shown for GM (A) GL (B) PTG (C) and R6 (D). The RVxF motif is shown in
green and the CBM domain is shown in blue with residue numbers labeling the C and Nterminus of the RVxF motif and CBM domain.
In 1997, the laboratory of Dr. Barford reported the structure of the GM RVxF
peptide bound to PP1 (Figure 21)35. This structure provided the first insight into the
interaction of PP1 with its regulatory proteins in general, and at the RVxF motif
specifically. However, this structure provides little information about the mechanism for
altering the substrate specificity of PP1 by the G subunits. Additionally, since the CBM
domains act as substrate interaction domain, questions about the mechnism by which
48
the G subunits alter the substrate specificity of PP1 must be answered: Do the CBM
domains of GM and GL extend substrate binding grooves on PP1, or do the recently
identified substrate interaction sites on the CBM domain act independent from PP1? By
exploring the regulation of PP1 initially by GM, we will gain critical understand of the
mechanism by which a targeting protein enhances the activity of PP1 for its substrates.
Figure 21. The PP1:GM RVxF peptide crystal structure.
PP1 is shown as a surface representation in grey with the water molecules that were
2+
built in the position of Mn ions shown as pink spheres. The GM RVxF peptide is
shown in green.
The aim of this project is to determine the structure of the PP1:GM holoenzyme. A
direct comparison of the PP1:GM structure with the PP1:spinophilin structure can begin
to explain the differences and similarities between the mechanisms for inhibiting and
enhancing the activity of PP1 for its substrates. Here, we describe the initial structural
and biochemical analysis of this important regulator of PP1. This includes a detailed
characterization of the interaction between PP1 and GM, a protocol for the production of
the PP1:GM complex, and the initiation of the structure determination of the CBM domain
using NMR spectroscopy.
3.2. Methods.
3.2.A. Protein expression.
49
Rabbit GM 2-240, in the expression vector PGEX 4T2, was obtained from Dr. David
Brautigan (University of Virginia)56. Constructs corresponding to GM 2-237, 2-93, and 63237 were subcloned into pETM30-MBP and RP1B99. pETM30-MBP encodes a His6
purification tag, MBP solubility tag and a TEV (tobaco etch virus protease) cleavage site.
The RP1B vector encodes a Thio6His6 expression and purification tag, as well as a TEV
cleavage site. All constructs were transformed into E. coli BL21 (DE3) RIL cells
(Stratagene). Freshly transformed cells were grown in luria broth (LB) shaking at 250
rpm at 37°C until an OD600 of 0.6. The incubator was cooled to 18°C and protein
expression was induced by the addition of 1 mM IPTG. Cells were grown overnight at
18°C under vigorous shaking (250 rpm). The next morning cells were harvested by
centrifugation (10 min, 6000 xg, 4°C) and stored at -80°C.
For PP1 expression, E. coli BL21 (DE3) cells (Invitrogen) were co-transformed
with rabbit PP1α 7-330 in RP1B and the vector pGro7 (Takara), which encodes the
GroEL/ES chaperone. Cells were grown at 30°C under vigorous shaking (250 rpm) in LB
supplemented with 1 mM MnCl2, which is required for the proper folding and activity of
PP1, until an OD600 of 0.5 was reached. Arabinose was added at 2 g/L of LB to induce
the expression of the chaperone. Cells were subsequently grown to an OD600 of 1.0 and
then placed in ice water baths. The incubator was cooled to 10°C, and expression of
PP1 was induced by the addition of 0.1 mM IPTG. Cells were grown under vigorous
shaking (250 rpm) at 10°C overnight. After ~20 hours cells were harvested by
centrifugation (10 min, 6000 xg, 4°C). Cells were resuspended in fresh LB containing
1 mM MnCl2 and 200 μg/ml of chloramphenicol. Cells were incubated at 10°C under
vigorous shaking (250 rpm) for two hours to allow for additional refolding. Cells were
harvested by centrifugation (10 min, 6000 xg, 4°C) and stored at -80°C.
3.2.B. Purification of GM and PP1:GM.
50
Thio6His6 tagged GM 2-237 from RP1B and His6-MBP tagged GM 63-237 from pETM30MBP were purified as follows. Cells were lysed using high pressure cell homogenization
(Avestin EmulsiFlex C-3) and cellular debris was removed by centrifugation (45,447 xg,
45 min, 4°C). Soluble protein was filtered using a 0.22 μm filter (Millipore) and loaded
onto a His trap column (GE Healthcare). The column was washed with buffer containing
20 mM Tris pH 8.0, 500 mM NaCl and 5 mM imidazole. Protein was eluted using a
gradient of 5-500 mM imidazole. Fractions were subjected to SDS-PAGE to confirm the
presence of protein. Fractions containing GM were pooled, TEV was added and the
sample was dialyzed overnight at 4°C against a buffer containing 20 mM Tris pH 8.0,
500 mM NaCl. Cleaved protein was added to Ni-NTA resin (Qiagen). The flowthrough
and 5 mM imidazole wash were collected and pooled. For GM 2-237 the sample was
concentrated to ~5 ml, filtered through a 0.22 μm filter and subjected to SEC using a
Superdex 75 (16/60) column (GE Healthcare) equilibrated with 20 mM Tris pH 7.3, 250
mM NaCl and 0.5 mM TCEP. Fractions were subjected to SDS-PAGE to confirm the
presence of protein. Fractions containing protein were pooled, concentrated to ~500 μl,
and 1D 1H NMR spectra were recorded to test if GM 2-237 is folded. For GM 63-237, the
sample was concentrated to ~10 ml, filtered through a 0.22 μm filter and subjected to
SEC using a Superdex 75 (26/60) column (GE Healthcare) equilibrated with SEC buffer
(20 mM Tris pH 7.5, 50 mM NaCl). Fractions containing protein were pooled,
concentrated to ~50 μM, frozen in liquid nitrogen and stored at -80°C for use in PP1:GM
complex formation.
His6MBP-GM 2-237, 63-237 and 2-93 were purified for use in isothermal titration
calorimetry (ITC) experiments. Cells were lysed using high pressure cell homogenization
45 min, 4°C). Soluble protein was filtered using a 0.22 μm filter and loaded onto a His
51
trap column (GE Healthcare). The column was washed with buffer containing 5 mM
imidazole and protein was eluted with a gradient of 5-500 mM Imidazole. Fractions were
subjected to SDS-PAGE to confirm the presence of protein. Fractions containing protein
were pooled and dialzyed overnight at 4°C against buffer (20 mM Tris pH 7.5, 50 mM
NaCl). Samples were concentrated to ~10 ml, filtered and subjected to SEC using a
Superdex 200 (26/60) column (GE Healthcare) equilibrated with 10 mM Tris pH 7.5, 100
mM NaCl and 0.5 mM TCEP. Fractions were subjected to SDS-PAGE to confirm the
presence of protein. Fractions containing protein were pooled, concentrated, frozen
using liquid nitrogen and stored at -80°C.
For PP1 purification, cells were lysed by high presure cell homogenization and
centrifuged at 97,210 xg for 45 min at 4°C to remove celullar debris. Protein was filtered
through a 0.22 μm filter and added to Ni-NTA resin. Beads were washed with buffer
containing 20 mM Tris pH 8.0, 700 mM NaCl, 1 mM MnCl2 and 15-25 mM imidazole,
depending on the number of times the resin had been used. Protein was eluted with 15
ml of 20 mM Tris pH 8.0, 700 mM NaCl, 1 mM MnCl2 and 250 mM imidazole. Eluted
protein was collected in 35 ml of 20 mM Tris pH 8.0, 700 mM NaCl, 1 mM MnCl2 and 5
mM imidazole.
The PP1:spinophilin complex purification protocol used in Chapter 2 was
modified for use in the purification of the PP1:GM complex. Cells containing PP1 were
lysed, subjected to high presure cell homogenization and centrifuged (97,210 xg, 45 min,
4°C). All steps for this purification were carried out at 4°C. Protein was filtered through a
0.22 μm filter and added to Ni-NTA resin. The column was washed in buffer containing
25 mM Tris pH 8.0, 700 mM NaCl, 1 mM MnCl2 and 10-20 mM imidazole depending on
the number of times the resin had been used. The column was washed in buffer
containing 25 mM Tris pH 8.0, 50 mM NaCl, 1 mM MnCl2, 5 mM imidazole. Purified GM
52
63-237 was added to the column and incubated for 1 hour. The column was washed with
buffer containing 25 mM Tris pH 8.0, 50 mM NaCl, 1 mM MnCl2, 5 mM imidazole and the
PP1:GM complex was eluted using 25 mM Tris pH 8.0, 700 mM NaCl, 1 mM MnCl2 and
250 mM imidazole. Protein was immediately subjected to purification using SEC with a
Superdex 200 (26/60) column equilibrated with 20 mM Tris pH 7.5, 50 mM NaCl and 0.5
mM TCEP. Fractions containing PP1:GM complex were pooled, TEV protease was
added and incubated overnight. The next day the protein was loaded onto Ni-NTA resin
and the flowthrough and 5 mM imidazole washes were collected. The protein was
concentrated to ~10 ml and purified using SEC with a Superdex 75 (26/60) column
equilibrated with 20 mM Tris pH 7.5, 50 mM NaCl, 0.5 mM TCEP. Fractions containing
protein, as verified by SDS-PAGE, were pooled and concentrated to 2.2, 2.8, 3.6, 3.8,
4.8, or 6.4 mg/ml for crystal trials.
3.2.C. Dephosphorylation reactions.
Phosphorylase a (Sigma) was prepared as described91. PP1α 7–330 was incubated
alone or in the presence of GST-GM 2-240 at room temperature (~22°C) for 15 min.
Dephosphorylation reactions were initiated by the addition of either GST-GluR1 809–889
or phosphorylase a. Reactions were incubated for 30 min at 30°C and quenched by the
addition of SDS-PAGE loading buffer without loading dye and boiling for 10 min at
100°C. Samples were separated by SDS-PAGE, phosphorylation was measured using
Pro-Q Diamond phosphoprotein gel stain (Invitrogen) and visualized using a Typhoon
9410 scanner (Amersham Bioscience) with an excitation wavelength of 532 nm and
emission wavelength of 560 nm. Total protein was measured using Sypro Ruby protein
gel stain (Invitrogen). Band densities were quantified using ImageQuant TL software (GE
Healthcare). The quantified band densities from ProQ were corrected for total protein as
previously described92.
53
pNPP dephosphorylation reactions were initiated by the addition of pNPP to a
final concentration of 2 mM. Reactions were incubated for 30 min at 30°C and quenched
by addition of 1 M NaOH. The amount of p-nitrophenolate product was determined using
the absorbance at 405 nm with a molar extinction coefficient of 18,000 M−1 cm−1. All data
were normalized to control reactions without PP1.
3.2.D. Isothermal titration calorimetry.
For ITC experiments, PP1 was purified the day before measurment. PP1 and GM were
dialyzed overnight in identical buffer containing 20 mM Tris pH 7.5, 500 mM NaCl and
0.5 mM TCEP at 4°C. ITC was perormed using a VP-ITC (Microcal) at 23°C with 31 μM,
20 μM and 19 μM of MBP-GM 2-93, 63-237 and 2-237, respectively, against 1.5 μM
PP1. Data was analyzed using Origin software and the data were fit to a 1 site binding
model.
3.2.E. Crystal trials.
For crystal trials using sitting drop vapor diffusion, purified protein was mixed with
precipitant conditions using the Phoenix Liquid Handling System (Art Robbins
Instruments) in 96-well Intelli-plates (Art Robbins Instruments). 50 μl of the precipitant
conidition was added to the reservoir. The crystal screens used included the Protein
Complex Suite(Qiagen), JCSG+ Suite (Qiagen), PEG/Ion(Hampton), PEG/Ion 2
(Hampton), Index (Hampton), Wizard I (Emerald), Wizard II (Emerald), Cryo I (Emerald),
Cryo II (Emerald) and MR001. MR001 is a homemade screen combining solutions from
the PEG 6000 grid screen (Hampton), PEG/LiCl grid screen (Hampton), PEG/Ion,
Protein Complex Suite, Index, Wizard I and Wizard II screens that contained conditions
that were similar to the crystallization condiditons for PP1:spinophilin (Table 3). For
crystal trials using microbatch, 0.5 μl protein and 0.5 μl precipitant conditions were mixed
54
by hand in 72 well Teraski plates (Grenier bio-one) and parafin oil (Hampton) was added
directly over the drop.
Table 3. Crystallization conditions of the MR001 screen.
The 96 conditions of the MR001 screen shown in the position they appear on the plate.
The screens are abbreviated as PEG 6000 grid screen (PEG6K), PEG/LiCl grid screen
(PEG/LiCl), Procomplex (PC), PEG/ion (PI), Wizard I (WizI), and Wizard II (WizII) and the
condition number or position (for grid screens) is listed directly after the screen.
3.3. Results.
3.3.A. Sequence analysis of the PP1 binding and CBM domains of GM.
To investigate the regulation of PP1 by GM, we received a vector encoding GM residues
2-240 from Dr. David Brautigan56. The vector used to express GM contained a N-terminal
GST tag and thrombin cleavage site. The first 240 residues of GM include all regions that
are necessary for altering the substrate specificity of PP1 including the RVxF motif, CBM
domain, and glycogen synthase interaction regions35,56. Additionally, this construct is
similar to the spinophilin417-583 construct that was used for structure determination of the
PP1:spinophilin complex. Both the spinophilin and GM constructs contain the PP1
binding as well as the targeting domains: the PDZ domain for spinophilin and the CBM
55
domain for GM. Using two domains of spinophilin proved beneficial as the PDZ domain
aided in the stability and solubility of the PP1 binding domain which is entirely
unstructured in solution. Therefore, GM 2-240 is an excellent “starting construct” for the
proposed studies of the PP1:GM complex.
Many PP1 regulators are predicted to interact with PP1 via an intrinsically
unstructured domain21. An intrinsically unstructured domain is defined as≥ 50 residues
that are unstructured75. To investigate if GM interacts with PP1 through an unstructured
domain a prediction of intrinsic unstructured regions (IUPRED) was performed (Figure
22)100. A disorder tendency score of 0.5 or higher predicts that the region is unstructured,
while a score of 0.5 or less predicts that the region is structured. The first 56 residues of
GM have an average disorder tendency score of 0.56 predicting that these residues are
disordered while the remaining residues (57-240) are predicted to be structured. This is
interesting, as it was previously shown that residues 1-51 do not play a signifigant role in
Figure 22. IUPRED analysis of GM 2-240.
IUPRED analysis for GM 2-240: a score of > 0.5 indicates regions of disordered while a score of
< 0.5 indicates structured regions. Data corresponding to the RVxF motif is shown in green and
the CBM21 domain is shown in red.
the binding of GM to PP1 using pull down assays56. IUPRED predicts that the RVxF motif
of GM is structured, and therefore GM may interact with PP1 through a structured domain.
56
This is in contrast to spinophilin, whose entire PP1 binding domain is unstructured. As
expected the CBM domain of GM is predicted to be structured.
3.3.B. Purification of GM.
The purification protocol for the PP1:spinophilin complex used in Chapter 2 served as a
starting point for the PP1:GM purification. In the PP1:spinophilin purification scheme,
spinophilin was purified to completion before the formation of the PP1:spinophilin
complex. Therefore, GM 2-240 must be purified to > 98% purity. However, so far all of
the previous biochemical assays performed by the Brautigan laboratory used GST-GM 56.
As a result, we developed a protocol for the purification of GM.
Initially, GST-GM 2-240
was purified using glutathionesepharose
resin
Healthcare).
Protein
eluted
(GE-
using
glutathione
was
reduced
and
dialyzed
overnight against PBS in the
presence
cleave
of
thrombin
GST-GM.
to
Upon
successful cleavage, a second
Figure 23. Attempted purification of GM2-240.
Chromatogram from size exclusion purification of GM and
GST (blue). SDS-PAGE reveals the presence of GM in
both peak 1 and 2. Peak 1 was pooled, DTT was added
and the size exclusion purification was repeated (red).
GST
purification
was
performed where only the flow
through
was
collected.
However, this purification step
removed only some of the GST. Subsequently, SEC was performed using a Superdex
75 (26/60) column. GM eluted from the column at 110 ml, which contained the majority of
57
the GM and at a second peak at 160 ml, which contained a smaller fraction of protein
(Figure 23 blue). The elution volumes for both of these peaks correspond to the elution
volumes for proteins of larger molecular weights than GM 2-240. Peak one eluted at the
void volume of the SEC column. In this peak, GM eluted with a large number of other
unidentified proteins, suggesting that this peak contains mainly aggregated GM. These
aggregates are soluble, since no precipitated protein was observed.
Additionally, elution of GM in this peak is independent of disulfide bond formation
as addition of DTT leads to an identical SEC chromatogram (Figure 23 red). The second
peak corresponds to a MW
of ~44 kDa, suggesting that
GM, which is 27 kDa, elutes
as a dimer. However, GST
co-eluted with GM in this
peak. To further purify GM
from
GST,
fractions
containing protein from peak
2
were
pooled
and
subjected to a third GST
purification. The resin bound
Figure 24. Purified GM 2-237.
SEC purification for GM 2-237 using the Superdex 75
(16/60) column.
both GM and GST equally,
preventing the separation of
GM and GST. The results from this experiment suggest that a large portion of GM is
aggregated in solution.
The sequence of GM was examined in an attempt to design a stable construct
that could be readily purified. The last three amino acids of the GM 2-240 construct are
58
Pro Glu Pro. We speculated that this largely hydrophobic sequence enhances
aggregation. Therefore, in an attempt to facilitate purification, GM was recloned as 2-237.
This sequence was subcloned into petM30-GST, petM30-MBP, and RP1B and tested for
expression. His6 tagged (RP1B) GM 2-237 was successfully purified using Ni-NTA resin
and SEC. Additionally, GM 2-237 elutes from SEC using a Superdex 75 (16/60) column
in a single peak (not in the void volume) demonstrating that GM 2-does not aggregate.
3.3.C. Substrate specificity of PP1:GM.
Spinophilin alters the substrate specificity of
PP1
by
inhibiting
phosphorylase a.
its
activity
against
GM is responsible for
targeting PP1 to phosphorylase a in vivo101.
Since spinophilin inhibits the activity of PP1
against specific GM substrates, we tested
whether GM binding results in the inhibition
of
PP1
against
specific
spinophilin
substrates. We have measured the activity
Figure 25. Substrate specificity of
PP1:GM.
The activity of PP1 and PP1:GM was
measured for three substrates pNPP
(green), GluR1 (blue) and phosphorylase
a (red).
of PP1 for three substrates, the model
substrate (pNPP), a GM specific substrate
(phosphorylase a) and a spinophilin specific
substrate (GluR1). GM does not inhibit the activity of PP1 for these three substrates
(Figure 25). Therefore, while spinophilin showed an inhibition of substrates that are not
specific for the PP1:spinophilin holoenzyme, GM must use a different mechanism to alter
the substrate specificity of PP1.
3.3.D. Initial structural characterization of GM 2-237.
59
NMR spectroscopy can be used to investigate whether a protein is folded. 1D 1H
experiments contain one peak for each hydrogen atom in the protein. The position of
each peak in the spectra is dependent on the chemical environment of the hydrogen
atom that gives rise to it. In a folded protein amide hydrogens (HN) are often part of
hydrogen bonds. These interactions result in HN peaks with chemical shift resonances
between 6 and 10 ppm (Figure 26 A). However, unfolded proteins contain high flexibility
and HN no longer form
hydrogen
bond
interactions. As a result,
the
HN
chemical
shift
range for unfolded protein
is much narrower (Figure
26 B)102. Therefore, the
amount of HN chemical
shift dispersion can be
used to test whether a
protein
unfolded.
is
folded
or
Additionally,
methyl groups that are in
1
Figure 26. 1D H NMR spectra.
1
1D H NMR spectra for a A) folded protein (calmodulin) B)
unfolded protein (RCS) and C) GM 2-237. Arrows highlight the
N
chemical shift dispersion in the backbone H region of the
spectra. The area of the spectra where ring current shifts
occur is highlighted by a box.
close proximity to aromatic
residues, as is the case in
the hydrophobic core of a
folded protein, have hydrogen chemical shifts that are upfield shifted. The presence of
these upfield shifted peaks, called ring current shifts, is another indication that the
sample is folded (Figure 26)103. A 1D 1H NMR spectrum was recorded for GM 2-237
(Figure 26 C). The amide region of this spectrum contains peaks that are well dispersed
60
demonstrating that GM contains a folded domain. However, GM 2-237 may still contain
some regions that are unfolded.
One hallmark of unfolded regions in proteins is that they are highly susceptible to
proteolytic cleavage. In fact, some intrinsically unstructured proteins or domains that
have been purified in the laboratory are only stable for a few hours to days at room
temperature. After the 1D 1H spectrum shown in Figure 26 was recorded, GM 2-237 was
left at room temperature for 24 hours. The next morning the sample was subjected to
SDS-PAGE. Approximately 25% of the total protein was cleaved into a smaller fragment.
This demonstrates that GM 2-237 contains regions that are both structured and
unstructured. This is in good agreement with the IUPRED prediction discussed in
Chapter 5.2.B.1. To investigate which regions of GM are protease labile and therefore
unstructured the sample was left at room temperature for two weeks. After two weeks,
only a single lower molecular weight band was identified by SDS-PAGE. This region
likely corresponds to the well folded region of GM. To determine the molecular weight of
this region, electrospray ionization (ESI) mass spectrometry was performed on the
digested sample. Three fragments of GM were detected with MWs of 15990, 16294 and
16769 Da. These MWs correspond well with GM fragments of residues 92-237, 96-237
and 99-237, 15992, 16290 and 16768 Da, respectively. These results demonstrate that
GM 2-237 includes two different structural regions. GM 2-98 is prone to degradation and
therefore likely unstructured, whereas GM 99-237 is structured and likely corresponds to
the domain boundaries for the CBM domain.
3.3.E. Thermodynamic interaction profile of GM and PP1.
Prior to the work described in Chapter 2, spinophilin residues containing the PP1 binding
activity
had
been
extensively
characterized
by
measuring
the
amount
of
coimmunoprecipitation of PP1 by various spinophilin constructs and mutants68. These
61
results identified the smallest spinophilin construct that contained the total PP1 binding
activity. A construct derived from these results was used in crystallography, as it
contained only residues important for binding and not additional residues that may
remain flexibile upon complex formation and inhibit crystallization. However, the PP1
binding domain of GM has been significantly less characterized. Currently, the only
identified PP1 interaction motif on GM is the RVxF motif (62RRVSF66). Therefore, it is
unclear if GM contains secondary interaction sites for PP1, as have been identified for
spinophilin, I-2, and MYPT1.
Proteolysis experiments demonstrate that GM 2-237 contains two discrete
structural regions. Residues 1-98 are unstructured and contain residues that are
important for its interaction with PP1 and residues 99-237 are structured and contain
regions that are important for the binding of glycogen and PP1 substrates. A crystal
structure of the GM RVxF peptide:PP1 complex has been determined35. However, no
regions outside of the RVxF motif have been shown to play a role in PP1 binding.
Isothermal titration calorimetry (ITC) was used to measure the binding affinity of three
GM constructs for PP1 (Figure 27). The dissociation constants (Kd) for GM 2-237, 63-237
Figure 27. ITC of GM for PP1.
Isothermal titration calorimetry was performed by titrating A) GM 2-237 B) GM 63-237 or C) GM
2-93 into PP1α 7-330.
62
and 2-93 were 50 nM, 65 nM and 518 nM repectively. This demonstrates that residues
C-terminal of residue 93 enhance the binding of GM to PP1 ~10-fold. Additionally, these
data demonstrate that residues 2-63 have no PP1 binding activity. This is in good
agreement with previously published data which demonstrate that residues 2-50 have no
effect on binding using pull down assays56.
3.3.F. Purification of the PP1:GM63-237 complex.
GM 63-237 binds PP1 as tightly as GM 2-237
suggesting that residues 2-62 remain flexible upon
binding to PP1. Flexible protein regions often
inhibit
crystallization
and
thus
structure
determination. Therefore, to minimize flexibility in
the PP1:GM complex, GM 63-237 was used to form
Figure 28. Purification of PP1:GM
A) Final sample from PP1:GM
purified using the protocol for
PP1:spinophilin. B) Final sample of
PP1:GM purified using the new
protocol.
the PP1:GM complex. The initial purification of the
PP1:GM complex followed the protocol for the
PP1:spinophilin complex, as described in Chapter
2.
However,
using
the
identical
purificiation
protocol for PP1:GM produced a sample that was not pure enough for structural studies
(Figure 28 A). Additionally, the yield was 0.5 mg PP1:GM starting from 2.5 liters of PP1
culture, which is only ~20% of the yield for PP1:spinophilin complex. Therefore, a
PP1:GM specific production protocol was developed to enhance production of the
PP1:GM complex. His6 tagged PP1 was bound to Ni-NTA resin, the column was washed
extensively (25 mM Tris pH 8.0, 700 mM NaCl, 1 mM MnCl2, 20 mM imidazole) and 3 ml
of 52 μM GM was added. However, to increase the total yield of PP1:GM, the complex
was eluted in a higher volume (30 ml) and a higher concentration of imidazole (25 mM
Tris pH 8.0, 700 mM NaCl, 1 mM MnCl2 and 250 mM imidazole). The elution volume
63
was too large for a single SEC purification step, which has a maximum injection volume
of 13 ml (Superdex 200 26/60). Furthermore, when the PP1:GM sample was
concentrated for SEC, the complex started to precipitate. Instead, two sequential SEC
purifications were performed using a Superdex 200 (26/60) column. Fractions containing
protein complex were pooled, TEV was added and incubated overnight at 4°C. The
following day subtraction Ni-NTA purification was performed to remove the his6 tagged
TEV protease, contaminants and cleaved purification tag. The PP1:GM complex was
concentrated to 10 ml using Amicon Ultra centrigual filters (Millipore) with a 10 kDa MW
cutoff. Concentrated protein was further purified using SEC on a Superdex 75 (26/60)
column (20 mM Tris pH 7.5, 50 mM NaCl, and 0.5 mM TCEP). This new purification
scheme increased the yield from 0.5 mg to 2.3 mg starting from 2.5 liters of PP1 culture,
as well as increased the purity of the complex (Figure 28 B).
3.3.G. Crystallization trials of the PP1:GM63-237 complex.
The necessary first step for structure determination using X-ray crystallography is to
obtain crystals from purified protein. This is accomplished by mixing the protein solution
with various precipitants. The goal of this process is to cause the protein to precipitate in
an ordered fashion with kinetics that supports protein crystallization. Sixteen different
plates, each containing 96 different conditions, were set up using various screens,
techniques, and temperatures (Table 4). While the PP1:spinophilin complex remained
soluble at a concentration of 4.5 mg/ml in ~50% of the precipitant conditions, the PP1:GM
complex precipitated in almost every condition at 3.8 mg/ml from the PEG/Ion, MR001,
JCSG+, Protein Complex, Wizard, Cryo, and Index screens. The concentration was
decreased to 2.2 mg/ml and additional screens were set up. However, the PP1:GM
complex still precipitated in the majority of the conditions. Based on these results the
PP1:GM complex is less soluble in the final purification buffer (20 mM Tris pH 7.5, 50 mM
64
NaCl, and 0.5 mM TCEP) than the PP1:spinophilin complex. In order to produce crystals
the final purification buffer must be modified to enhance the solutibility of PP1:GM. One
possibility is to increase the salt concentration of the final SEC buffer, which has already
enhanced the solubility of other PP1:regulatory protein complexes studied in the
laboratory.
Concentration
Protein:Solution (μl:μl)
Type
Temperature
MR001
6.4 mg/ml
1:1
SD
4°C
MR001
2.8 mg/ml
2:1
SD
4°C
MR001
4.8 mg/ml
1:2, 1:1, 2:1
SD
4°C
MR001
3.8 mg/ml
1:2, 1:1, 2:1
SD
4°C
MR001
3.8 mg/ml
1:2, 1:1, 2:1
SD
RT
Protein
complex suite
3.8 mg/ml
1:2, 1:1, 2:1
SD
4°C
Protein
complex suite
3.8 mg/ml
1:2, 1:1, 2:1
SD
RT
PEG/ion 1/2
3.8 mg/ml
1:2, 1:1, 2:1
SD
4°C
PEG/ion 1/2
3.8 mg/ml
1:2, 1:1, 2:1
SD
RT
ProComplex
2.2 mg/ml
1:2, 1:1, 2:1
SD
4°C
ProComplex
2.2 mg/ml
1:2, 1:1, 2:1
SD
RT
MR001
2.2 mg/ml
1:2, 1:1, 2:1
SD
4°C
MR001
2.2 mg/ml
1:2, 1:1, 2:1
SD
RT
MR001
3.6 mg/ml
1:1
MB
4°C
MR001
2.0 mg/ml
1:1
MB
4°C
JCSG
3.6 mg/ml
1:2, 1:1, 2:1
SD
4°C
Index
3.6 mg/ml
1:2, 1:1, 2:1
SD
4°C
Wizard I/II
3.6 mg/ml
1:2, 1:1, 2:1
SD
4°C
Cryo 1/2
3.6 mg/ml
1:2, 1:1, 2:1
SD
4°C
Screen
Table 4. Crystallization conditions setup for PP1:GM.
A list of the crystallization conditions that were set up for the PP1:GM complex. Abbreviations:
MB, microbatch; RT, room temperature; SD, sitting drop vapor diffusion.
65
Since GM degraded rapidly when individually purified, the stability of the PP1:GM
complex was tested by placing a 0.8 mg/ml sample at 4°C for three days. The stability of
PP1:GM was evaluated using SDS-PAGE. After three days the PP1:GM complex
degraded into smaller fragments. ESI mass spectrometry was performed on the sample
and three protein fragments were detected with molecular weights of 3563, 16289, and
34884 Da. These molecular weights correspond to GM 63-90, GM 95-237 and PP1 7-307
which have the molecular weights of 3563.9, 16290.4, and 34900 Da, respectively. This
suggests that residues 91-94 stay flexible bound to PP1 and therefore are susceptible to
proteolytic digestion. PP1 contains a flexible C-terminal tail, which includes residues
300-330. PP1 residues 308-330 have been digested suggesting that these residues
remain flexible in the PP1:GM complex and may be inhibiting crystallization.
3.4. Discussion.
We have obtained structural and biochemical data for the regulation of PP1 by GM. First,
we identified that GM 2-237 behaves as a two domain protein. The first domain, residues
1-98, is likely the unstructured PP1 binding domain and the second domain, residues 99237, is the structured CBM domain. Second, using ITC we have shown that GM residues
1-62 do not enhance the interaction of GM with PP1. Instead the complete binding
activity for PP1 is contained in GM residues 63-237. Third, GM residues 91-94 are the
only GM residues susceptible to proteolysis in the PP1:GM 63-237 complex and therefore
likely the only residues that remain flexible upon binding to PP1. Thus, GM residues 6390 are important for PP1 binding and most likely become restricted in conformation upon
binding to PP1, preventing proteolysis. Fourth, we have determined that GM binding to
PP1 does not inhibit the activity of PP1 for any substrates we tested. This leads to the
conclusion that GM does not block any PP1 substrate binding grooves since GM did not
inhibit the dephosphorylation reactions of either GluR1 or phosphorylase a by PP1.
66
To obtain the three-dimensional structure of the PP1:GM complex, we must
modify our purification protocol in multiple ways. First, increasing the salt concentration
of our purification buffers from 50 to 500 mM NaCl may increase the solubility of the
PP1:GM complex. This has already increased the solubility other PP1:regulatory protein
complexes in the laboratory. Second, our proteolysis results indicate that there is
flexibility in the C-terminal 22 residues of PP1. This flexibility may be inhibiting
crystallization. Therefore, we will attempt crystallization using alternate constructs of PP1
to minimize flexible regions. Third, there is no structure available for the CBM domain of
GM and it is possible that this domain is undergoing conformational exchange or that it
contains flexible regions that are further inhibiting crystallization. Therefore, the GM
construct used in the formation of the PP1:GM complex may need to be redesigned to
minimize flexibility. As a result structure determination of the CBM domain by itself will
not only provide valuable insight into the targeting and regulation of PP1 by G M, but will
likely aid the structure determination of the PP1:GM complex.
Chapter 4. The CBM21 domain of GM.
All of the experiments in this section were performed by M.J.R. NMR experiments were
set up with the help of Dr. Anderson Pinhiero and Dr. Wolfgang Peti.
67
68
4.1. Introduction.
CBMs are conserved protein domains that interact with carbohydrates. CBMs consist of
seven families that interact with glycogen: CBM20, 21, 25, 26 34 41 and 45104. The
CBM domain of GM belongs to the CBM21 family, which are characterized by sequence
similarity and localization close to the N-terminus of a protein104. Structures for this family
Figure 29. The CBM21 domains from RoCBM21 and GL.
The structure of RoCBM21 shown as a cartoon representation (A) and as a surface
representation (C) with glycogen binding sites shown in blue. The structure of GL CBM21
shown as a cartoon representation (B) and as a surface representation (D) with glycogen
binding sites shown in blue. E) Sequence alignment for RoCBM21, the GL and GM CBM21
domains. Conserved residues are highlighted in blue and the glycogen binding sites are
outlined in red.
of CBMs have only recently been determined. Currently, only two structures of CBM21
69
domains have been determined including the CBM21 domains from GL (unpublished
PDB ID:2EEF) and Rhizopus oryzae glucoamylase (RoCBM21, Figure 29)105. Two
binding sites for glycogen have been identified on RoCBM21 using the glycogen analog
β-cyclodextrin (Figure 29 C)106. These binding sites were also shown to be co-operative.
The binding sites are also conserved in GM and GL (Figure 29 D,E). Suprisingly, one of
the glycogen binding sites in GM and GL has been previously shown to interact with
glycogen synthase56. Thus there may be either competition for these binding sites in
vivo, or glycogen synthase binding may be co-operative with glycogen binding. Despite
the sequence conservation of the glycogen binding sites, the CBM21 domain of GL
shows only 31% sequence identity with the CBM21 domain of GM. This low sequence
identity suggests that the GM and GL domains may support different protein interactions,
and as a result, have different mechanisms for the regulation of PP1. Therefore, to gain
insight into the regulation of PP1 by GM we have decided to determine the structure of
the GM CBM21 domain.
4.2. Methods.
4.2.A. Expression and Purification of GM102-237.
Rabbit GM 102-237 was subcloned into the RP1B vector. The vector was transformed
into E. coli BL21 (DE3) RIL cells (Stratagene). Freshly transformed cells were grown in
luria broth (LB) shaking at 250 rpm at 37°C until an OD600 of 0.6. The incubator was
cooled to 18°C and protein expression was induced by the addition of 1 mM IPTG. Cells
were grown overnight at 18°C under vigorous shaking (250 rpm). The expression of
uniformly
13
C- and/or
15
N-labeled protein was carried out by growing freshly transformed
cells in M9 minimal media containing 4 g/L [13C]-d-glucose and/or 1 g/L 15NH4Cl (CIL) as
the sole carbon and nitrogen sources, respectively, after induction with 1 mM IPTG.
70
Cells containing GM 102-237 were lysed using high pressure homogenization
45 min, 4°C). Soluble protein was filtered using a 0.22 μm filter and loaded onto a His
trap column (GE Healthcare). The column was washed with buffer containing 20 mM
Tris pH 8.0, 500 mM NaCl and 5 mM imidazole. Protein was eluted using a gradient of 5500 mM imidazole. Fractions containing GM were pooled, TEV was added and the
sample was dialyzed overnight at 4°C into a buffer containing 20 mM Tris pH 7.5, 500
mM NaCl. Cleaved protein was then added to Ni-NTA resin for subtraction purification.
The flowthrough and 5 mM imidazole wash were collected and pooled. The sample was
concentrated to ~10 ml, filtered through a 0.22 μm filter and subjected to SEC using a
Superdex 75 (26/60) column (GE Healthcare) equilibrated with 20 mM Na-PO4 pH 6.5,
50 mM NaCl. Fractions containing protein were verified by SDS-PAGE and pooled. For
NMR spectroscopy,
N GM 102-237 was concentrated to 800 μM and
15
15
N,13C GM 102-
237 was concentrated to 860 μM. 10% D2O was added to all NMR samples. The
concentrated samples were either used directly for NMR measurments or frozen in liquid
nitrogen and stored at -80°C until usage.
4.2.B. Crystallization of GM 102-237.
For sitting drop vapor diffusion trials, precipitant and concentrated, purified protein were
mixed using the Phoenix Liquid Handling System (Art Robbins Instruments) in 96-well
Intelli-plates (Art Robbins Instruments). 50 μl of the precipitant conidition was added to
the reservoir. For crystal trials using microbatch, 0.5 μl precipitant was mixed with 0.5 μl
concentrated, purified protein by hand in 72-well Teraski plates (Grenier bio-one) and
parafin oil (Hampton) was added directly over the drop. For hanging drop vapor diffusion
protein and precipitant conditions were mixed by hand on siliconized glass cover slips
71
(Hampton). The cover slips were inverted and placed onto greased 24-well VDX plates
(Hampton) containing 500 μl precipitant.
4.2.C. NMR spectroscopy of GM 102-237.
NMR measurements were performed at 298 K on a Bruker Avance II 500 MHz
spectrometer equipped with a TCI HCN z-gradient cryoprobe. 2D [1H,15N] HSQC, 3D
HNCACB, 3D CBCA(CO)NH, and 3D HNCA spectra were used to complete the
sequence specific backbone assignment. 3D CC(CO)NH were used to assign the side
chain carbon atoms. 3D HNCO, 3D HN(CA)CO, 3D HBHA(CO)NH, 3D HCCH TOSY, 3D
15
N-resolved [1H,1H] NOESY (mixing time, 70 ms) were also recorded and will be used
for the assignment of sidechain hydrogen atoms and for structure determination. A 3D
13
C-resolved [1H,1H] NOESY (mixing time, 70 ms) spectrum was recorded on an Avance
800 MHz spectrometer at Brandeis.
Chemical shift indices (CSI) were calculated as previously described57. The
RefDB, Wang and Jardetzky, and Wishart random coil databases were used in
calculations107-109. Data from residues preceding prolines were excluded.
4.3. Results.
4.3.A. Purification and initial
characterization of GM102-237.
The purification for GM 102-237 followed standard
protocols in the laboratory and produced a > 98%
pure sample with a yield of 5.8 mg/liter of cell
culture (Figure 30). The structures of both
RoCBM21 and the CBM domain of GL are
Figure 30. The final GM 102-237
sample.
Concentrated and purified GM
102-237 was subjected to SDSPAGE immediately before NMR
spectra were recorded.
predominantly composed of β-strands (Figure 29
A,B). Therefore, after purification circular dichroism
72
(CD) spectra were recorded to determine if the CBM21 domain is folded and
predominantly β-strands as expected. The CD spectrum of GM 102-237 showed a
minimum at 218 nm corresponding well with the expected spectra for proteins containing
mostly β-strands. No minima were observed at either 222 or 208 nm which corresponds
to α-helicical proteins. A melting curve of GM 102-237 was performed by monitoring the
absorbance at 218 nm while increasing the temperature from 23°C to 100°C. Two
melting curves were recorded and averaged providing a melting temperature of 72 ± 6
°C.
Figure 31. Circular dichroism spectrscopy of GM 102-237.
A) CD spectra recorded in triplicate on 10 μM GM 102-237. B) A melting curve of GM 102-237
measuring the absorbance at 218 nm which corresponds to the minimum of A.
4.3.B. Crystallization of GM 102-237.
Structures for CBM21 domains have been determined using both X-ray crystallography
and NMR spectroscopy105,106. As a result both methods were considered for the structure
determination of the GM CBM21 domain. We initially focused on X-ray crystallography.
After purification of GM 102-237, crystal trials were set up and crystals appeared
overnight at room temperature in 30% PEG-8000, 0.1 M Na Acetate pH 4.5, 0.2 M
Li2SO4 from the Wizard I screen, using sitting drop vapor diffusion (Figure 32 A). These
73
crystals
contained
however
weakly
they
to
~8
protein,
only
Å.
diffracted
For
crystal
optimization a fine screen based
Figure 32. Crystallization of GM 102-237.
A) The initial crystal hits that were produced
overnight at room temperature in 30% PEG-8000,
0.1 M Na Acetate pH 4.5, 0.2 M Li2SO4. B)
Optimized crystals using microbatch at room
temperature.
on the initial condition, as well as
different
additive
and
dilution
screens were used. Hanging drop
vapor diffusion and microbatch
techniques were used, along with incubating plates at room temperature and 4°C. Lastly,
the concentration of GM 201-237 was varied from 5 to 10.6 mg/ml. Through optimization,
the size and quantity of the crystals were increased and a substantial improvement in
the diffracting resolution to ~3 Å was observed (Figure 32 B). However, the crystals grew
as multiple latices that could not be separated, meaning that multiple diffraction patterns
were observed from a single crystal. Additionally, all tested cryoprotectants including
glucose, PEG-400, MPD, glycerol, and sucrose caused the crystals to crack or dissolve.
As a result no crystal was obtained that was suitable for structure determination and
instead NMR spectroscopy was pursued to investigate the structure of GM 102-237.
4.3.C. NMR spectroscopy of GM 102-237.
In order to perform structure determination using NMR spectroscopy, it is necessary to
assign each chemical shift resonance to the corresponding atoms. 2D [1H,15N] HSQC
spectra correlate the magnetization of amide hydrogens and nitrogens atoms in the
protein backbone (Figure 33 A). Since each amino acid except for proline contains a
backbone amide and nitrogen this experiment produces a spectrum that contains one
peak for each amino acid in the protein except for prolines (Figure 33 B). To evaluate if
structure determination is possible by NMR spectroscopy a 2D [1H,15N] HSQC spectrum
74
is recorded and the number of peaks in the spectrum are counted. If the number of
peaks is identical or close to the number of residues in the protein it is possible to use
NMR spectroscopy for structure determination.
15
N GM 102-237 was purified in buffer
containing 20 mM NaPO4 pH 6.5, 50 mM NaCl, 10% D20 and a 2D [1H,15N] HSQC
spectrum was recorded. GM 102-237 contains 130 non proline residues and 128 peaks
were counted demonstrating that structure determination by NMR spectroscopy is
readily feasible.
The next step in NMR based
structure determination is to perform
the
sequence
specific
backbone
assignment. This relates each peak in
the 2D [1H,15N] HSQC spectrum to a
single amino acid in the protein. To
perform
the
sequence
specific
backbone assignment double labeled
15
N,13C GM 102-237 was produced
and 3D HNCA, 3D HNCACB, and 3D
CBCA(CO)NH spectra were recorded.
These
1
15
Figure 33. 2D [ H, N] HSQC spectrum.
1
15
A) A 2D [ H, N] HSQC experiment correlates the
magnetization between amide hydrogen and
nitrogen atoms. B) 3 peaks are observed in a
1
15
theoretical 2D [ H, N] HSQC spectrum resulting
for the 3 amino acid peptide shown in A.
spectra
correlate
magnetization from amide hydrogens
and nitrogen with the magnetization of
specific side-chain carbon atoms. The
sequence
specific
backbone
assignment for GM 102-237 has been completed and the fully annotated 2D [1H,15N]
HSQC spectrum is shown in Figure 34. To finish the complete chemical shift assignment
75
Figure 34. The backbone assignment of GM 102-237.
1
15
Fully annotated 2D [ H, N] HSQC spectrum of GM 102-237. Each peak is labeled with the
corresponding single letter amino acid code and residue number.
of GM we will need to assign the carbons and protons of all aliphatic and arromatic side
chains. All necessary spectra have already been recorded. Interactomic distances of
76
less than 5Å will be measured using NOE spectra. Lastly, these distance restraints will
be used to determine the structure of GM 102-237.
In the absence of the 3D structure, structural information on the protein using
chemical shift information is readily available. Deviation of the Hα, Cα, Cβ, and CO
chemical shifts from their random coil values, termed the chemical shift index (CSI),
provides information about the secondary structure of a protein110. CSI values have been
calculated for GM 102-237. The values are largely negative indicating that the majority of
the protein contains β-strands. One helix was observed at the N-terminus of this domain
Figure 35. Chemical shift index for GM 102-237.
ΔCα – ΔCβ CSI values are reported using the RefDB database for the random coil chemical
shifts. Positive values indicate the presence of an α-helix, while negative values indicate the
presence of a β-strand. The secondary structure elements of RoCBM21 and the CBM21
domain of GL are shown above the graph for comparison. β-strands are represented as
arrows and α-helices are represented as cylinders.
which is in aggreement with the position of an α-helix in GL. The CSI scores are in good
77
aggreement with our CD data and the secondary structure content of previous CBM21
domain structures, indiciating that the overall fold will likely be similar.
4.4. Discussion.
We are making progress towards the structure determination of the CBM21 domain of
GM. CBM21 domains contain an overall β-sandwich fold containing eight β-strands
(Figure 29) labelled β1-8 in Figure 35106. Our preliminary data demonstrates that the
central eight β-strands of the CBM21 domain are conserved in GM. Both RoCBM21 and
the GL CBM21 domain structure contain a 310 helix. Interestingly, the position of this helix
is different between RoCBM21, and GL CBM21. Our CSI data also suggests that GM
may contain a 310 helix in a different position than either RoCBM21, or GL CBM21, in
between β4 and β5. Additionally, due to the low sequence identity between the CBM
domains of GM and GL it will be important to determine the structure of GM as the
properties of specific binding pockets may change resulting in different binding
properties for these domains. It will also be interesting to investigate the role of glycogen
and glycogen synthase binding as they have been shown to contain overlapping binding
sites.
Chapter 5. Small Angle X-ray Scattering.
78
79
5.1. Introduction and Background.
In Chapter 2 of this thesis, X-ray crystallography was used to investigate the structure of
the PP1:spinophilin holoenzyme. The structure reveals that the PP1 binding domain of
spinophilin, which is highly flexible when not bound to PP1, becomes locked into a single
conformation, with little residual flexibility upon PP1 binding. This decrease in flexibility
upon binding, allowed for crystallization and structure determination of the complex by Xray crystallography. However, numerous PP1 regulators retain a high degree of flexibility
when bound to PP1, which makes structure determination by X-ray crystallography
difficult for two reasons. First, flexibility can often prevent crystallization, which is the
necessary first step of structure determination by X-ray crystallography. Second, if
crystals are produced, flexible regions often do not give rise to any discernable electron
density and therefore cannot be accurately built into the structure. This has already been
observed for the PP1: inhibitor-2 (I-2) complex, in which no electron density was
observed for ~70% of I-2 residues. This suggests that it might be difficult to use X-ray
crystallography to obtain structural information for many PP1:regulatory protein
complexes. Instead, small angle X-ray scattering (SAXS), which is performed on
samples in solution rather than in crystals, can be used to gain low resolution structural
information for proteins with high flexibility. SAXS theory and methodology have been
discussed in detail in three recent review articles which were used as references for this
chapter: Small-angle scattering: a view on the properties structures and structural
changes of biological macromolecules in solution by M.H.J. Koch et al.111, X-ray solution
scattering (SAXS) combined with crystallography and computation: defining accurate
macromolecular structures, conformations and assemblies in solution by C.D. Putnam et
al.112, and Small-angle scattering studies of biological macromolecules in solution by D.
Svergun and M.H.J. Koch113.
80
SAXS, NMR spectroscopy and X-ray crystallography are complimentary
techniques that are used to obtain different structural information for biological
macromolecules. For both SAXS and X-ray crystallography, monochromatic X-rays are
focused onto a sample cell. When the X-rays interact with the sample positioned in the
sample cell, they are scattered and a detector measures the intensity of the scattered Xrays. The intensity and angle of the scattered X-rays contains structural information
about the sample. In X-ray crystallography this information can be used to obtain atomic
level detail of the structure of a biological macromolecule, while in SAXS it can be used
to obtain low resolution (≥10Å) information regarding the shape and size of a
macromolecule in solution. This difference in information results from the organization of
molecules within the sample and, as a result, the type of X-ray scattering that occurs. A
SAXS sample consists of a purified biological macromolecule tumbling in solution with
macromolecules oriented and distributed randomly relative to one another. In SAXS, Xrays will interact with the molecule in all orientations and will scatter in a radially
symmetric and continuous pattern around the beam center. The majority of the X-rays
are scattered at small angles of less than 4°. Since the resolution of the experiment is
directly related to widest angle of scattering measured, the resolution of SAXS
experiments is often low (~10-20 Å). In X-ray crystallography, crystals are produced
which contain a repeating array of a biological macromolecules in identical orientations.
X-rays will interact with aligned macromolecules and give rise to a specific type of
scattering called diffraction. Scattered X-rays interfere with one another producing a
discrete diffraction pattern with individual maxima called reflections. The position and
intensity of the reflections contain information about the orientation and position of
electrons within the crystal. By sampling the crystal in many different orientations, the
diffraction pattern can be used to obtain an electron density map of the sample. In
81
crystallography, the majority of X-rays are scattered over a larger angular range (4-170°)
than in SAXS, thus producing high resolution data (≥1 Å).
Despite the much lower resolution and the lack of orientational information,
SAXS can still be used to obtain low resolution information about a macromolecule in
solution, such as molecular weight, folding, overall shape and size. Additionally, SAXS
has several advantages over other structural biology techniques including X-ray
crystallography, NMR spectroscopy, and cryo-electron microscopy (cryo-EM). First,
SAXS experiments are generally recorded on dilute samples in solution and therefore do
not require extensive sample preparation, as is the case in X-ray crystallography and
cryo-EM. Second, SAXS experiments can be performed on large molecular weight
protein complexes, such as the 30S ribosome (Bioisis ID:12), which are challenging to
study using NMR spectroscopy, a technique that becomes increasingly more difficult
with larger molecular weight proteins. Third, SAXS can be used to investigate molecules
that are too small for detection by cryo-electron microscopy, including the 8.5 kDa
PF1061 protein (Bioisis ID:31). These advantages clearly show that SAXS
measurements can be used to obtain structural information on a variety of samples that
are not applicable to other structure determination techniques.
5.2. Beamline design.
Careful design of a SAXS beamline is essential to maximize the information that can be
obtained from a SAXS experiment. For SAXS, monochromatic X-rays are typically
focused with a series of crystals and slits onto a sample. The SAXS detector is
positioned to measure the intensity and angle of the scattered X-rays. In contrast to Xray crystallography where the angle of scattering measured is typically large, 4-170°, the
angle of scattering that is of interest for SAXS is 0-4°. To gain the maximal amount of
information from such a small angular range, SAXS beamlines must be able to measure
82
Figure 36. SAXS beamline design.
A diagram of a SAXS beamline that is used for recording small angle (SAXS) and wide angle
(WAXS) data simultaneously.
scattering at the smallest angle possible. As a result, it is important to decrease the
amount of parasitic scatter, which is the scatter that results from an X-ray beam passing
through a focusing slit. Small amounts of parasitic scatter are not normally a concern in
X-ray crystallography because they occur at a too small angle to detected. However, in
SAXS experiments, parasitic scattering at small angles can mask intensities which are
important for data analysis. Decreasing the parasitic scatter at many beamlines,
including NSLS X9, has been accomplished using a three slit system. In this system the
first two slits collimate the beam, while the third slit is wider than the first two and blocks
the parasitic scatter at small angles114. Another consequence of measuring small angle
scatter is that the detector must be placed much farther away (1.5 m or more) from the
sample than in X-ray diffraction (0.2 m) to gain resolution at low scattering angles. Since
the distance between the sample and the detector is much larger than for
crystallography and the intensity is lower, it is essential to decrease the amount of
intensity loss due to scattering from air. As a result, the region between the sample and
the detector in SAXS beamlines is regularly vacuum enclosed. In some instances, it is
also important to record scattering at wider angles, which is referred to as wide angle Xray scattering (WAXS). Some beamlines, including X9 at NSLS and X33 at DESY, have
been designed to measure both SAXS and WAXS data simultaneously. These setups
contain a SAXS detector that is centered behind the beam with the WAXS detector
offset as not to block the majority of the small angle scattering but still detect wide angle
83
scattering (Figure 36). This allows the wide angle scattering to be measured without
blocking the small angle scattering. Since the scattering pattern is radially symmetric
around the beam center it is not necessary to measure the complete circle of intensity
around the beam center (Figure 37).
5.3. Acquisition and analysis.
5.3.A. Data acquisition.
SAXS
experiments
are
sensitive to the quality of
the
sample.
sample
The
is
ideal
pure,
homogenous
and
monodisperse.
This
is
often accomplished using
standard
purification
methods for proteins or
nucleic acids. After the
final
which
purification
is
exclusion
often
step,
size
Figure 37. Scattering data.
Scattering images of buffer (A) and sample (B). C) These
images are integrated, scaled and averaged to give scattering
intensity plots for buffer (blue) and sample (red). D) The buffer
image is subtracted from the sample to give the final
scattering pattern.
chromatography, fractions containing the sample are pooled but not concentrated.
Instead the sample is left dilute in solution at 4oC until just prior to data collection to
minimize any concentration dependent aggregation. Aggregation is a major concern for
SAXS, since larger samples, like aggregates, scatter more at smaller angles than
smaller samples. As a result, small amounts of aggregation can mask the signal from the
sample of interest. Immediately before data collection the protein is concentrated and
filtered using a 0.02 μm Anotop filter (Whatman). This is the method we use when we
84
collect SAXS data at NSLS X9. At other beamlines, including 12.3.1 at ALS, size
exclusion chromatography is performed immediately before acquisition to maximize
sample homogeneity. In this case, users purify 25 μl of sample using a small gel filtration
column immediately before each data collection.
In general, the best data are recorded on samples that are purified as close to
the time of data collection as possible. Once the sample is produced it is placed in one
of two types of sample chambers; a static cell or flow cell. The sample is exposed to Xrays and scattering is measured. Since SAXS experiments are performed at room
temperature with much lower concentrations than in a protein crystal, the samples are
exceedingly more susceptible to radiation damage than crystals, which are frozen and
exposed to X-rays at 100 K. To minimize the effects from radiation damage flow cell
apparatus have been designed to continually push fresh sample through the X-ray
beam. This prevents the sample from being exposed to the beam for too long. At NSLS
X9 we use a flow rate of 0.1 μl/s. In the absence of a flow cell, it is important to monitor
for radiation damage. This is accomplished by measuring scattering using short
exposures before and after each experiment to determine if the amount of scattering is
decreasing. After each experiment scattering is recorded on buffer alone. This is
important because the average electron density of water (0.33 e-/Å3) is similar to the
average electron density for proteins (0.44 e-/Å3). Therefore, the intensity of scatter from
the protein sample is small compared to the total scattering, thus accurate buffer
subtraction is essential.
During data acquisition for each experiment, a 2D image is recorded (Figure 37
A,B). Scattering in SAXS is radially symmetric around the beam center. Thus, each
intensity value is averaged around the image at the radius corresponding to the
momentum transfer (q), where q=(4π sinθ)/λ. This is accomplished using aquistion
85
software. At NSLS X9, the software used is view.gtk written by Dr. Lin Yang. It is
important to note that the momentum transfer is expressed as either q or s where q = Å-1
and s = nm-1 and that many of the programs used in data analysis can use either as
input data. Here I will solely use q(Å-1) as this is what is used at X9. After intensity values
are averaged, a buffer alone dataset is subtracted from the sample data set (Figure 37
C). The program PRIMUS, which is freely available as part of the ATSAS suite of
programs
(http://www.embl-hamburg.de/ExternalInfo/Research/Sax/software.html),
is
capable of performing all data manipulations. PRIMUS is designed for data
manipulations and initial analyses of scattering data. The sample data and buffer data
are opened in PRIMUS by clicking on the tools tab and then the select button in the data
processing windows (Figure 38). The intensity values of each dataset
must be
normalized to the total beam intensity. At NSLS X9 this parameter is called usmon and
Figure 38. Snapshot of PRIMUS.
Main window (left) and the data processing windows of PRIMUS are shown. Clicking the tools
button highlighted in red brings up the data processing window. Clicking the select button
highlighted in red allows the user to load there data.
can be found in the experimental log file. Normalization is accomplished via the multiplier
column on the right side of the data processing window. The buffer alone dataset is
subtracted from the sample dataset using the subtract button in the data processing
86
window (Figure 38). The final scattering data is plotted as log intensity (relative units)
against q (Å-1) (Figure 37 D).
Scattering intensity is related to the electron distribution of the sample (equation
3.1) where Dmax is the maximal distance in the sample. The electron distribution (p vs. r)
𝐷
𝐼(𝑞) = 4𝜋 ∫0 𝑚𝑎𝑥 𝑝(𝑟)
sin(𝑞𝑟)
d𝑟
𝑞𝑟
(3.1)
is the real space representation of the molecule, while the scattering data (I vs. q) is the
reciprocal space representation. They can be readily interconverted by Fourier
transformation. If the sample has the shape of a box and the length of a side is equal to
d, then the majority of the
scattering will occur with a
momentum transfer of less
than 2π/d (Figure 39). This
demonstrates
that
the
scattering intensity from
Figure 39. Real and reciprocal space.
A) Two rectangles of length d1 and d2 will produce the
corresponding scattering patterns shown in B. Adapted from
Svergun and Koch 2003.
larger samples decreases
at smaller values of q than
for
smaller
samples.
Importantly, basic information about the sample can be obtained directly from the
scattering data including 1) the foldedness, 2) the radius of gyration (Rg), the root mean
square distance of all atoms from the center of the molecule, and 3) whether the sample
is aggregated. How these data are extracted from the scattering curve is described in
the following sections.
5.3.B. The Guinier analysis.
87
At small values of q the scattering intensity is solely defined by the Rg. The relationship
between intensity and Rg is called the Guinier approximation (equation 3.2) and is only
correct for q values smaller than 1.3/Rg115. Rearrangement of this equation (equation
3.3) shows that when ln(I) is plotted against q2, the radius of gyration can be determined
from the slope and the forward scattering intensity I(0) from y intercept. The resulting
𝐼(𝑞) = 𝐼(0)𝑒
−(𝑞2 𝑅𝑔2 )
�
3
�
ln 𝐼(𝑞) = ln 𝐼(0) − �
𝑅𝑔2
� 𝑞2
3
(3.2)
(3.3)
plot is called the Guinier plot and can also be used to readily detect the presence of
aggregation in the sample. The linearity of the Guinier plot is dependent solely on the Rg.
If the sample is aggregated, it will have multiple Rg’s and therefore the plot will deviate
from linearity (Figure 40A). The deviation from linearity is often visualized in PRIMUS
using a residuals plot, which plots the differences between the data points and the
Guinier trendline (Figure 40B). Both the Guinier plot and residuals plot can be visualized
in PRIMUS by using the Guinier button in the data processing window. After clicking the
Guinier button the entire scattering curve is plotted as ln(I) against q2 and the number of
points will need to be adjusted to display only the Guinier range. The range displayed
can be adjusted by changing nBeg, the first data point used, and nEnd, the last point
used. The y intercept of the Guinier plot is equal to the natural log of the I(0), which is
the theoretical scattering at a q value of zero. It is impossible to measure I(0) directly, as
it would overlay with the incident beam. I(0) is related to the number of electrons in the
scatterer and is independent of shape or size. Thus, I(0) can be used to determine the
molecular weight of the sample by direct comparison with the I(0) for standards with
known molecular weights. The forward scattering intensity can also be used to
investigate the concentration dependence on the multimerization state of the sample.
88
Scattering data measured at different concentrations must produce a linear plot of I(0)
against concentration assuming that no concentration dependent multimerization is
occurring. If a measurment is outside the linear range, than the data is either
Figure 40. Guinier analysis.
A) A Guinier plot of an aggregated (blue) and a non aggregated sample (red). B) Residuals
plot of an aggregated (blue) and a non aggregated sample (red) generated in primus. C) Plot
of I(0) against concentration. A trendline is drawn for points less than 1 mg/ml.
multimerized or aggregated and should not be used for analysis. In Figure 40, data for
the PP1:spinophilin complex recorded at 0.5 mg/ml, 0.9 mg/ml and 1.7 mg/ml is shown.
The data at 0.5 mg/ml and 0.9 mg/ml adhere to a linear trendline but the data for 1.7
mg/ml falls outside of this linear range and should therefore be discarded.
5.3.C. The Kratky plot.
An additional plot that can be directly created from the scattering data is the Kratky
plot116. In the Kratky plot, q2I is depicted as a function of q. For a folded protein this plot
89
is parabolic at low q values, however at higher q values the q2I returns to approximately
zero. This results from Porod’s law, which states that the scattering intensity decays at a
rate proportional to q-4 for a globular protein112. However, in the case of partially folded or
completely unfolded proteins the intensity decay is often closer to q-5/3, which causes the
plot to no longer tend towards zero. This gives characteristic plots for unfolded (I-2) and
folded (PP1:I-2) protein shown in Figure 41. A Kratky plot, which looks like a
combination of the folded and unfolded plots, is indicative of partially folded proteins or
multidomain proteins with flexible linkers. One example of a protein that has an
intermediate Kratky plot is
spinophilin417-583,
contains
one
which
unfolded
(residues 417-494) and one
folded PDZ domain (residues
495-583, Figure 41). Like the
Guinier
approximation,
Porod’s law only holds true for
Figure 41. Kratky plot.
Kratky plot for an unfolded protein (blue) and a folded
protein (red) is shown. A Kratky plot for spinophilin 417583 shows the characteristic plot for a partially folded
protein (green).
a particular region of the
scattering curve. At higher
angles,
scattering
is
dependent on the internal structure of the protein in addition to its overall shape. The
effects of the internal structure on scattering cause the intensity to no longer decrease at
q-4 at higher angles. The cutoff for this region is not as clearly defined as the region for
the Guinier approximation.
5.3.D. The pair distance distribution function.
90
As discussed in section
5.3.A,
the
electron
distribution of a sample
can be determined via a
Fourier transformation of
the
Figure 42. The pair distance distribution function.
A) The p(r) plot for five objects with the same maximum
length but unique shapes. B) The shapes corresponding to
the p(r) curves in A. Adapted from Svergun and Koch 2003.
scattering
experimental
curve.
Since
SAXS is performed on
biological macromolecules
in solution, and biological macromolecules rotate on a nanosecond timescale in solution,
the scattering curve is the average scattering of the sample in all orientations. As a
result, the electron distribution obtained during SAXS represents the probability of two
atoms being separated by specific distances. A plot of the electron distribution is the
pair distance distribution function (PDDF or p(r)). The p(r) can provide useful information
about the overall shape of the sample. Typical p(r) functions for five objects with the
same Dmax but different shapes and Rg values, are shown in Figure 42. A spherical
object gives rise to a symmetric bell shaped p(r) with a maxima at around Dmax/2, while
more elongated shapes (yellow & green) gives rise to p(r) functions whose maxima is no
longer located in the center of the function. A two domain protein separated by a linker
(purple) gives rise to two distinct peaks. Lastly, a hollow sample (blue) has a maxima at
greater than Dmax/2. Determining the p(r) from I(q) follows equation 3.4 and requires
measurement of the scattering intensity over the entire q range. This is technically
impossible
as
the
scattering
𝑝(𝑟) =
intensity
∞
𝑟
∫ 𝐼(𝑞)𝑞 sin(𝑞𝑟)d𝑞
2𝜋2 0
decreases
rapidly
at
(3.4)
91
higher scattering angles and therefore, the intensity is recorded only over a small q
range, typically 0.01-0.3. Instead, an indirect Fourier transformation is used by the
program GNOM to calculate p(r)117,118. GNOM calculates p(r) over a defined real space
range and uses the regularization parameter alpha to minimize the discrepancy between
the theoretical scattering data for the p(r) and the experimental scattering data, while
producing a realistic p(r) function119. Alpha is a combined fit of 6 parameters: 1)
discrepancy (discp), 2) systematic deviations (sysdev), 3) stability (stabil), 4) oscillations
(oscill), 5) positivity (positv), and 6) validity in the central part of p(r) (valcen). The
optimal value for each parameter is listed in Table 5. Discrepancy, systematic deviations
and stability define the quality of the fit of the theoretical scattering data for p(r) to the
Calculation
Dmax
Discrp
Oscill
Stabil
Sysdev
Positv
Valcen
Total
Optimal
−
0.700
1.100
0.000
1.000
1.00
0.95
≥0.5
1
50
10.357
1.140
0.075
0.068
1.00
0.866
0.552
2
100
0.708
1.403
0.012
0.355
1.00
0.881
0.668
3
180
0.695
2.521
0.014
0.387
1.00
0.490
0.414
Table 5. GNOM parameters.
The optimal values for parameters from GNOM and the values for calculations 1,2 and 3 from
Figure 21 are listed.
experimental data. Discrepancy, is the χ2 of the fit to the experimental data. Systematic
deviations refer to non-random fluctuation of the residuals between the experimental
data and the fit. Stability refers to the change in the fit to the experimental data over
different alpha values. Oscillations, positivity, and validity in the central part of p(r) define
the realism of the p(r) based on the p(r) for a solid sphere. Oscillations assumes that the
p(r) function is a smooth function. Positivity assumes that the function should not be
negative over the range 0 to Dmax. Validity in the central part of p(r) assumes that the
majority of the p(r) distribution should occur in the middle of the range 0 to Dmax. Three
unrealistic p(r) functions are shown in Figure 43, with the values for oscillations,
positivity, and validity in the central part of p(r) listed. Since not enough independent
92
experimentally determined parameters are measured, an alpha value is selected that
has the highest total value. The total value is the overall measure of how close all of the
parameters are to their optimal values, where a total value of≥0.5
represents an
acceptable solution and a value of 1 means that each parameter is optimal.
Figure 43. Different P(r) functions and their respective parameters.
A) Ideal P(r) for a globular protein. B) P(r) resulting from over fitting the experimental data. C)
P(r) resulting from incorrect error estimates for the experimental data. D) Typical P(r) for a
lipid bilayer. This figure is adapted from Svergun 1992.
Since the indirect Fourier transform is performed over a defined real space range
(0 - Dmax), it is necessary to repeat GNOM over a range of Dmax values to obtain the best
p(r). To select the best p(r), it is important to consider all of the discussed parameters.
GNOM selects a p(r) function based on six parameters and therefore it is possible that a
solution contains an acceptable total value but has a poor fit to the experimental data
93
and is therefore a poor solution (Table 5, calculation 1). In general the best solution is
one that has an alpha value greater than 0.5, with a good fit to the data and a
reasonable p(r) solution. As a practical example, GNOM was used to calculate p(r)
solutions for SAXS data from the PP1:spinophilin complex at a specified Dmax of 50, 100,
Figure 44. Selecting the best output from GNOM.
A series of calculations from GNOM using Dmax of 50(A), 100 (B) and 180 Å (C) to
demonstrate underfitting, correctly fitting and overfitting your data. The calculations are
referred to as 1(A),2 (B) and 3(C) in Table 5.The plots on the left are the scattering data as
circles and the fit to the data as a solid line. The right plots are the P(r) functions.
and 180 Å (Figure 44). In general, it is good practice to use Dmax ~ 3(Rg) as a starting
point for GNOM analysis and to subsequently repeat GNOM with steps of ±10 Å for Dmax
around the starting Dmax. Here only three GNOM solutions are discussed in detail. In
order to identify the best solution it is important to analyze three parts of the output data.
94
First, the fit of theoretical scattering data from the p(r) function to the experimental
scattering data (Figure 44, left); second, the p(r) curve (Figure 44, right), and third, the
six parameters output from GNOM (Table 5). Clearly, the fit to the experimental
scattering data in Figure 44 A is poor. This is confirmed by the high discrepancy value of
10.357 (Table 5, calculation 1). In addition, the p(r) function is unrealistic for a globular
protein, as the value drops suddenly to zero at Dmax. The fit for calculations 2 and 3 to
the experimental data is good both visually (Figure 44 B,C) and statistically, with
discrepancy values of 0.708 and 0.695, respectively (Table 5). Since these two criteria
suggest that calculation 2 and 3 are equally good solutions, the p(r) functions must be
examined. For calculation 2, the p(r) remains positive throughout the entire distance
range and contains a clear peak that tapers towards zero at Dmax. The function suggests
that the protein is elongated and asymmetric with a central spherical region. Previous
knowledge of the expected shape of the protein is useful for the selection of the correct
solution. For example if the protein is spherical, a p(r) that is similar to the red p(r) from
Figure 42 is expected. However, as we have determined the crystal structure of the
PP1:spinophilin holoenzyme we know that the structure is elongated and asymmetric. In
contrast, the p(r) for calculation 3 oscillates betweeen negative and positive values at
larger distances. This suggests that the p(r) function is wrong and that the actual Dmax
value must be smaller. Furthermore, both the values for oscillitaions and validity have
deviated from their optimal values of 1.100 and 0.95 and are 2.521 and 0.490,
respectively. Thus, calculation 2 is the best choice and Dmax is ~100 Å. Subsequently it is
important to perform additional GNOM calculation with Dmax values of 90, 95, 105, and
110 Å to chose the most optimal solution based on the discrepancy and p(r).
5.3.E. Ab initio envelope determination.
95
Multiple programs can generate
an
ab
initio
low
resolution
envelope via a direct fit to the
experimental
The
most
programs
scattering
data.
commonly
used
are
DAMMIN
and
GASBOR120,121. There are three
major differences between these
two programs. First, DAMMIN
creates a model using densely
packed dummy atoms, while
Figure 45. Structural information contained in
scattering curves.
Twenty five scattering curves of proteins with different
molecular weight, Dmax, shape and volume are
overlayed. The resolution corresponding to the s value
is listed above the graph. Adapted from Svergun et al.
2001.
GASBOR
uses
a
chainlike
model of
connected dummy
residues, where the number of
dummy residues is equal to the
number
of
residues
in
the
protein. Second, GASBOR can create an ab initio envelope from either the p(r) or the
experimental scattering curve while DAMMIN can only use experimental scattering data
as input. Fitting the p(r) drastically decreases the time it takes calculate the ab initio
envelope. Third, DAMMIN can only support scattering data up to a q value of 0.3, which
corresponds to a resolution of 21 Å. GASBOR can fit scattering data to a q value of 1.2,
which corresponds to a resolution of 4 Å. Therefore, GASBOR can be used to obtain
higher resolution ab initio envelopes than DAMMIN. However, the resolution limit for
obtaining structural information from experimental scattering data is ~5 Å. This is
because scattering curves converge at high resolution, demonstrating that it is nearly
impossible to extract detailed structural information at the level of secondary structure
96
(Figure 45). Since we use GASBOR much more frequently than DAMMIN, I will only
discuss ab initio envelope determination using GASBOR for the remainder of this
section.
GASBOR uses processed scattering data to create an ab initio envelope with
chain like dummy residues121. These dummy residues are separated by ~3.8 Å, which is
the average distance between two Cα atoms in proteins. GASBOR uses the output from
the best GNOM analysis, described in the previous section, and the number of residues
in the protein as the input for ab initio calculations. The program creates an initial model
by placing the dummy atoms randomly within a sphere of radius Dmax/2 + r0, where Dmax
is obtained from the GNOM analysis and r0 is the origin. Next, a single dummy residue is
selected at random and moved to a new location, while always adhering to the ~3.8 Å
distance restraint. The new position of this residue can either be selected or rejected
based on a goal function E(r) = χ2 + αP(r). This function takes into account the fit of the
theoretical scattering curve from the model to the regularized scattering data from
GNOM (χ2) and the realism of the model using a penalty term (P(r)) where α is a
weighting term.
𝑥2 =
2
𝐼 −𝐼
1
∑𝑛1 � 𝐷𝑅 𝑒𝑥𝑝 �
𝜎
𝑛−1
𝑃(𝑟) = ∑𝑘[𝑊(𝑅𝑘 )(𝑁𝐷𝑅 (𝑅𝑘 ) − 〈𝑁(𝑅𝑘 )〉)]2 + 𝐺(𝑟) + [max {0, (|𝑟𝑐 | − 𝑟0 )}]2
(3.5)
(3.6)
Equation 3.5 defines χ2. IDR is the intensity from the theoretical scattering curve
derived from the dummy residue model. Iexp is the experimental scattering and σ is the
error from the experimental intensities. The penalty term P(r) contains three parts that
ensures the realism of the model. The first part [𝑊(𝑅𝑘 )(𝑁𝐷𝑅 (𝑅𝑘 ) − 〈𝑁(𝑅𝑘 )〉)]2 describes
the similarity of the model to the average distribution of nearest neighbors based on the
97
Figure 46. Ab initio envelope calculation using GABSOR.
A-C) Progression through GASBOR using scattering data for the PP1:spinophilin complex. A
represents the initial model and C represents the final model. The DR model is shown as pink
spheres. The red line is the theoretical scattering curve from the model, the green line is the
regularized scattering data from GNOM and the experimental data are blue circles.
98
radius. The second part, G(r), describes the connection of the dummy residues making
sure that residues are spaced ~3.8 Å apart. The third part [max {0, (|𝑟𝑐 | − 𝑟0 )}]2
describes the distribution of the residues around r0, ensuring that the center of mass lies
close in proximity to the origin. The new position of the residue is selected for when
ΔE(r)= E(r’)- E(r) is < 0 where E(r) is the goal function for the original position and E(r’) is
the goal function for the new position of the residue. If ΔE(r) is > 0 than the position of
the residue is rejected. A new residue is selected at random and moved to a new
position. This process is repeated until ΔE(r) no longer produces a value of < 0. A
summary describing a GASBOR calculation is shown in Figure 46. The initial model is
shown in A. The final model is shown in C and is output as a PDB file with both dummy
residues and solvent atoms. Since each calculation starts with randomized positions of
dummy atoms it is important to repeat the process typically 10-20 times to generate a
series of models which represent the error in generating a model from a randomized
initial model. The envelopes from each calculation are averaged using the program
DAMAVER, which also gives a measure for the similarity of each envelope, the
normalized spatial discrepancy (NSD)122. The NSD 𝜌(𝑆1 , 𝑆2 ) is the normalized average of
the minimum distance between a point in the first structure with all points in second
structure (equation 3.7). A value of 1 or less indicates the models are identical.
1
2
1
𝑁1
� ∑𝑖=1
𝜌2 (𝑠1𝑖 , 𝑆2 ) +
𝑁1 𝑑22
𝜌(𝑆1 , 𝑆2 ) = {� � ��
1
1
𝑁2
2 (𝑠
2
∑
)
�
𝜌
,
𝑆
�}
2𝑖 1
2
𝑖=1
2 𝑑1
�𝑁
(3.7)
The more similar the models are the better the averaged envelope calculation is.
Calculations that fall outside the standard deviation of the NSD values are rejected and
not used for the determination of the average structure.
5.4. SAXS compliments high resolution techniques.
99
One of the most significant strengths of SAXS is that it can be used in combination with
X-ray crystallography and NMR spectroscopy to gain additional structural information. In
this regard, SAXS has been used to investigate flexibility between domains with known
structures and to guide for protein docking experiments. These techniques require the
calculation of scattering curves from high resolution structural models and comparison of
these curves to experimental scattering data. There are two commonly used programs
for the calculation of scattering curves from protein structures: FOXS and CRYSOL123,124.
FOXS, created by the SIBYLS beamline at LBL, is used as part of the flexibility and
docking programs BILBOMD and BILBODOCK125. CRYSOL developed by the Svergun
group at the EMBL is used with the docking program SASREF and the modeling
program BUNCH. Both FOXS and CRYSOL were used to calculate theoretical scattering
curves from the PP1:spinophilin holoenzyme structure (Figure 47). The programs
calculate similar scattering curves from a q value of 0 to ~0.15. The scattering curves
then diverge. This divergence is the result of different interpretations in regard to the
solvent layer around the protein. Despite the divergence, both curves fit the experimental
scattering data for the PP1:spinophilin complex similarly, with a χ of 3.32 for FOXS and
3.31 for CRYSOL.
A recently developed program, BILBOMD, has been used to create structural
models for proteins with flexible domain linkers and loops125. The program uses
experimental SAXS data to select for models that agree best with the experimental data.
Again, I will use the PP1:spinophilin holoenzyme as an example to explain a BILBOMD
calculation. Comparison of the two molecules in the asymmetric unit of the
PP1:spinophilin holoenzyme structure reveals that electron density was only detectable
for one of the two PDZ domains of spinophilin (PDB ID 3EGG). This suggests that the
linker between the spinophilin PP1 binding and PDZ domains is flexible and that the
100
PDZ domain in the structure is
restricted in conformational space. To
investigate the flexibility of this linker,
we have recorded SAXS data for the
holoenzyme and used the program
BILBOMD to analyze the structure.
The
Figure 47. Theoretical scattering data.
Scattering data was computed for the
PP1:spinophilin holenzyme using FOXS (blue)
and Crysol (red).
BILBOMD
input
model
represents the entire sample used in
scattering
experiments.
Here,
we
have used the molecule of the
PP1:spinophilin holoenzyme which includes the PDZ domain. First, rigid domains are
identified. Of these domains, one is defined as fixed and the remaining domains are
allowed to move as rigid bodies. In the PP1:spinophilin holoenzyme, two rigid domains
were defined based on the electron density for molecule two in the asymmetric unit. PP1
7-330 and spinophilin 424-489 (the PP1 binding domain) were defined as the fixed
domain and spinophilin 495-583 (the PDZ domain) was defined as the moving domain.
Residues 490-494 were defined as the flexible linker between the domains. An energy
minimization of this initial model is the first step of the calculation, which includes
clearing of steric hindrances if any exist. The model is then heated to 1500 K and MD
simulations are performed using the CHARMM force field. The domains are treated as
rigid bodies while only the flexible linker, residues 490-494, is allowed to move. The
program generates 200 MD conformations for each Rg. In this case, I have used an Rg
range of 35-45 Å with a step of 2 Å to produce 1200 conformations. Theoretical
scattering curves are calculated using FOXS for each calculated MD conformation. The
scattering curves are directly compared to the experimental scattering and the single
best-fit model and a minimal ensemble model (MES) are determined. MES uses an
101
Figure 48. Investigating flexibility using SAXS.
BILBOMD was used to model the flexibility between the spinophilin PP1 binding and PDZ
domains. Scattering from the starting model and final model are shown in blue and red
respectively against the experimental data (black). The starting model is shown in blue (PDB
ID 3EGG). The final MES model (red) contains two structures representing 82.3% and 17.7%
of the populations in solution.
algorithm, which determines the minimal number of conformations (2,3,4 or 5) that
optimally describe the experimental SAXS data. The determination of MES is important,
since SAXS is performed in solution and the experimental data represent the average of
all the conformations in solution. Therefore, the description of the experimental SAXS
data by a single conformation may not be an accurate representation of the protein in
solution. For the PP1:spinophilin holoenzyme, the MES includes two structures (Figure
48). Both structures contain a weighting term, which describes the relative populations of
each structure contributing to the fit of the experimental data. Scattering from the input
model fits the experimental data with a χ2 of 11.1, while the MES fits the experimental
data with a significantly improved χ2 of 2.3. This demonstrates that the crystal structure
contains a conformation that is restricted in its conformational space. Therefore, while
the crystal structure represents the local structure well, the inherent flexibility of the
protein in solution is misrepresented. Using the combination of the crystal structure and
the MES model from BILBOMD, we now have a more complete view of the protein than
we would have gained from using either technique alone.
102
I also want to provide an example for the usefulness of SAXS data when
combined with NMR spectroscopy. In NMR spectroscopy, it can be difficult to define the
homodimer interface of two macromolecules if only a small number of intermolecular
NOEs are observed. This can be even more difficult in the case of RNA interfaces where
long distance NOEs are less abundant. To this end, Zuo et al 2008 have demonstrated
that SAXS can be used to refine the dimer interface of NMR structures in the absence of
intermolecular NOEs126. The structure of the complex of a tetraloop receptor RNA
homodimer was determined using NMR spectroscopy. The orientation of the two
subunits was refined using 66 intermolecular NOEs and 18 RDCs measured in a single
alignment media. The orientation was also
refined
using
SAXS
in
place
of
the
intermolecular NOE restraints. Since the
RDCs were only measured in a single
alignment media, four orientations of the
domains can satisfy the RDC restraints.
Figure 49. Overlay of tetraloop receptor
homodimer.
The structure was refined using
intermolecular NOEs and RDCs (red) or
RDCs and SAXS (Green). Adapted from
Zuo et al 2008.
However, using information about the overall
shape of the complex obtained from the
SAXS, the number of orientations was
reduced to one. The structures were refined
with and without intermolecular NOEs and superimpose with a backbone RMSD of ~0.4
Å (Figure 49)126.
5.5. Time-resolved SAXS.
Time-resolved experiments are commonly detected using the change of fluorescence
over time. While this gives information about the kinetics of the reaction or
rearrangement, these experiments contain very little information about structural
103
rearrangements. Rapid mixing devices
have also been used in SAXS for timeresolved
experiments.
These
experiments have been used to monitor
induced conformational rearrangements
in proteins, as well as protein folding. I
will discuss one example briefly. The
folding of cytochrome-c was studied by
transfer from the acid denatured state at
pH 2.0 to the natively folded state at pH
4.5 using a rapid mixing device127.
Time-resolved SAXS experiments were
Figure 50.Time-resolved SAXS experiments
of cytochrome-c.
A) Scattering data shown as kratky plots for
160us (red circle) to 500ms (black squares). B)
Radius of gyration calculated using the Guinier
approximation against time. Adapted from
Akiyama et al 2002.
recorded with a resolution of 160 μs on
28.9 mg/ml cytochrome-c with a q range
of 0.0056 to 0.0367. This q range
contains the highest scattering intensity and allowed for the determination of the Rg
using the Guinier approximation in real-time throughout the experiment (Figure 50).
These experiments identified two intermediates within the folding process with Rgs of
20.5 and 17.7 Å. The rate of formation of these intermediates confirmed previous
fluorescence data. However, until the time-resolved SAXS experiments were measured
and analyzed, no structural information was available for these intermediate states. p(r)
functions were determined from the SAXS data showing that intermediate one has a
Dmax of 66 Å and intermediate two has a Dmax of 58 Å and that the largest folding
collapse occurs between intermediate two and the native state.
5.6. Summary.
104
SAXS can be used to obtain low resolution structural information (~10 Å) on dilute
macromolecules in solution. A flowchart summarizing the process of data acquisition and
analysis as described in this chapter is shown in Figure 51. Since SAXS is performed on
dilute samples in solution, sample preparation is much simpler than for X-ray
crystallography and NMR spectroscopy, which requires crystallization and the
purification of isotopically labeled proteins, respectively. Additionally, since SAXS is
performed in solution, samples in near physiological condition can be used. Recent
computational
advances,
including
programs
created
for
ab
initio
envelope
determination, flexibility modeling, and protein docking, have been used in combination
with high resolution techniques to gain novel insights into various macromolecules.
These novel insights have led to more realistic descriptions of protein structures in
solution.
105
Figure 51. Flowchart for SAXS data acquisition and analysis.
SAXS data can provide a wealth of information including the Rg, overall shape of a molecule,
and even an ensemble model (green squares). However, it is important to verify the quality of
the data. The above flowchart details important questions (orange diamonds) that should be
considered as checkpoints to proceed through different steps in SAXS data analysis (blue
square).
Chapter 6. SAXS analysis of PP1 complexes.
106
107
6.1. Structural diversity in free and bound states of intrinsically disordered
protein phosphatase 1 regulators.
This manuscript is currently in press in the journal structure. My contribution to this work
is the SAXS analysis of the PP1:I-2 complex.
Joseph A. Marsh1, Barbara Dancheck2, Michael J. Ragusa2, Marc Allaire3, Julie D.
Forman-Kay1,* & Wolfgang Peti2,*
1
Molecular Structure & Function, Hospital for Sick Children, Toronto, Ontario M5G 1X8,
Canada & Department of Biochemistry, University of Toronto, Toronto, Ontario M5S
1A8, Canada; 2Department of Molecular Pharmacology, Physiology and Biotechnology &
Department of Chemistry, Brown University, Providence, RI, 02912, USA; 3National
Synchrotron Light Source, Brookhaven National Laboratory, Upton, NY, 11973-5000,
USA.
*to whom correspondence should be addressed:
Julie D. Forman-Kay: Molecular Structure & Function, Research Institute, Hospital for
Sick Children, 555 University Avenue, Toronto, Ontario M5G 1X8, Canada; Phone 416813-5358; Fax 416-813-5022; forman@sickkids.ca.
Wolfgang Peti: Department of Molecular Pharmacology, Physiology and Biotechnology &
Department of Chemistry, Brown University, 70 Ship Street, GE-3, Providence, RI,
02912, USA; Phone 401-863-6084; Fax 401-863-6087; Wolfgang_Peti@brown.edu.
108
6.1.A. Abstract.
Complete folding is not a prerequisite for protein function, as disordered and partially
folded states of proteins frequently perform essential biological functions. In order to
understand their functions at the molecular level, we utilized diverse experimental
measurements to calculate ensemble models of three non-homologous, intrinsically
disordered proteins: I-2, spinophilin and DARPP-32, which bind to and regulate protein
phosphatase 1 (PP1). The models demonstrate that these proteins have dissimilar
propensities for secondary and tertiary structure in their unbound forms. Direct
comparison of these ensemble models with recently determined PP1 complex structures
suggests a significant role for transient, pre-formed structure in the interactions of these
proteins with PP1. Finally, we generated an ensemble model of partially disordered I-2
bound to PP1 that provides insight into the relationship between flexibility and biological
function in this dynamic complex.
6.1.B. Introduction.
In recent years, many exceedingly flexible proteins with important biological functions,
known as intrinsically disordered proteins (IDPs), have been described61,128-135. IDPs lack
the typical hydrophobic cores that generally stabilize folded proteins, and their highly
dynamic nature prevents their description as single, rigid structures. However, structural
studies of a number of IDPs indicate that they are not random-coil polymers; rather they
often have preferences for transient secondary and tertiary structure elements. It has
been proposed that pre-organization and spatial restriction in IDPs expose primary
contact sites to enable faster and more effective binding to target molecule136-141. Binding
is often accompanied by stabilization of transient secondary structure or folding-uponbinding transitions78. However, the abundance and diversity of transient structural
elements in IDPs and their importance in binding interactions are still unclear. In
109
addition, it is uncertain whether functionally related IDPs (e.g. IDPs that bind identical
targets) show conservation of functional motifs with conserved structural features, as is
commonly seen in folded proteins.
Protein phosphatase 1 (PP1) is a major serine/threonine phosphatase with roles
in diverse cellular processes such as muscle contraction and cell signaling34. The activity
of PP1 is controlled by two types of proteins: targeting proteins, which direct PP1 to
specific subcellular locations and alter its substrate specificity, and inhibitor proteins.
Nearly all PP1 regulators interact with PP1 via a common “RVxF” motif71. However, this
motif can be found in one third of all eukaryotic proteins, many of which do not interact
with PP1, indicating that, although it is necessary for the interaction with PP1, it alone is
not sufficient to provide specificity for PP1 binding142. Biochemical data34,71 and
structures of PP1 in complex with its targeting proteins MYPT143, spinophilin58 and the
inhibitor protein inhibitor-2 (I-2)62 show that PP1 regulators interact with multiple sites on
PP1. Furthermore, both types of PP1 regulators are known to contain intrinsically
disordered regions57,58. The flexible nature of these IDPs allows them to wrap around
PP1, forming contacts with distal sites and burying large surface areas43,58. While
multiple, distant interaction sites on PP1 provide some specificity in the interaction with
its intrinsically disordered regulators, it is unclear whether specificity is enhanced by preformed structural motifs present in the IDPs.
Here we describe the comprehensive structural analysis of three PP1 regulators.
I-2, one of the most ancient PP1 regulatory proteins, is ubiquitously expressed from
yeast to humans143,144 and is important for the proper regulation of cell division145.
DARPP-32 (dopamine- and cyclic AMP-regulated phosphoprotein with molecular weight
of 32 kDa) is a pseudo-substrate inhibitor of PP1 that requires phosphorylation on Thr34
to become a nanomolar PP1 inhibitor. It has been established as an essential link
between the neuronal dopaminergic and glutamatergic signaling pathways. In this
110
position it plays a critical regulatory role for numerous neurological diseases and a large
number of drugs of abuse146. Finally, spinophilin is a PP1-targeting protein46. The
PP1:spinophilin holo-enzyme dephosphorylates Ser845 on the GluR1 AMPA receptor
subunit and regulates the closed/open state of the AMPA receptor. It is therefore
essential for the regulation of long-term potentiation51,68,147. Furthermore, spinophilin
facilitates dephosphorylation of doublecortin by PP1 to mediate microtubule bundling at
the axonal wrist, playing an important role in the organization of the neuronal
cytoskeleton72.
Previously we reported nuclear magnetic resonance (NMR) spectroscopy data
on the dynamics and transient secondary and tertiary structural preferences of I-2 and
DARPP-3257 and the PP1-binding domain of spinophilin58. Here we have investigated
the relationship between transient structure and the literature reported function in these
three PP1 regulators by calculating detailed ensemble models of them using the
program ENSEMBLE148-150. These models demonstrate that functionally related IDPs
can have diverse structural properties. In addition, the availability of PP1-bound crystal
structures for I-2 and spinophilin58,62 allows for the direct comparison of the free and
bound states of these proteins. This has enabled us to investigate the role of pre-formed
transient structure in the molecular recognition of PP1. Finally, we have calculated an
ensemble model of the PP1:I-2 complex, in which 75% of I-2 remains disordered,
providing insight into both the function of I-2 and the structural properties of highly
dynamic complexes.
6.1.C. Results.
6.1.C.1. Calculation of unbound intrinsically disordered ensemble models.
ENSEMBLE was used to calculate models of I-2 residues 9-164 (I-29-164), the spinophilin
PP1-binding domain residues 417-494 (spinophilin417-494) and DARPP-32 residues 1-118
(DARPP-321-118). Three independent ensembles were calculated for each system, with
111
the final ensembles containing 10-24 structures each according to the simplestensemble approach, in which the smallest number of conformers possible are fit to the
experimental restraints149. The experimental restraints included NMR chemical shifts,
distance restraints derived from PRE measurements,
15
N R2 relaxation rates and
hydrodynamic radii (Rh) from dynamic light scattering57,58. These four different
experimental restraint types report on diverse structural properties including secondary
and tertiary structure and overall compaction. Table 7 shows the number of experimental
restraints used for each system, the agreement between these experimental restraints
and the values predicted from the final calculated ensembles. Table 6 shows the general
structural properties of the ensembles compared to reference “coil” ensembles149,151.
Plots comparing our experimental measurements to the ensemble-predicted values are
shown in the Supporting Information (Figure 75-Figure 81).
6.1.C.2. Unbound ensembles possess secondary structure of varying similarity to
the PP1-bound complexes.
I-29-164
Calculated
Number of
structures
Radius of
gyration (Å)
Hydrodynamic
radius (Å)
Solvent
accessible
surface area
(Å2)
Coil
Spinophilin417-494
Calculated
Coil
DARPP-321-118
Calculated
Coil
21.00 ± 1.63
21 ± 0
10.33 ± 0.47
11 ± 0
14.00 ± 0.82
14 ± 0
34.59 ± 0.98
36.34 ±
1.57
16.39 ± 0.33
25.71 ±
1.68
28.28 ± 0.52
31.94 ±
2.03
33.09 ± 0.08
34.16 ±
0.63
19.92 ± 0.05
28.09 ± 0.06
30.15 ±
0.74
18326 ±
200
8267 ± 155
12595 ± 57
13783 ±
213
17259 ± 133
24.89 ±
0.63
9213 ±
144
Table 6. General properties of calculated unbound intrinsically disordered ensembles
and TraDES ‘Coil’ ensembles.
Average values and standard deviations for three independently calculated ensembles and
100 TraDES ‘Coil’ ensembles are presented.
The residue-specific secondary structure content of the ensembles, including the fraction
112
Restraint type
13

C chemical shifts

C chemical shifts
13
C´ chemical shifts
13 
H chemical shifts
13 N
H chemical shifts
PREs
Rh
15
a
N R2
13
#
151
143
151
151
139
631
1
141
I-29-164
RMSD
0.35 ppm
0.38 ppm
0.41 ppm
0.08 ppm
0.17 ppm
0.99 Å
1.9 Å
0.85
Spinophilin417-494
#
RMSD
77
0.35 ppm
74
0.39 ppm
76
0.41 ppm
75
0.08 ppm
67
0.17 ppm
42
0.86 Å
1
1.9 Å
67
0.62
DARPP-321-118
#
RMSD
112 0.34 ppm
107 0.39 ppm
105 0.42 ppm
112 0.08 ppm
99 0.16 ppm
437
0.70 Å
1
1.8 Å
102
0.68
I-2:PP1 complex
#
RMSD
100
0.33 ppm
94
0.33 ppm
100
0.40 ppm
100
0.08 ppm
93
0.17 ppm
b
514
20.5 Å
0
92
0.68
Table 7. List of experimental restraints used in ENSEMBLE calculations and agreement
between experimental and predicted values.
a
15
The N R2 restraints are applied by enforcing a correlation between experimental values
and contacts within the ensemble and thus these values represent a Pearson correlation, not
b
an RMSD. The PRE restraints were utilized in a very different way in the I2:PP1 complex
than for the unbound intrinsically disordered ensembles which accounts for this large RMSD
value – see the Methods for more details.
of residues within the broad α, left-β and right-β regions of Ramachandran space
(previously defined149) and the fraction of residues identified as α-helical by STRIDE152
are presented in Figure 52. The left-β region includes Φ angles <-100°, typically
associated with β-strand formation, while the right-β region includes Φ angles ≥ -100°,
referred to as the polyproline II (PPII) region. Secondary structure elements identified in
the PP1-bound complexes of I-2 and spinophilin are shown at the top of the plots.
The most notable secondary structure element observed in the three ensembles
is the ~70% populated α-helix in I-29-164 spanning residues 130-142. This α-helix
corresponds directly to an α-helix present in the PP1:I-2 complex, which folds over the
PP1 active site and is important for PP1 inhibition. Interestingly, two other α-helices
(residues 47-54 and 153-164) seen in the PP1:I-2 complex are not extensively populated
in the unbound I-29-164 ensembles, demonstrating that transient secondary structure
formation in free states is not necessary for all helices observed in bound complexes.
Conversely, two transient α-helices are observed from residues 36-40 (~20% populated)
and 97-105 (~35% populated) that are in regions of I-2 that are not visible in the PP1:I-2
complex structure.
113
Spinophilin417-494
has
less
secondary structure than I-29-164. However,
a ~25% populated α-helix (residues 477487) corresponds perfectly to the only αhelix present in the PP1-bound spinophilin
structure. In addition to this helix, two βstrands (residues 430-434 and 456-460)
are
also
observed
in
PP1-bound
spinophilin. Interestingly, a peak in the leftβ (i.e. β-strand region) plot of spinophilin
(residues 456-461) corresponds very well
Figure 52. Secondary structure content
of calculated ensembles.
(a) I-29-164, (b) spinophilin417-494 and (c)
DARPP-321-118. Lines represent the α-helix
populations (green), populations in the
broad α (blue), left-β (red) and right-β
(purple) regions of the Ramachandran
diagram. All values are plotted with a threeresidue smoothing. Secondary structures
observed in the bound-state crystal
structures of I-2 and spinophilin are shown
above the plots with α-helices in green, βstrands in red and other observed regions
in blue.
with this second β-strand. However, there
is no significant peak in the left-β plot
corresponding
to
the
first
β-strand
(residues 430-434). Thus, while the region
corresponding to the second β-strand
tends to sample strand-like Ramachandran
angles, there is no evidence for hydrogenbonded β-sheet formation in disordered
spinophilin417-494.
Finally, for DARPP-321-118, a ~30% populated α-helix from residues 22-29 and a
10% populated α-helix near the C-terminus (residues 97-114) were identified. However,
the lack of a PP1:DARPP-32 complex structure prohibits further analysis of this data.
6.1.C.3. Tertiary contacts in the unbound ensembles.
The identification and interpretation of tertiary structure in IDPs is much more
complicated than secondary structure. Although contact plots are commonly used to
114
demonstrate tertiary structure in folded and disordered proteins, low populated contacts
within a heterogeneous ensemble can be difficult to discern. Thus, to analyze tertiary
structure in our ensembles, we used a clustering algorithm to divide the ensembles into
clusters of structurally similar conformers149. The tertiary structure of each cluster is
presented in Figure 53 using fractional contact plots149 which show tertiary contacts
within these clusters. Contact plots for the PP1-bound structures of I-2 and spinophilin
are shown in the bottom halves of the plots. Populations and structural properties of the
clusters are presented in Table 14. For both I-29-164 and DARPP-321-118, a large number
of PRE distance restraints were utilized in the ENSEMBLE calculations (Table 7).
However, the spinophilin PP1-binding domain is particularly difficult to work with as it is
rapidly degraded by proteases. Thus, accurate data from only one spin-label site was
available.
Contact patterns observed in the three major clusters of I-29-164 were analyzed.
Cluster 1 is highest populated (44%) and most compact. It is dominated by contacts
involving the C-terminal region of I-29-164, particularly residues 120-164. Cluster 2 (17%
populated) has significant tertiary contacts between regions around residues 50 and 90.
Cluster 3 (14% populated) has contacts between its N-terminus and the region around
residue 80. Direct comparison of tertiary structure between I-29-164 clusters and PP1bound I-2 is difficult as electron density for I-2 was only observed for three isolated
segments which lack intramolecular contacts. Nevertheless, although none of these
clusters resemble the bound state, it is interesting that I-2 regions that directly interact
with PP1 in the crystal structure (residues 11-16, 42-54 and 128-167) participate in a
substantial fraction of the contacts detected in unbound I-2.
Upon binding to PP1, spinophilin417-494 undergoes a folding-upon-binding
transition and forms secondary and tertiary structure elements including two β-strands
that extend a large β-sheet in PP1 and an α-helix. Spinophilin417-494 cluster 1 (68%
115
Figure 53. Significantly populated clusters and their fractional contact plots.
(a) I-29-164, (b) spinophilin417-494 and (c) DARPP-321-118. Fractional contact plots represent the
fractional formation of contacts between pairs of residues, with a contact being defined as any
two heavy atoms being within 10 Å of each other. The upper halves of the contact plots show
contacts present in the clusters while the lower halves show contacts present in PP1-bound I2 and spinophilin. White represents regions not present in the PP1-bound structures, either as
no electron density was detected or because different construct lengths were used in the
studies. Structures are colored as follows: I-2 – residues 6-58 red, 59-111 green, 112-164
blue; spinophilin – residues 417-442 red, 443-468 green, 469-494 blue; DARPP-32 – residues
1-39 red, 40-78 green, 79-118 blue.
116
populated) has a large number of tertiary contacts, particularly near the N- and Ctermini; none of these contacts resemble the topology of PP1-bound spinophilin. Cluster
2 (18% populated), on the other hand, has tertiary interactions that clearly resemble the
β-strand contacts of the PP1-bound state, seen as contacts perpendicular to the
diagonal in the contact plot. Although this cluster also contains contacts not seen in the
bound state, this suggests that free spinophilin417-494 has a minor population in which
Figure 54. PP1:I-2 NMR study.
1
15
15
15
2
(a) 2D [ H, N] HSQC/TROSY spectra of unbound N-labeled I-29-164 (black) and N, Hlabeled I-29-164 in complex with unlabeled PP1 (red). Spectra were aligned by overlaying the
ε1
peaks belonging to the N of Trp46. (b) Chemical shift changes in I-29-164 upon addition of
PP1 are small and are primarily localized to regions observed in the crystal structure of the
bound complex. The chemical shift differences (Δδ) between peaks in unbound and bound
spectra were calculated as
(∆δ H )2 +  ∆δ N 
 10 
2
.
117
contacts resembling the bound state are beginning to form. Notably, these contacts are
close to the position of the spin label, providing confidence in their accuracy.
Three clusters are identified for DARPP-321-118. Cluster 1 (50% populated) is very
compact. Cluster 2 (29% populated) is dominated by contacts involving the C-terminal
region. Finally, cluster 3 (12% populated) is extended and has only very few tertiary
contacts. Although these results suggest interesting tertiary structure in unbound
DARPP-321-118, further analysis is limited due to the lack of a PP1:DARPP-32 complex
structure.
Notably, all three IDPs have at least one cluster with tertiary structures involving
the N- and/or C-termini. This is interesting, given the observation that contacts involving
the termini should be more likely due to the excluded volume of the polymer chain153.
This was also previously noted in the unfolded state of the drkN SH3 domain149. Thus, it
seems likely that contacts involving the termini are a common structural feature of IDPs.
6.1.C.4. Calculation of an ensemble model of the partially folded PP1:I-2 complex.
As noted earlier, electron density for I-2 in the PP1:I-2 complex was only detected for
~25% of the protein and, thus, I-2 remains largely disordered, even upon binding to
PP162. The 2D [1H,15N] HSQC/TROSY spectra of unbound
15
N-labeled I-2 and
15
N,2H-
labeled I-2 in complex with unlabeled PP1 are shown in Figure 54. Chemical shift
changes between unbound I-2 and PP1-bound I-2 were limited to residues that are
present in the PP1:I-2 crystal structure. Signals of regions with no electron density in the
crystal structure retained the characteristic traits of disordered proteins, including very
narrow line-widths typical for fast reorientation in solution. This strongly suggests that
regions not observed in the PP1:I-2 crystal structure are highly similar in the free and
bound states of I-2. Therefore, we used ENSEMBLE to determine a model of the full
118
Figure 55. The PP1:I-2 complex.
(a) PP1 is shown in a grey surface representation while I-2 is shown in various colors: the
region observed in the PP1:I-2 crystal structure is shown in light blue while the three
disordered regions of I-2, residues 6-10, 17-41 and 55-127, are colored orange, green and
red, respectively. Thr72 is shown in black. (b) Fitting of the PP1:I-2 complex model into the
crystallographic unit cell. The most extended conformer of the PP1:I-2 complex model was
overlaid with the PP1:I-2 crystal structure, and unit cell information from the crystal structure
was applied to the complex model. Symmetry mates were then added to determine whether
the PP1:I-2 model fits into the crystallographic unit cell. (c) Fractional contact plots for the
unbound I-2 ensembles (top left) and the PP1-bound I-2 ensemble model (bottom right). The
color scale is the same as in Figure 2.
PP1:I-2 complex using the assumption that experimental restraints from the free state of
I-2 could be used to describe the disordered regions of PP1- bound I-2.
The resulting PP1:I-2 complex model is shown in Figure 55A. In Figure 55B, the
most extended conformation from the model is shown to fit into the PP1:I-2
crystallographic unit cell, providing substantial support for this model. If I-2 adopted a
more compact conformation upon PP1 binding, our extended model would be unlikely to
fit into the unit cell. The most remarkable feature of this complex is the long loop
119
connecting
residues
55-127
(Figure
55A).
Although
this
loop
is
somewhat
heterogeneous, with varying conformations between the different ensemble conformers,
a significant fraction of structures adopt a distinct loop structure. This loop contains
residue Thr72 which is essential for the biological function of the PP1:I-2 complex, as
phosphorylation of Thr72 by GSK-3β reactivates the PP1:I-2 complex63. Interestingly,
this loop also contains a transiently populated α-helix (residues 97-105, ~35%
populated). Finally, although most conformers show this loop region extending away
from PP1, several conformers show the loop closer to PP1, indicating that there may be
transient interactions between this loop and PP1which potentially adding stability to the
overall complex.
4.1.C.5. The PP1:I-2 complex: assessment using SAXS measurements.
PP1:Inhibitor-2
Guinier approximation
Rg (Å)
P(r) function calculation
q-range (Å-1)
Rg (Å)
Dmax (Å)
Structure modeling
χ
NSDa
NSDe
29.8 ± 0.4
0.015-0.300
30.2
100
1.22 ± 0.02
1.13 ± 0.03
1.50 ± 0.14
Table 8. Statistics for SAXS analysis of the
PP1:I-2 complex.
NSDa is the discrepancy between the ab
initio models. NSDe is the discrepancy
between the envelope and the ensemble
models.
To experimentally validate the PP1:I-2
complex ensemble model we utilized
small angle X-ray scattering (SAXS)
measurements, which report on the
overall size and shape of a protein in
solution.
Using
approximation
of
the
four
Guinier
independent
SAXS samples a radius of gyration
(Rg) of 29.8 ± 0.4 Å was calculated for
PP1:I-2 (Figure 56A, all measurement
statistics are reported in Table 8). This is in excellent agreement with the average Rg
(28.5 ± 1.6 Å) for the nineteen calculated ensemble conformations of PP1:I-2.
Furthermore, theoretical scattering curves were generated for each ensemble model,
which fit the experimental scattering data well with an average χ of 2.1 ± 0.9 (Figure
56B). Finally, a molecular envelope of the PP1:I-2 complex was calculated using
120
Figure 56. SAXS data and analysis for the PP1:I-2 complex.
(a) The Guinier region of the scattering curve for 0.8 mg/ml and 0.5 mg/ml PP1:I-2. (b)
Average theoretical scattering curve for the ensemble models (red line) fit to the experimental
scattering data for PP1:I-2 (squares). (c) Overlay of the SAXS envelope (surface
representation, blue) with the PP1:I-2 ensemble models. PP1 is shown as a cartoon
representation in black and I-2 is shown as a cartoon representation in red.
GASBOR
121
by averaging sixteen independent ab initio models. These models
represent the scattering data well with an overall fit of χ = 1.22 ± 0.02 (Figure 82). The
models were generated with a maximum length of 100 Å as determined by the analysis
of the pair-distance distribution function (Figure 82). The individually calculated ab initio
models were averaged and the resulting envelope was aligned with the nineteen
ensemble models of PP1:I-2. The envelope and ensemble models are visually and
statistically similar, aligning with an average normalized spatial discrepancy (NSD)154 of
1.50 ± 0.14 (Figure 56C). The calculated PP1:I-2 envelope represents the largest
population of the ensemble models with the distinct loop structure of I-2 extended away
from the center of PP1. Although the minor populations deviate slightly from the overall
121
envelope, this is likely due to the averaging of SAXS data in solution. Thus the SAXS
data supports the correctness of the overall ensemble model of the PP1:I-2 complex.
6.1.D. Discussion.
6.1.D.1 Structural properties of the unbound ensembles and their implications for
PP1 binding.
Two questions are fundamental for the molecular understanding of IDP interactions: 1)
Are the conformational propensities of unbound states similar to their bound structures?
2) What is the role of transient secondary and tertiary structure for specific interactions
with their targets? In the “conformational selection” model of binding, the IDP adopts the
bound-state conformation in its free state at a limited population155. Only these prepopulated binding-competent conformations will interact and thus shift the equilibrium
towards the bound state. In contrast, in the “induced fit” model, the bound conformation
only becomes energetically accessible in the presence of the binding partner156. Current
evidence suggests that both induced fit and conformational selection mechanisms are
important for IDP function157. Certainly the interaction of any IDP undergoing a large
folding-upon-binding transition must involve an induced-fit component, as the probability
of all elements of the bound-state structure occurring simultaneously with the correct 3D
geometry seems low. However, this does not exclude conformational selection, as
individual elements of secondary or tertiary structure might be fractionally populated and
selected for during the binding process.
The most remarkable feature of unbound I-29-164 is the presence of a highly
populated α-helix (residues 130-142) that is also observed in the PP1:I-2 complex. This
α-helix is vital for PP1 inhibition, as it folds over and blocks the PP1 active site. The high
population of this helix suggests it is likely to be critical for the association with PP1. A Cterminal adjacent helix (residues 153-159) is ~10% populated in the unbound state and
primarily folds upon binding to PP1, possibly mediated by the binding of the first I-2 helix.
122
Additional secondary structural elements observed in the I-2-bound state, including an αhelix (residues 47-54), are lesser populated in the unbound state and are thus induced
upon binding. In addition, the transient intramolecular contacts of unbound I-29-164 show
no significant resemblance to I-2 in the complex, although many contacts are in regions
that directly interact with PP1. Therefore, the initial binding of the C-terminal helix may
initiate the release of many of these intramolecular contacts in unbound I-2. In particular,
the region surrounding Trp46, which includes the RVxF motif, may become more
accessible, thus permitting complex formation.
Spinophilin417-494 contains a ~30% populated α-helix in the unbound state near
the C-terminus that becomes fully formed upon PP1-binding. Furthermore, a region
corresponding to the second β-strand (residues 456-461) shows a significant peak in βstrand-like Ramachandran angles, suggesting it transiently populates conformations that
are optimal for PP1 binding. Finally, there is evidence for low populated tertiary contacts
occurring in spinophilin417-494 that resemble its structure in the complex. Thus, many of
the intramolecular structural features observed in PP1-bound spinophilin are also
detected in unbound spinophilin, suggesting that conformational selection is likely
important for PP1 binding.
Although no structure of the PP1:DARPP-32 complex has been reported, it is
likely that transient structure present in the free state of DARPP-32 is important for this
interaction. In particular, the fractionally formed α-helices probably play an important role
in PP1 binding, just as α-helices identified in the ensembles of both I-2 and spinophilin
correspond closely to their PP1-bound crystal structures. The ~30% populated α-helix
lies between the RVxF motif of DARPP-32 (KKIQF, residues 7-11) and Thr34.
Phosphorylation of Thr34 by PKA activates DARPP-32 as a PP1 inhibitor158. Just Cterminal to this transient α-helix is a stretch of basic residues (RRRRPT, residues 2934), which complement the acidic groove of PP1. The acidic groove, one of three
123
predicted substrate binding grooves on the PP1 surface, connects the RVxF motif
binding site and the active site of PP1. Thus, binding of the helix and the C-terminal
positively charged residues is likely important in the positioning of Thr34 in the active site
of PP1 for inhibition.
6.1.D.2. Biological implications of the PP1:I-2 complex model.
The ensemble model of the PP1:I-2 complex provides molecular insight into the dynamic
regions of I-2 that are not observed in the crystal structure, but which fulfill essential
biological functions. The most notable structural feature of the complex is the 73 amino
acid loop that connects I-2 residues 55 to 127. This loop contains Thr72, a GSK-3β
phosphorylation site, which is important for the formation of the reactivated ATPMgdependent PP1 phosphatase complex (i.e. phosphorylation of Thr72 reverses the
inhibition of PP1 by I-2 without release of I-2 from PP163). Phosphorylation of Thr72 on I2, when bound to PP1, has been shown to increase the fluorescence anisotropy of a
label covalently linked to a cysteine engineered into residue 129 of I-264. This suggests a
conformational change in PP1-bound I-2 upon phosphorylation at Thr72, corresponding
change in unbound I-2 was not seen. In agreement with this data, we observe contacts
between residues near 50-70 and 130 in PP1-bound I-2 (Figure 55C). These interactions
only exist in the PP1-bound form of I-2 and are likely required for the structural
rearrangement induced by phosphorylation.
The flexibility of this loop likely guides the intricate regulation of PP1 by I-2 for at
least three reasons. First, phosphorylation of Ser86 on I-2 has been shown to prime I-2
for GSK-3β phosphorylation at Thr72159. However, the thirteen residue linker between
these two residues is much larger than the typical three residue linker between
phosphorylation sites for GSK-3β recognition160. Interestingly, in conformers that adopt
the extended loop structure of our PP1:I-2 model, Ser86 and Thr72 are in close
proximity (average distance of 16.4 Å ± 5.8 Å). This is in contrast to the second set of
124
conformers, where the loop is folded back onto PP1. In this set, the average distance
between these two residues is 30.6 Å ± 8.5 Å. Thus, the positioning of the loop allows for
regulation of the phosphorylation state of Thr72 on I-2, and thus regulation of the
activation state of PP1. Second, Thr72 must be accessible for phosphorylation by GSK3β as well as for subsequent dephosphorylation by PP1 in order to reactivate the
complex63. In the inactivated PP1:I-2 complex, Tyr147 of I-2 is placed in the active site of
PP1 for inhibition62. Therefore, both Tyr147 and Thr72 must be able to reach the active
site of PP1, which is easily accomplished with the long, flexible loop of PP1-bound I-2.
Finally, the extended loop structure of PP1-bound I-2 creates an additional protein
interaction surface on I-2 that is extended away from PP1. This loop contains a transient
α-helical region from residues 97-105 (~35% populated). Given the tendency of transient
helices in IDPs to serve as binding sites, it seems quite likely that this α-helix is
important for I-2 function. Its accessible position would facilitate interactions of the
PP1:I2 holoenzyme with other proteins. Alternatively, this α-helix might facilitate the
binding of I-2 to other proteins, as I-2 has previously been shown to activate Aurora-A
kinase161 and regulate the activity of the prolyl isomerase Pin1162.
6.1.D.3. A structural continuum from folded to disordered states.
There is increasing evidence that proteins do not exist in either discrete random coil or
folded states – rather, they exist in a continuum of possible states ranging from folded to
disordered. The ensembles calculated in this study convincingly demonstrate this
continuum: disordered states can have varying amounts of secondary and tertiary
structure, ranging from highly disordered to the almost fully formed α-helix in I-2. In
addition, it is now understood that complete folding-upon-binding is only one possible
scenario for IDPs. An IDP can also retain substantial flexibility upon complex
formation163-167. The model of the PP1:I-2 complex presents a unique structural picture of
a protein that undergoes partial folding but remains largely disordered upon complex
125
formation. In fact, dynamic complexes such as PP1:I-2 are likely far more common than
might be expected based upon current experimental evidence, given that X-ray
crystallography
is
heavily
biased
against
disorder
and
flexibility.
Structurally
characterizing these highly dynamic complexes is not simple, but the presented
approach of combining various experimental measurements with previously known
structural information should be applicable to numerous systems. Thus, these combined
experimental and computational ensemble-modeling techniques will likely prove to be
extremely valuable for improving our understanding of the structural properties and the
biological functions of these dynamic complexes.
6.1.E. Experimental Procedures.
15
N-labeled and 2H,15N-labeled I-29-164 and PP1 were produced as described27,57. For the
purification of the PP1:I-2 complex, cells were lysed by high-pressure homogenization
(Avestin C-3 Emulsiflex). The soluble protein fraction was loaded onto a Ni-NTA column
(Invitrogen). Purified I-29-164 was incubated on the PP1-bound Ni-NTA beads, and the
complex was eluted with imidazole. The eluate was injected onto a Superdex 75 (26/60)
size exclusion column (GE Healthcare) equilibrated with 20 mM Tris (pH 7.5), 50 mM
NaCl, 0.5 mM TCEP. Fractions containing the pure protein complex, as confirmed by
SDS-PAGE, were pooled and concentrated to 60 µM for subsequent NMR
measurements. The 2D [1H,15N]-HSQC spectrum of unbound I-29-164 was measured at
298 K on a Bruker Avance II 500 MHz spectrometer; the 2D [1H,15N]-TROSY spectrum
of the PP1:I-2 complex was measured at 298 K on a Bruker Avance II 800 MHz
spectrometer. Both spectrometers were equipped with TCI HCN z-gradient cryoprobes.
The NMR spectra were processed with Topspin 1.3 (Bruker) and analyzed with the
CARA software package (www.nmr.ch).
Ensemble models of intrinsically disordered I-29-164, spinophilin417-494 and DARPP321-118 were calculated with ENSEMBLE using a nearly identical protocol to that
126
previously described for the drkN SH3 domain unfolded state
149
and the experimental
restraints given in Table 1. The model of the PP1:I-2 complex was calculated using
ENSEMBLE with modifications to the previous protocol. The PP1 structure was kept
identical to the crystallographic model and all refinement was performed on I-29-164. The
regions of I-2 observed in the crystal structure were restrained to be nearly identical to
the crystal structure (< 2 Å CαCαRMSD). Chemical shift and
15
N R2 restraints from
unbound I-2 were used for the disordered regions not observed in the bound complex.
All unbound state PRE restraints were used except those between pairs of residues both
observed in the bound crystal structure. The PRE restraints were weighted according to
the chemical shift changes that were observed upon binding PP1, whereby residues that
had large chemical shift changes had their PRE restraints significantly underweighted. In
addition, the maximum energy penalty from any single restraint was limited; this
prevented individual poorly fit restraints from dominating the ENSEMBLE calculations.
One of the final unbound I-29-164 ensembles was used as the starting ensemble and,
rather than using the simplest-ensemble approach, the number of structures was kept
constant at 19. The calculation proceeded until the chemical shift and
15
N R2 restraints
were entirely fit and then continued to fit the PRE restraints as optimal as possible. This
step-wise procedure was chosen as it was expected that the unbound-state PRE
restraints of I-2 would have a lower fit in the bound form of I-2.
The PP1:I-2 complex sample for SAXS data collection was produced within 24
hours of data acquisition and stored at 4°C. Prior to scattering experiments the sample
was filtered through a 0.02 µm filter (Whatman). Synchrotron X-ray scattering data were
collected at the National Synchrotron Light Source (NSLS) beamline X9. Small angle Xray scattering (SAXS) data were collected using a MarCCD 165 located at 3.4 m
distance from the sample. Wide angle X-ray scattering (WAXS) data were collected
simultaneously with SAXS data using a Photonic Science CCD located at 0.47 m from
127
the sample. 20 µl of sample was continuously pushed through a 1 mm diameter capillary
for either 180 or 360 s of total measurement time and exposed to a 400 x 200 microns
X-ray beam. Scattering data for the complex was collected at 0.5 mg/ml and 0.8 mg/ml
PP1:I-2 concentrations. Normalization for beam intensity, buffer subtraction and merging
of the data from both detectors were carried out using PRIMUS168. A Guinier
approximation, I(q) = I(0)exp(-q2Rg2/3), where a plot of I(q) and q2 is linear for q<1.3/Rg,
was performed on four independent scattering trials and averaged to determine the
radius of gyration115. HYDROPRO was used to calculate the Rg for the individual
ensemble models for comparison169. GNOM was used to determine the pair distribution
function [P(r)] and maximum particle dimension (Dmax)119. Sixteen independent ab initio
models were calculated using GASBOR121. The models were aligned and averaged
using DAMAVER122. The averaged ab initio envelope was aligned with nineteen
ensemble models for the PP1:I-2 complex using SUPCOMB154. The normalized spatial
discrepancy (NSD) values were averaged and used to represent the overall fit of the
ENSEMBLE model to the ab initio envelope. Theoretical scattering curves were
calculated for each of the ensemble structures using CRYSOL123. The scattering curves
and χ values were averaged to represent the overall fit of the theoretical scattering curve
to the experimental data.
6.1.F. Acknowledgments.
This material is based upon work supported by a Natural Sciences and Engineering
Research Council of Canada fellowship to J.A.M and a National Science Foundation
Graduate Research Fellowship to B.D. The authors thank Dr. Lin Yang for his support at
beamline X9. The project described was supported by Grant R01NS056128 from the
National Institute of Neurological Disorders and Stroke to W.P and funding from the
Canadian Institutes for Health Research to J.D.F.-K. 800 MHz NMR data were recorded
at Brandeis University (NIH S10-RR017269). Use of the National Synchrotron Light
128
Source, Brookhaven National Laboratory, was supported by the U.S. Department of
Energy, Office of Science, Office of Basic Energy Sciences, under Contract No. DEAC02-98CH10886. W.P. is the Manning Assistant Professor for Medical Science at
Brown University.
129
6.2. Molecular investigations of the structure and function of the protein
phosphatase 1:spinophilin:inhibitor-2 heterotrimeric complex.
This manuscript is currently in preparation. My contribution to this work is the SAXS
analysis of the PP1:spinophilin:I-2 complex.
Barbara Dancheck1,$, Michael J. Ragusa2,$ and Wolfgang Peti1,*
From Department of Molecular Pharmacology, Physiology and Biotechnology1, Brown
University, Providence, RI, 02912, USA and Department of Molecular Biology, Cell
Biology, and Biochemistry2, Brown University, Providence, RI, 02912, USA
$
These authors contributed equally to this work
Running head: Protein Phosphatase 1 Regulation
Address correspondence to: Wolfgang Peti, 70 Ship Street, GE-3, Providence, RI,
02912, Fax: (401) 863-6087; Email: wolfgang_peti@brown.edu
130
Regulation of the major ser/thr phosphatase Protein Phosphatase 1 (PP1) is controlled
by a diverse array of targeting and inhibitor proteins. Though many PP1 regulatory
proteins share at least one PP1 binding motif, the RVxF motif, it was recently discovered
that certain pairs of targeting and inhibitor proteins bind PP1 simultaneously to form PP1
heterotrimeric complexes. To date, structural information for these heterotrimeric
complexes, and, in turn, how they direct PP1 activity is entirely lacking. Using a
combination of NMR spectroscopy, biochemistry and small angle X-ray scattering
(SAXS), we show that major structural rearrangements in both spinophilin (targeting)
and Inhibitor-2 (I-2, inhibitor) are essential for the formation of the heterotrimeric
PP1:spinophilin:I-2 (PSI) complex. The RVxF motif of I-2 is released from PP1 during the
formation of PSI, making the less prevalent SILK motif of I-2 essential for complex
stability. The release of the I-2 RVxF motif allows for enhanced flexibility of both I-2 and
spinophilin in the heterotrimeric complex. In addition, we used inductively coupled
plasma atomic emission spectroscopy to show that PP1 contains two metals in both
heterodimeric complexes (PP1:spinophilin and PP1:I2) and PSI, demonstrating that PSI
retains the biochemical characteristics of the PP1:I2 holoenzyme. Finally, we combined
the NMR and biochemical data with SAXS and molecular dynamics simulations to
generate a structural model of the full heterotrimeric PSI complex. Collectively, these
data reveal the molecular events that enable PP1 heterotrimeric complexes to exploit
both the targeting and inhibitor features of the PP1-regulatory proteins to form multifunctional PP1 holoenzymes.
6.2.A. Introduction.
PP1 is a major ser/thr protein phosphatase with roles in numerous cellular processes
including cell cycle progression, protein synthesis, muscle contraction, transcription and
neuronal signaling25,34. Two metals in the active site of PP1 mediate a single step
131
dephosphorylation reaction. The metal ions enhance the nucleophilicity of a water
molecule bound at the active site, which performs nucleophilic substitution with the
phosphate moiety26,31. In vitro these metals are often manganese, as PP1 is expressed
in the presence of manganese. In vivo it is believed that these metal ions are iron and
zinc26.The surface of PP1 is characterized by extensive grooves, including three
substrate interaction sites: the acidic, the hydrophobic and the C-terminal substrate
binding grooves. Due to its large number of functions, PP1 is active towards a broad
range
of
substrates21.
Despite
this
inherent
lack
of
substrate
specificity,
dephosphorylation events by PP1 are under exceedingly tight control. This strict
regulation of PP1 activity is mediated by a large number of diverse regulatory proteins
(~200)21,33. There are two types of regulatory proteins: 1) targeting proteins that target
PP1 to specific locations in the cell and alter its substrate specificity, and 2) inhibitor
proteins71,170. Notably, these regulatory proteins have very little sequence similarity. In
addition, structures of PP1 in complex with the targeting proteins spinophilin/neurabin58
and MYPT143 and the inhibitor protein Inhibitor-2 (I-2)
60,62
demonstrate the diversity of
regulatory protein interactions with PP1. From these structures it has become evident
that targeting proteins regulate PP1 by differentially altering its ability to bind substrates.
Spinophilin blocks the C-terminal groove58, while MYPT-1 appears to extend the acidic
groove of PP143. Recently, it was shown that specific pairs of targeting and inhibitor
proteins can bind PP1 simultaneously, adding an additional layer of complexity in PP1
regulation65-67. Despite the recent advances in structural information for heterodimeric
complexes of PP1, structural information for heterotrimeric PP1 complexes is entirely
missing.
One heterotrimeric PP1 complex of interest is PP1:spinophilin:I-2 (PSI).
Spinophilin is a multi-domain scaffolding protein which targets PP1 to dendritic spines
132
and is important for the regulation of excitatory synaptic transmission and synaptic
plasticity42,47,171. I-2 is a ubiquitous, single domain inhibitor of PP1 found in diverse
tissues such as brain, skeletal muscle, and sperm172. As cells progress to the S phase of
the cell cycle173, I-2 is translocated into the nucleus and its expression peaks during Sphase and mitosis174, indicating a role for I-2 in cell cycle regulation. In vitro, I-2 forms an
inactivate complex with PP1 that can be reactivated by phosphorylation of I-2 at Thr72
by glycogen synthase kinase-3 or cyclin dependent kinase 2. Recently, spinophilin (or its
neuronal isoform neurabin) and I-2 were shown to bind PP1 simultaneously to form a
heterotrimeric complex (PSI). Furthermore, I-2 and neurabin were shown to co-localize
in actin-rich adherens junctions and dendritic spines
65
, suggesting a role for the
heterotrimeric PSI complex in cytoskeletal rearrangement and/or neuronal signaling.
Once targeted to the dendritic spine and cytoskeleton, the PSI complex is poised for
immediate activation from signaling pathways which may lead to the phosphorylation of
I-2 and reactivation of the phosphatase.
Three additional heterotrimeric PP1 complexes have been identified, including
the PP1:GADD34:I-1, the PP1:Sds22:I-3 and the PP1:MYPT1:CPI-17 complexes 66,67,175.
However, GADD34 and I-1 also interact in the absence of PP1. Moreover, CPI-17 does
not contain an RVxF motif, rather it is a highly specific inhibitor of the PP1:MYPT1
holoenzyme. Lastly, Sds22 also does not contain an RVxF motif. Only in the PSI
complex do both PP1-regulators, spinophilin and I-2, contain a RVxF motif and require
PP1 for the formation of the heterotrimeric complex65.
The PP1:spinophilin (PS) and PP1:I-2 (PI) complex structures have been solved
by X-ray crystallography and NMR spectroscopy-based ensemble calculations58,62.
These structures provide insight into the mechanisms each protein uses to regulate PP1.
While both proteins form unique interactions with PP1, they also share common binding
133
sites. First, both spinophilin and I-2 bind PP1 via a common docking motif termed the
RVxF motif58,62. This short consensus motif, which is shared by most PP1 regulatory
proteins, binds PP1 in a hydrophobic pocket ~20 Å from the active site and is the reason
it was originally thought that only one regulatory protein could bind PP1 at a time35.
Second, helices from I-2 and spinophilin bind in a similar location on PP1 near the
hydrophobic groove, with residues Asp163 of I-2 and Glu482 and Glu486 of spinophilin
making contacts with Arg132 of PP1. Therefore, a structural rearrangement of I-2,
spinophilin, or both proteins is necessary for the formation of the heterotrimeric PSI
complex. We have used NMR spectroscopy, biochemistry and small angle X-ray
scattering (SAXS) to gain insight into the formation and regulation of PSI. In combination
with the previously reported structures of the heterodimeric complexes, our data
provides the first structural insights into a heterotrimeric PP1 complex and provide a
model for how these complexes direct PP1 activity in vivo.
6.2.B. Experimental Procedures.
6.2.B.1. Protein Expression and Purification.
Unlabeled PP1α7-330, unlabeled spinophilin417-583, and both unlabeled and 2H,15N-labeled
I-29-164 C85S were produced as previously described27,57,58,60. The C85S mutation of I-2
was used because wild-type I-2 dimerizes by forming an artificial disulfide bond (dimer
formation can be reversed by reducing agents) and this dimerization is not observed with
I-2C85S. The expression of 2H,15N-labeled spinophilin417-583 was carried out by growing
freshly transformed BL21-Codon-Plus (DE3)-RIL cells (Stratagene) in M9 minimal
medium containing 1 g/L
15
NH4Cl as the sole nitrogen source, prepared in D2O. Protein
expression was induced with 1 mM IPTG. 2H,15N-labeled spinophilin417-583 was purified
following the same protocol as unlabeled spinophilin417-583. The PS and PI complexes
were produced as previously described58,60 with the following minor modifications. After
134
elution from the Ni-NTA column, the complexes were purified using a Superdex 200
26/60 size exclusion column (GE Healthcare) equilibrated with PP1 complex buffer (20
mM Tris pH 7.5, 50 mM NaCl, 0.5 mM TCEP). Tobacco Etch Virus protease was added
to cleave the His6-tag from PP1α7-330. After digestion was complete, the complexes were
loaded onto Ni-NTA beads, from which they eluted in the flow through. For the final
purification step, the proteins were purified using a Superdex 75 26/60 size exclusion
column (GE Healthcare) equilibrated with PP1 complex buffer.
For the production of the PSI complex, PP1α7-330 was bound to Ni-NTA beads via
its His6-tag and incubated with purified, untagged spinophilin417-583. Following step elution
with buffer containing 250 mM imidazole, the eluate was loaded directly onto a Superdex
200 26/60 size exclusion column equilibrated with PP1 complex buffer. Fractions
containing the PS complex were confirmed by SDS-PAGE gel electrophoresis. The
complex was pooled and TEV protease was added to remove the His6-tag from PP1
overnight. After digestion was complete, the complex was incubated with a 1:2 molar
ratio of >98% pure I-29-164, and the complex was loaded onto a Superdex 75 26/60 size
exclusion column equilibrated with PP1 complex buffer. The heterotrimeric PSI complex
eluted from this column at ~127 mL. Fractions containing the PSI complex were
confirmed by SDS-PAGE gel electrophoresis.
The QuikChange site-directed mutagenesis kit (Stratagene) was used to create
the C85S, I13A/L14A/K15A, K42A/Q44G, W46A and R157A/D163A mutants of I-2
following the manufacturer’s protocol. The presence of the mutations was confirmed by
sequencing. All mutants were prepared following the same protocol as I-29-164 C85S. CD
spectroscopy was used to confirm that mutagenesis of I-2 did not alter the overall
secondary structure of the protein. CD spectra and estimation of percent α-helical
content of the I-2 mutants were performed as previously reported57.
135
6.2.B.2. NMR Spectroscopy.
Samples for NMR spectroscopy were prepared in 20 mM Tris, pH 7.5, 50 mM NaCl, 0.5
mM TCEP. The 2D [1H,15N]-HSQC spectra of unbound I-2 and unbound spinophilin were
measured at 298 K on a Bruker Avance II 500 MHz spectrometer equipped with a TCI
HCN z-gradient cryoprobe. The 2D [1H,15N]-TROSY spectra of the PS, PI and PSI
complexes were measured at 298 K on a Bruker Avance II 800 MHz spectrometer
equipped with a TCI HCN z-gradient cryoprobe. The NMR spectra were processed with
Topspin 1.3 (Bruker) and analyzed with the CARA software package (www.nmr.ch).
Δδ
Chemical shift differences (Δδ) were calculated as Δδ(ppm) = �(ΔδH )2 + ( N)2.
10
6.2.B.3. Inductively Coupled Plasma Atomic Emission Spectroscopy.
The manganese content of the PS, PI and PSI complexes was measured using a Jobin
Yvon JY2000 Inductively Coupled Plasma Atomic Emission Spectrometer. Complexes
were prepared as described, concentrated and then diluted in 2% nitric acid. Manganese
standards were prepared in 2% nitric acid. Three independent measurements were
performed in triplicate for each sample.
6.2.B.4. Capture Assay.
For the capture assay using PP1, cells expressing his6-tagged PP1α7-330 were lysed by
high pressure homogenization (Avestin C-3 Emulsiflex). His6-tagged PP1α7-330 was
bound to Ni-NTA beads (Qiagen, 6 parallel 1-mL columns with fresh beads). For the
capture assay using the PS complex, PS was prepared as described above up until the
TEV cleavage step. The PS complex was bound to Ni-NTA beads via the His6-tag on
PP1α7-330 (Qiagen, 6 parallel 1-mL columns with fresh beads). The beads were
extensively washed and subsequently incubated with buffer, I-29-164 C85S or one of the
four I-2 mutants for one hour at 4 oC. After extensive washing to remove any unbound I-
136
2, the PI or PSI complexes were eluted with elution buffer containing 1 M imidazole.
Bound complexes were analyzed by SDS-PAGE using Coomassie blue staining, and the
density of the protein bands was quantified using the Quantity One program (Bio-Rad
Laboratories).
6.2.B.5. Small angle X-ray scattering.
Small angle X-ray scattering (SAXS) data was recorded on the PSI complex. Samples
were prepared in 20 mM Tris pH 7.5, 50 mM NaCl, 0.5 mM TCEP within 24 hours of
data acquisition and were stored at 4 °C. Prior to scattering experiments the samples
were concentrated and filtered through a 0.02 µm filter (Whatman). Scattering data were
collected at the National Synchrotron Light Source (NSLS) beamline X9 using a
MarCCD 165 located at 3.4 m distance from the sample for small angle X-ray data. Wide
angle X-ray scattering (WAXS) data were collected simultaneously with SAXS data
using a Photonic Science CCD located at 0.47 m from the sample. 20 µl of sample was
continuously flowed through a 1 mm diameter capillary and exposed to a 400 x 200 µm
X-ray beam for 180 seconds of total measurement time. Scattering data for the PSI
complex was collected at 0.8 and 1.2 mg/ml. PRIMUS was used for normalization for
changes in the beam intensity, buffer subtraction and merging of the small angle and
wide angle scattering data168. The radius of gyration (Rg) was calculated using a Guinier
approximation, I(q) = I(0)exp(-q2Rg2/3), where a plot of I(q) and q2 is linear for q<1.3/Rg115.
Four independent scattering trials were averaged. The linearity of the Guinier plots was
evaluated to confirm that no aggregation was present in any of the samples.
HYDROPRO was used to calculate the Rg for the PSI model169. GNOM was used to
determine the pair distribution function [P(r)] and maximum particle dimension (Dmax)119.
Sixteen independent ab initio models were calculated using GASBOR121,176. The models
were aligned and averaged using DAMAVER122. The averaged ab initio envelope was
137
aligned with the PSI model using SUPCOMB154. BILBOMD was used to investigate the
flexibility of the PSI complex125. Molecular dynamics (MD) simulations as part of
BILBOMD were used to generate 12,000 structures with a Rg range of 20 – 50 Å (200
structures/1 Å; 2 total calculations). Theoretical scattering curves were calculated for
each structure using FOXS and compared to the experimental data124. The single best fit
structure is defined as the structure with the lowest discrepancy (χ2) between the
theoretical and experimental data. A minimal ensemble (MES) model was generated as
previously described; this is the minimal number of either 2, 3, 4, or 5 structures that
provided the smallest discrepancy between theoretical and experimental scattering
data125. This ensemble was selected as the best model for the protein complex in
solution.
6.2.C. Results.
6.2.C.1. Spinophilin417-583, I-29-164C85S and PP1α7-330 form a stable heterotrimeric
complex in vitro.
Size exclusion chromatography (SEC) was used to determine whether spinophilin417-583,
I-29-164C85S and PP1α7-330 form a stable heterotrimeric complex in vitro. Briefly, the PS
complex was incubated with a 1:2 molar ratio of I-29-164, and the formed complex was
injected onto a Superdex 75 26/60 size exclusion column. The heterotrimeric PSI
complex eluted at ~127 mL, prior to excess unbound I-2 (Figure 57). For comparison,
both PS and PI elute at ~155 mL, demonstrating a significant increase in molecular size
upon formation of the heterotrimeric PSI complex. SDS-PAGE gel electrophoresis
confirmed the presence of all three proteins in the appropriate stoichiometric ratios
(Figure 57). Taken together, this data demonstrated the stable formation of the
heterotrimeric PSI complex in vitro.
138
6.2.C.2. Two manganese ions
are present in the active site of
all three PP1 complexes
Two manganese ions are present
in the active site of PP1 in the
crystal
structure
of
the
PS
complex58 and the PP1:MYPT1
complex43. In contrast, only one
manganese ion was observed in
Figure 57. PP1, spinophilin and I-2 form a stable,
hetermotrimeric complex.
SEC 75 26/60 chromatograms are shown for PS
(dashed line) PI (dashed-dotted line) and PSI (solid
line). In the PSI chromatogram, distinct peaks are
observed for the PSI complex at 127 mL and for
excess, unbound I-2 at 149 mL. The PSI
chromatogram is overlaid with that of PS and PI alone,
illustrating the shift in elution position upon complex
formation. The SDS-PAGE gel to the right of the
chromatogram shows that I-2 and spinophilin are
bound to PP1 in a 1:1 stoichiometric ratio and that the
complex is highly pure.
the crystal structure of the PI
complex62, suggesting that I-2
may expel one of the metals from
the active site of PP1 as a
mechanism
of
inhibition.
To
determine the number of Mn2+
ions in the heterodimeric PS, the heterodimeric PI and the heterotrimeric PSI complex in
solution, we used inductively coupled plasma atomic emission spectroscopy (ICP-AES).
The results are summarized in Table 9. As expected, two moles of manganese were
present per mole of PS. In addition, two moles of manganese were also present per
mole of both PI and PSI. Thus, these data suggest that PSI retains the biochemical
characteristics of the PI holoenzyme.
6.2.C.3. The I-2 RVxF motif is not necessary for stable PP1 binding.
An overlay of the crystal structures of the PS and PI complexes reveals two regions of
steric clashing between I-2 and spinophilin (Figure 83). Thus, to form the heterotrimeric
PSI complex either spinophilin, I-2, or both must undergo a conformational
rearrangement. In order, to understand the structural rearrangements that occur during
139
the formation of PSI, we undertook
Complex
mol Mn2+ / mol
complex
Standard
Deviation
NMR spectroscopy studies of PS
PS
2.3
0.3
(55.9 kDa), PI (55.2 kDa) and PSI
(73.9 kDa). For these studies, only
PI
1.9
0.1
PSI
2.4
0.3
Table 9. Manganese content of PS, PI and PSI
complexes as measured by ICP-AES.
Values are the average of three independent
measurements performed in triplicate.
a single protein (either I-2 or
spinophilin) was 2H,15N isotopically
labeled and, thus, NMR active; the
other proteins of the heterotrimeric
complex remained unlabeled, thus
invisible in the NMR spectrum. This makes the analysis of the NMR spectra much more
feasible for these high-molecular weight complexes (>55 kDa).
We previously reported that regions of I-2 with significant chemical shift
differences between unbound I-2 and PP1-bound I-2 were limited to the three regions
that directly interact with PP1, specifically the RVxF motif, the SILK motif and the Cterminal helix60. Figure 58A depicts an overlay of the 2D [1H,15N]-HSQC spectrum of
unbound
15
N labeled I-2 and the 2D [1H,15N]-TROSY spectrum of 2H,15N labeled I-2 in
PSI while Figure 58B depicts an overlay of the 2D [1H,15N]-TROSY spectra of 2H,15N
labeled I-2 in the PI and PSI complexes. The spectra are highly similar, indicating that no
major structural rearrangement of I-2 between the PI and PSI complexes occurs. The
differences in the spectra are quantified using chemical shift difference plots, which are
shown in Figure 58C (unbound I-2 versus I-2 in PSI) and Figure 58D (I-2 in PI versus I-2
in PSI). The regions of I-2 with significant chemical shift differences between unbound I2 and PSI are the SILK motif, which binds the backside of PP1, and the C-terminal helix,
which lies over the PP1 active site. In contrast, there are essentially no differences in the
RVxF motif between unbound I-2 and PSI. This shows that the RVxF motif is not bound
140
Figure 58. The RVxF motif of I-2 is released from PP1 in the formation of PSI, increasing
the importance of the SILK motif.
1
15
1
15
A) Overlay of the [ H, N]-HSQC spectrum of unbound I-2 (pink) and the [ H, N]-TROSY
1
15
spectrum of PSI-bound I-2 (black). B) Overlay of the [ H, N]-TROSY spectra of PSI-bound I-2
(black) and PI-bound I-2 (red). C) Chemical shift difference plot of unbound I-2 versus PSIbound I-2 indicates that residues in the RVxF motif of I-2 are in similar environments in
unbound I-2 and in the PSI complex. Only residues in the SILK motif and the C-terminal helix
show significant differences in chemical shifts. D) Chemical shift difference plot of PSI-bound
I-2 versus PI-bound I-2 indicates that residues in the SILK motif and the C-terminal helix of I-2
are in similar environments in PSI and PI complexes. Only residues in the RVxF motif show
significant changes in chemical shifts between these two complexes. This indicates that the I2 RVxF motif releases from PP1 during the formation of PSI. E) Pull down assays (error bars
are based on three independent measurements performed in triplicate) of I-2 mutants using
PP1-bound Ni-NTA resin (black bars) or PS-bound Ni-NTA resin (grey bars). The SILK mutant
is I13A/L14A/K15A, RVxF mutant 1 is W46A, RVxF mutant 2 is K42A/Q44G, and the helix
mutant is R157A/D163A. Binding to PP1 alone is affected equally by mutations in all three
regions of I-2 that interact with PP1. However, mutation of the SILK motif has a dramatic
affect on the binding of I-2 to PSI, indicating its increased importance after release of the I-2
RVxF motif from PP1.
141
to PP1 in the PSI complex. This is further supported by that observation that the RVxF
motif is the only region of I-2 with significant chemical shift differences when comparing
the chemical shifts of I-2 in PI and PSI complexes. Collectively, these data show that the
RVxF motif of I-2 is in a similar environment in unbound I-2 and PSI, i.e. not bound to
PP1, while the SILK motif and C-terminal helix are in similar environments in the PI and
PSI complexes, i.e. bound to PP1.
6.2.C.4. Pull downs of I-2 mutants identify the SILK motif as critical for PSI
formation.
To compliment the NMR analysis of each of the three PP1-binding motifs of I-2 in the
formation of the PI versus PSI complexes, numerous I-2 point mutants were created and
used in pull down assays to test their ability to form the PSI complex. I-2 C85S was used
as the wild-type control62. The SILK mutant, I13A/L14A/K15A, includes residues forming
both electrostatic and hydrophobic interactions with PP1. The RVxF (residues KSQKW 4246
in I-2) mutant 1, W46A, was created because W46 is buried in a hydrophobic pocket
in the RVxF motif binding groove and plays an essential role in forming the PI complex.
However, we have previously shown that W46 is also important in the stabilization of
transient tertiary structure in unbound I-257,60. In order to confirm that our results were
not due to a destruction of important tertiary structure in unbound I-2, an additional RVxF
mutant 2, K42A/Q44G, was produced; these residues also make important contacts with
PP1. Finally, the C-terminal helix mutant, R157A/D163A, was created because these
residues form important electrostatic interactions with PP1 in the PI complex. CD
spectroscopy confirmed that mutagenesis did not alter the secondary structure present
in unbound I-2 (Figure 84).
As shown in Figure 58E, all three PP1 binding motifs of I-2 are equally important
in the formation of the PI complex, as all mutations affected the binding of I-2 to PP1
142
alone about equally (~20-40% reduction in binding; black bars). In contrast, the SILK
motif mutant, in agreement with our NMR data, resulted in significantly less binding of I-2
to the PS complex (~85% reduction in binding) when compared with all other I-2 mutants
(grey bars). Therefore, upon release of the I-2 RVxF motif from PP1, the SILK motif
becomes essential for I-2 to bind PS and form the heterotrimeric PSI complex.
6.2.C.5. NMR spectroscopy of PS and PSI.
To determine if a structural rearrangement of spinophilin plays a role in the formation of
PSI, we undertook NMR spectroscopy studies of spinophilin in the PS and PSI
complexes. The overlay of the 2D [1H,15N]-HSQC/TROSY spectra of unbound
15
N
labeled spinophilin and 2H,15N labeled spinophilin in the PS complex is shown in Figure
59A, while the overlay of the 2D [1H,15N]-TROSY spectra of 2H,15N labeled spinophilin in
the PS and PSI complexes is shown in Figure 59B. Unbound spinophilin shows limited
preference for transient secondary and tertiary structure throughout its PP1 binding
domain60. However, upon binding to PP1, secondary structure elements are formed in
the PP1 binding domain, following restriction to a single conformation58. As is expected
for such a significant folding-upon-binding transition, the 2D [1H,15N]-HSQC/TROSY
spectra of unbound spinophilin and spinophilin in the PS complex are significantly
different. Therefore, it was not possible to transfer the assignment of unbound
spinophilin onto PP1-bound spinophilin. Due to the low stability of the complex (~20 h,
120 µM) it was impossible to record the necessary 3D NMR spectra for the sequencespecific backbone assignment of spinophilin in the PS complex. However, ~15 peaks
differ in the 2D [1H,15N]-TROSY spectra of spinophilin in the PS and PSI complexes,
indicating that the local environment of these spinophilin residues are different in these
two complexes, due to I-2 binding.
143
Figure 59. Structural rearrangement is seen in spinophilin upon formation of PSI.
1
15
1
15
A) Overlay of the [ H, N]-HSQC spectrum of unbound spinophilin (pink) and the [ H, N]TROSY spectrum of PPI-bound spinophilin (black). Major changes in spinophilin between the
two spectra are due to the folding-upon-binding transition spinophilin undergoes upon binding
1
15
to PP1. B) Overlay of the [ H, N]-TROSY spectra of PPI-bound spinophilin (black) and PSIbound spinophilin (red). Limited differences between the two spectra indicate that spinophilin
does not undergo a large conformational rearrangement between the PS and PSI complexes,
however ~15 residues exhibit significant chemical shift changes.
6.2.C.6. SAXS analysis of the heterotrimeric PSI complex.
To gain further structural insights into the heterotrimeric PSI complex, we performed
small angle X-ray scattering (SAXS), which reports on the overall size and shape of a
protein in solution (Figure 85, Table 10). SAXS data was recorded at multiple
concentrations and Guinier analysis of the PSI complex reported a radius of gyration of
37.1 ± 1.0 Å (Figure 60A). For the initial analysis of the SAXS data we created a PSI
model (thereafter referred to as PSIm) based on a superposition of PS X-ray crystal
structure and the PI ensemble structure60. Interestingly, the calculated Rg of the PSIm
structure is only 33.7 Å and thus significantly smaller then the experimental value based
144
Figure 60. SAXS analysis of the PSI complex.
A) The Guinier region of the scattering curve for data recorded at 0.8 mg/ml and 1.2 mg/ml. B)
Ab-initio envelope for PSI. Overlay of the ab-initio envelope (gray surface representation) with
the PS crystal structure and one structure from the PI ensemble model. All structures are
shown as cartoon images with spinophilin in red, PP1 in black and I-2 in cyan. BILBOMD
2
analysis of the PSI complex was carried out. C) Graph comparing the Rg to the χ for the
2
12,000 structures generated during MD simulations (grey squares). χ is the discrepancy
between the theoretical scattering and the experimental data. The data point corresponding to
the single best fit structure is shown blue and the data points for the five structures of the MES
model shown in red. D) Comparison of the theoretical scattering data (red line) for the best fit
MES model to the experimental scattering data (black squares). E) The five structures of the
MES model are shown with PP1 in gray as a surface representation, and spinophilin and I-2
as cartoon representations in red and cyan, respectively. The relative percent populations are
listed below each structure along with the Rg for each structure.
145
on Guinier approximation. This shows that the PSIm superposition model does not
provide a good representation of the PSI complex and the formation of PSI results in a
more extended structure.
Next, GASBOR was used to
PSI
Guinier approximation
Rg (Å)
P(r) function calculation
q-range (Å-1)
Rg (Å)
Dmax (Å)
Structure modeling
χ
NSDa
NSDm
calculate an ab initio molecular
37.1 ± 1.0
envelope
0.015-0.300
36.5
125
1.30 ± 0.03
1.37 ± 0.04
2.02
Table 10. Statistic for SAXS analysis of the PSI
complex.
a
NSD is the discrepancy between the ab initio
m
models. NSD is the discrepancy between the
envelope and the PS and PI overlay model.
of
the
PSI
60B)121,176.
(Figure
complex
Sixteen
independent ab initio models were
created with a maximal length of
125
Å
analysis
as
of
determined
the
by
the
pair-distance
distribution function (Figure 85C).
The ab initio models were then
averaged. However, the normalized spatial discrepancy (NSDa) between the models was
a modest 1.37 ± 0.04154, showing that the envelopes do not converge to a single model
and instead suggests that the PSI complex is best represented as a conformational
ensemble. Interestingly, the ab-initio envelope contains additional density around both
the RVxF motif and PDZ domain of spinophilin, further supporting our NMR and
biochemical data that show that both spinophilin and I-2 exhibit increased flexibility due
to the structural rearrangements the accompany PSI complex formation. This was
further confirmed by the alignment of the PSI envelope with PSIm, which resulted in a
poor NSDm of 2.02. All these results confirmed that the PSI complex cannot be
accurately described by a single structure.
6.2.C.7. Ensemble model for PSI.
146
Thus to further refine the conformation of the PSI complex in solution, we used the
program BILBOMD125, which samples conformational space using molecular dynamics
(MD) simulations and subsequently selects for the models that have the best agreement
between the theoretical and experimental scattering data. The PSIm superposition
model and information from the NMR spectroscopy and biochemistry experiments were
used to define regions of rigidity and flexibility. The NMR analysis of I-2 in the PSI
complex shows that the chemical shifts of the I-2 C-terminal helix in the PI and PSI
complexes are identical, indicating that the I-2 helix interactions observed in the PI
crystal structure are also present in the PSI complex. However, this implies that the
position of the spinophilin helix must change upon heterotrimeric PSI complex formation.
Thus, PP1 (residues 7-300; electron density in the PP1:spinophilin structure), spinophilin
residues 424-476 (electron density in the PP1:spinophilin structure) and the I-2 SILK
motif (residues 11-16; electron density in the PP1:I-2 structure) and helix of I-2 (residues
130-164; electron density in the PP1:I-2 structure) were fixed as rigid body 1, which was
not allowed to move. The spinophilin PDZ domain (residues 494-583; electron density in
the PP1:spinophilin structure) was defined as rigid body 2, which was allowed to move
relative to rigid body 1. Based on our NMR data, spinophilin residues 477-493 (helix)
and I-2 residues 9-10 (N-terminal to the SILK motif) and I-2 residues 17-129 were
defined as flexible.
As expected, the generated PSI structures sample a broad conformational space
with an Rg between 30.8 to 48.7 Å. Theoretical scattering data from a single best fit
structure fits the data with a χ2 = 1.47. An minimal ensemble search (MES) model of five
structures provides a better fit to the experimental data with a χ2 = 1.08 (Figure 60C,D).
Individually, each structure of the ensemble has a significantly worse fit to the data,
demonstrating that the structure of the PSI complex can only be described as an
147
ensemble of structures. The five ensemble structures that form the PSI MES include an
Rg range from 30.8 to 48.7 Å and are represented from 13 to 32% (Figure 60E).
6.2.D. Discussion.
PP1 is a major ser/thr protein phosphatase, whose specificity is defined through its direct
interaction with >200 regulatory proteins. Recently, it was also shown that pairs of PP1
targeting and inhibitor proteins can interact with PP1 simultaneously65-67. However, to
date, no detailed biochemical and structural information for any heterotrimeric
PP1:targeting protein:inhibitor protein complex has been reported. As most PP1
regulators share interaction sites on PP1, significant conformational rearrangements
must accompany the formation of these heterotrimeric complexes. For example, both the
PP1 targeting protein spinophilin and the inhibitor I-2 interact with PP1 via their RVxF
motifs. In addition, helices within the spinophilin and I-2 PP1-binding domains interact
with a common region on PP1, close to the PP1 hydrophobic substrate binding groove.
Yet, PP1, I-2 and spinophilin still form a heterotrimeric complex. Therefore, we used
NMR spectroscopy and SAXS measurements, supported by biochemistry, to determine
the structural rearrangements that occur upon the formation of the heterotrimeric PSI
complex.
To further our understanding of the mode of action of PP1 and, in parallel, the
mode of inhibition by I-2 in vivo, we determined the number of manganese ions that are
present in the active site of PP1 in the PS, PI and PSI complexes using inductive
coupled plasma atomic emissions spectroscopy. Our results show that one PP1
molecule binds to two manganese ions in solution in all three complexes. This indicates
that I-2 uses the same mode of PP1 inhibition in both the PI and PSI complexes. While
two metals were seen in the active site of PP1 in the crystal structure of the PS
complex58, only one was present in that of the PI complex when co-expressed with I-2 in
148
the presence of manganese62. However, our data indicate that if I-2 binds to PP1 after it
is folded in vitro, it is not able to expel a metal from the active site.
Two aspects make the structural analysis of PP1 complexes challenging: 1) their
large size (>50 kDa) makes NMR spectroscopy demanding and 2) PP1 complexes are
highly flexible. To overcome the NMR size limitation, we established a robust PSI
complex formation protocol, which enables us to form stable complexes with individually
produced proteins/protein domains. This was essential for the current study, as this
enables any single protein of the three that form the heterotrimeric PSI complex to be
individually labeled, allowing us to overcome any chemical shift overlap uncertainties.
Because of the high flexibility of the PSI complex, X-ray crystallography is challenging at
best and often impossible (1000’s of crystallization trials of the PSI complex were
unsuccessful). Thus only techniques that are able to tolerate flexibility, such as NMR
spectroscopy and SAXS, can be used to gain molecular structural information. Lastly,
due to the increased flexibility, these structures cannot be described as single
conformers, rather they should be described as ensembles of structures.
Because of its vital importance for PP1 binding, numerous bioinformatics and
biochemical analysis of the amino acid composition of the RVxF motif have been
described21,33,35,177. For instance, a biochemical RVxF motif analysis identified that
mGluR7b with K, H and F in RVxF motif positions two, four and five, respectively, pulled
down more PP1 than S, K and W in these same positions177. This predicts that the RVxF
motif of spinophilin (RKIHF447-451) interacts more tightly with PP1 than that of I-2
(KSQKW 42-46). In line with this, our NMR spectroscopy data show that the RVxF motif of
I-2 is released from PP1 upon formation of PSI, while the RVxF motif of spinophilin
remains bound to PP1. The same will likely hold true for other heterotrimeric PP1
complexes, indicating that it will be possible to predict which regulatory protein releases
149
from the RVxF site based on the RVxF sequences. Release of the RVxF motif of I-2
from PP1 increases the importance of the SILK motif in the formation of PSI. I-2 is one of
the most ancient PP1 regulatory proteins, and the SILK motif is highly conserved among
opisthokonts and amoebozoa143. Therefore, it is predicted that SILK motifs are present in
other PP1 regulatory proteins. So far, seven additional PP1 regulators have been
identified with a SILK or SILK-like motif21,178. Critically, these studies show the SILK motif
is an essential motif for anchoring regulatory proteins to PP1, especially when other
binding motifs, such as the RVxF motif, are released or displaced. Additionally, our
mutagenesis data have also revealed the importance of residues forming electrostatic
interactions between the helix of I-2 and PP1 for binding. In contrast, mutations of similar
residues on spinophilin had no effect on the binding to PP158. This clearly lends further
support to our model in which I-2 displaces the helix of spinophilin during formation of
the PSI complex.
Based on the NMR spectra as well as biochemical and SAXS data, we have
generated a MES model for the PSI complex. The MES model contains five structures
each represented at similar populations but with varying Rgs. The release of the RVxF
motif of I-2 and helix of spinophilin create a heterotrimeric complex that contains
enhanced flexibility compared to the already dynamic heterodimeric complexes. While it
is currently unclear what the role of this enhanced flexibility is, it could alter the function
of both spinophilin and I-2. The enhanced flexibility of the heterotrimeric complex creates
a more extended and flexible PDZ domain of spinophilin than in PS. This could allow the
PDZ domain to bind its targets more efficiently, or create a larger substrate capture
radius, in a model similar to the proposed “fly casting” mechanism179. Similarly, the more
extended nature of I-2 could also create new protein interaction sites which were
previously buried in the PI complex. Thus, to summarize, by forming the heterotrimeric
150
PSI complex the best of both ‘PP1 worlds' is optimally combined: 1) PP1 targeting is
mediated by the multi-domain protein spinophilin and 2) PP1 inhibition is mediated by I2, although I-2 most likely does not activate PP1 until 'triggered' by outside inputs such
as phosphorylation of Thr-72 on I-2.
Chapter 6. Discussion and outlook.
151
152
About 30% of all intracellular eukaryotic proteins are regulated by serine/threonine
phosphorylation69. Due to the large number of phosphorylation events and the complex
cellular process that they regulate, it is essential that these events are under tight
control. It was previously believed that Ser/Thr phosphatases were simple housekeeping
enzymes with little regulation and specificity, while Ser/Thr kinases were tightly regulated
enzymes180. In recent years, it has become clear that this assumption is incorrect for
Ser/Thr phosphatases. Ser/Thr phosphatases undergo specific regulation through
targeting, activation and inhibition. These mechanisms allows phosphatases to play an
active role in the regulation of many cellular events.
Ser/Thr protein phosphatase 1 (PP1) interacts with ~200 regulatory proteins to
dephosphorylate hundreds of substrates21. The majority of these regulatory proteins are
targeting proteins, which alter the substrate specificity of PP1 and localize PP1 to its
point of action. Targeting proteins localize PP1 to many cellular compartments including
the plasma membrane, the nucleus, sarcoplasmic reticulum, and dendritc spines34.
However, significantly less work has been done to characterize the role of these proteins
in altering the substrate specificity and activity of PP1. Currently there are two major
classes of targeting proteins that alter the substrate specificity of PP1. The first class,
which includes spinophilin, bind to PP1 and inhibit the activity of PP1 for phosphorylase
a, while the second class, which includes GM do not42,45. The mechanism for both of
these modifications of the activity of PP1 is unclear. To this end, structural biology and
biochemistry have been used to investigate the regulatory mechanisms for PP1 by the
targeting proteins spinophilin and GM.
6.1. The PP1:spinophilin complex.
Spinophilin is part of a larger class of PP1 targeting proteins that inhibit the acitivty of
PP1 against the most commonly used substrate, phosphorylase a42. However, no
153
mechanism for this inhibition of phosphorylase a had been previously described. Here
we have determined the structure of the PP1:spinophilin holoenzyme and measured
changes in the activity of PP1 for various substrates. The structural and biochemical
data provides, for the first time, a mechanism for altering the substrate specificity of PP1.
Spinophilin, which is unstructured in the absence of PP1 and folds upon binding to PP1,
interacts with one of three PP1 substrate binding grooves, the C-terminal substrate
binding groove. As we and others have shown, phosphorylase a requires binding to the
C-terminal substrate binding groove of PP1. Therefore, spinophilin’s interaction with
residues in the C-terminal substrate binding groove causes an inhibition in the activity of
PP1 for phosphorylase a, as it prohibits phosphorylase a binding. By blocking only a
single substrate binding site, PP1 is still able to dephosphorylate substrates that interact
with the remaining two PP1 substrate binding grooves, including the physiological target
of the PP1:spinophilin holoenzyme, GluR1. This demonstrates that spinophilin does not
inhibit the activity of PP1 for all of its subtrates as had been previously postulated42.
Additionally, our data demonstrates that the PP1:spinophilin holoenzyme can
dephosphorylate GluR1 in vitro, which suggests that PP1 does not need to release from
spinophilin to dephosphorylate GluR1 in vivo.
We can expand our understanding of PP1 regulation by targeting proteins via
comparison of the PP1:spinophilin structure with the only other PP1:targeting protein
structure, the PP1:MYPT1 structure43. MYPT1 and spinophilin bind to PP1 through
completely different interaction sites, except for the conserved RVxF motif, which is
shared by most PP1 interactors. The unique modes of binding for each protein, suggest
that they modify the substrate specificity of PP1 using two different mechanisms.
Spinophilin binds and decreases the accessible surface area for substrate binding on
PP1 (Figure 61 A). MYPT1 binds to PP1 and extends substrate binding grooves which is
154
hypothesized to create additional surface area for substrate binding (Figure 61 B).
Despite this difference, MYPT1 has also been shown to inhibit the activity of PP1 for
phosphorylase a41. However, it is unclear how this occurs since the C-terminal substrate
binding groove is exposed and therefore inhibition of PP1 must occur via a different
mechanism than by spinophilin. This suggests that the group of targeting proteins that
inhibit the activity of phosphorylase a may use many different mechanisms to alter the
Figure 61. Comparison of the PP1:spinophilin and PP1:MYPT1 structures.
A) Surface representation of the PP1(grey):spinophilin(red) crystal structure. B) Surface
representation of the PP1(grey):MYPT1(blue) crystal structure. Theoretical substrates are
shown in black. The PP1 C-terminal substrate binding groove is boxed to highlight the
difference in binding between spinophilin and MYPT1 in this groove.
substrate specificity of PP1. Also, while many targeting proteins have been shown to
inhibit the activity of PP1 for phosphoryalse a, it is unclear whether these targeting
proteins inhibit the activity of PP1 for phosphorylase a with the same IC50. It seems
plausible that since these proteins may contain different mechanisms for altering
substrate specificity, they may also inhibit the activity of PP1 to differing degrees. For
example, targeting proteins that directly block binding pockets that are necessary for
substrate binding, like spinophilin, may have a more drastic effect on the activity of PP1
than others.
155
Spinophilin and MYPT1 target PP1 to different cell types and substrates,
therefore, each of these proteins have likely evolved mechanisms for altering the
substrate specificity of PP1 that are advantageous for the specific dephosphorylation of
their substrates. However, it is currently unknown what the role of these mechanisms are
in vivo. We believe that the decreased accesible surface area for substrate binding in the
PP1:spinophilin complex has possible advantages for the dephosphorylation of PP1
substrates in the post synaptic density (PSD). The PSD contains a high concentration of
proteins when compared with the cytosol and other cellular compartments. As a result,
when spinophilin targets PP1 to the plasma membrane in the PSD there may be many
possible substrates for PP1181. Therefore, spinophilin may have evolved to decrease the
accessible surface area for binding of substrates as a mechanism to decrease
competition for PP1 in the PSD. Through decreasing competiton for the enzyme this
would result in an relative increase in the activity of PP1 for GluR1 in the PP1:spinophilin
holoenzyme compared to PP1 alone. MYPT1, which targets PP1 in muscle cells, may
not require a decrease in the accessible surface area for binding because there is a
significantly lower concentration of protein and therefore less competition for PP1
binding.
A second possibility for the evolution of decreasing the accessible surface area
for substrate binding could be to prevent non-specific dephosphorylation reactions from
taking place in the cell. This is important since the substrate specificity of PP1 is broad
in the absence of any targeting protein and because neuronal signaling in the PSD
requires tight regulatory control.
Third, spinophilin can actively inhibit PP1, as a means of increasing the
phosphorylation state for specific PP1 substrates. This is supported by a recent study
that investigated the activity of PP1 in a mouse model for Parkinsons disease20. The
156
study identified that the activity of PP1 was inhibited as a result of an overassociation of
PP1 with spinophilin. The decrease of PP1 activity led to the hyperphosphorylation of
CaMKII thereby affecting synaptic plasticity. We now know that the association of PP1
and spinophilin creates a selective inhibition against certain substrates, but not a
complete inhibition of PP1, and it is likely that the dephosphorylation of other proteins
was increased in the mouse model. Therefore, one means to actively increase the
phosphorylation state of CaMKII is to increase the interaction of spinophilin with PP1.
This result also lends support to the first two possible advantages of decreasing the
number of accessible PP1 substrate binding sites, as it confirms that competition exists
for PP1 in the post synaptic density.
One final observation from our work is related to the preferences of spinophilin
for different PP1 isoforms. Spinophilin has been shown to specifically interact with the α
and γ isoforms of PP1 but not to interact strongly with the β isofrom. This characteristic
has been shown to supposedly reside within residues 464-470 of spinophilin182. These
residues correspond to the interaction of spinophilin within the C-terminal substrate
binding groove of PP1. This groove is conserved between all three of the PP1 isoforms
and therefore the mechanism for isoform selectivity is unclear. A potential
phosphorylation site at PP1 residue 298 in the β isoform is absent in the α and γ
isoforms and may play a role in isoform selectivity. Using NetPhosK, a phosphorylation
prediction program, residue 298 received a likelihood score of 0.86 in a range from zero
to 1, where a score of 1 has the highest likelihood of being a phosphorylation site.
Additionally, this residue scores higher than any other residue in PP1 further supporting
the likelihood that this residue is a phosphorylation site. This site is positioned in close
proximity to the C-terminal substrate binding groove, and it is possible that
phosphorylation of this site creates a repulsive or steric effect on spinophilin’s interaction
157
in this area. This would significantly decrease the binding strength of spinophilin since
mutation of residues that interact in the C-terminal substrate binding groove decreased
spinophilin binding to PP1, as shown in Chapter 2.
6.2. The regulation of PP1 by GM.
GM targets PP1 to dephosphorylate phosphorylase a in vivo45. Therefore, unlike
spinophilin, GM does not inhibit the activity of PP1 for phosphorylase a. To investigate
the role of GM in the regulation of PP1 we have performed an intial characterization of
the substrate specificity of the PP1:GM complex, created a purification scheme for the
PP1:GM complex and begun NMR spectroscopy studies to determine the structure of the
CBM21 domain of GM. While we have not been able to determine the structure of
PP1:GM and therefore a mechanism for the altered substrate specificity of PP1 by G M,
we have gained valuable insights into the regulation of PP1 by GM.
The activity of PP1 and the PP1:GM complex was measured towards GluR1,
pNPP and phosphorylase a, which are the same substrates that were used for
measuring the activity of the PP1:spinophilin complex. Spinophilin inhibits the activity of
PP1 for phosphorylase a, a PP1 substrate that is not the target of the PP1:spinophilin
holoenzyme in vivo. Therefore, it seems possible that GM targets PP1 to phosphorylase
a, but inhibits the activity of PP1 for non-specific targets. However, our data demonstrate
that GM does not inhibit the activity of PP1 for GluR1, which is not a substrate of the
PP1:GM complex in vivo. Since GluR1 is believed to be dephosphorylated by interacting
in the acidic and/or hydrophobic substrate binding grooves of PP1 this, suggests that GM
does not bind to PP1 through either of these binding grooves. This suggests that GM
may not inhibit the activity of PP1 for any PP1 substrates. This seems even more
plausible when the targeting of PP1 to glycogen granules is considered. The majority of
proteins that are localized to glycogen granules are glycogen metabolic enzymes, the
158
majority of which PP1 has been shown to dephosphorylate. Therefore, the only proteins
that PP1:GM comes in contact with are PP1 substrates that GM is targeting towards and
inhibition of the activity of PP1 may not be necessary.
Our data also demonstrate that residues important for the interaction of GM with
PP1 are 63-237. This suggests that residues in the CBM domain may be important for
the interaction of GM with PP1. If correct, this adds an additional interaction site to the
CBM domain of GM which already contains defined glycogen and glycogen synthase
binding sites. This suggests that the CBM21 domain plays a different role in the
regulation of PP1 by GM than the PDZ domain of spinophilin, which appears to play only
a role in targeting and not in binding or altering the substrate specificity of PP1.
6.3. The role of flexibility in the heterotrimeric PP1:spinophilin:I-2 complex.
PP1 interacts with a second class of regulatory proteins, inhibitory proteins. These
proteins bind to PP1 and block the dephosphorylation of all PP1 substrates by blocking
the PP1 active site. Recently, it has been shown that PP1 can form heterotrimeric
complexes with specific pairs of targeting and inhibitory proteins, further enhancing the
diversity of PP1 interactions and mechanisms for altering the activity of PP165-67. One
identified heterotrimeric complex is the PP1:spinophilin:Inhibitor-2 (I-2) complex (PSI) 65.
It has previously been shown that Inhibitor-2, a single-domain, intrinsically
unstructured protein, binds to PP1, however ~75% of the total protein remains highly
flexible62. Therefore ~75% of I-2 was not built in the PP1:I-2 crystal structure. To gain
insight into the heterotrimeric PSI complex structure we initially investigated the PP1:I-2
complex structure in solution. We used a combination of NMR spectroscopy, ensemble
modeling and small angle X-ray scattering (SAXS) to create a more complete structural
view of this complex including the flexible regions of I-2. The ensemble model and the
159
low resolution molecular envelope calculated from SAXS data were determined
independently and are in good agreement confirming the accuracy of this PP1:I-2 model.
Our model shows that despite I-2's remaining flexibility, I-2 residues 55 to 127 have a
structural preference: they extend away from PP1 forming a distinct loop structure. This
loop is not fully extended and instead folds slightly onto itself.
Phopshorylation of Thr72 on I-2 results in the release of the I-2 helix from the
active site of PP1, thereby reactivating the enzyme63,64. Our model shows that Thr72 is
located at the top surface of this loop (Figure 62). The extension of this phosphorylation
site away from the surface of PP1 may create a binding site for the kinase which does
not compete with binding sites for PP1. One possible mechanism for the reactivation of
PP1 is that phosphorylation of Thr72, which resides in the most packed loop region,
causes a destabilization of the loop structure, which in turn, causes the helix to be
released from PP1 (Figure 62 C). It is also known that after reactivation, Thr72 becomes
dephosphorylated by PP1, resulting in a return to the inactivation of PP1. It seems likely
that the flexibility in this region allows this loop to fold down into the active site of PP1,
where Thr72 can become dephosphoryalted thereby causing re-inhibition of PP1.
Figure 62. Threonine 72 of I-2 in the PP1:I-2 ensemble model.
A) An overlay of the nineteen ensemble models of the PP1:I-2 complex with PP1 shown as a
surface representation in grey and I-2 shown as a cartoon representation in cyan. I-2 Thr 72 is
shown as a stick model in red. B) 90 rotation of A. C) A close-up view of the box shown in B.
160
After the generation of the PP1:I-2 ensemble model we investigated the
formation of the heterotrimeric PP1:spinophilin:I-2 complex. By overlaying the
PP1:spinophilin crystal structure and the PP1:I-2 ensemble model it is readily visible that
spinophilin and I-2 have two overlaping interaction sites with PP1, suggesting that a
structural rearrangement must occur for the formation of the heterotrimeric complex
(Figure 83). We used NMR spectroscopy and biochemistry to identify interaction regions
that rearrange upon formation of the heterotrimeric PP1 complex. Subsequently, we
have used these results, in combination with SAXS and molecular dynamics simulations,
to create a structural model for the heterotrimeric PSI complex. Our data demonstrate
that when the heterotrimeric complex is formed, I-2 releases from the RVxF binding site,
leaving only the SILK motif and helix interactions to stabilize its interaction with PP1.
Additionally, the helix of I-2 remains unchanged in the heterotrimeric complex,
suggesting that the helix of spinophilin must release from or shift its position on PP1.
The release of these binding sites creates additional flexibility that was not observed in
the heterodimeric PP1 complexes.
Previously it was demonstrated that mutation of the RVxF motif, or the addition
of a synthetic RVxF peptide, completely ablates the binding of many PP1 regulators35,68.
This suggests that the RVxF motif is necessary for the binding of PP1 regulators.
However, our data demonstrates that I-2 can remain bound to PP1 without interacting
through the RVxF motif and therefore raises two interesting questions about the role of
the RVxF motif in the binding of I-2 to PP1. 1) Why does I-2 contain an RVxF motif if it is
able to interact with PP1 without it? 2) Is there a role for the RVxF motif that is
functionally distinct in the PP1:I-2 and PSI complexes? We have begun to hypothesize
what the role of the RVxF motif of I-2 is in an effort to answer these questions. In the
PP1:I-2 complex, phosphorylation of Thr72 results in the release of the I-2 helix from
161
PP1. It is possible that without the RVxF motif interraction, the SILK motif interaction
might not be strong enough to maintain the PP1:I2 interaction. As a result, the
phosphorylation of Thr72 in the heterotrimeric PSI complex may cause the dissociation
of PP1 and I-2, instead of just the release of the helix. A second possibility is that
phosphorylation of this site may no longer lead to release of the helix from the active site
and that reactivation of PP1 may not occur in PSI or it may occur by other means. Lastly,
the release of the RVxF motif may create additional binding surfaces on I-2 that were
previously not accesible because of steric hinderance by PP1. For example, residues in
I-2 that are directly N and C-terminal to the RVxF motif lie in close proximity to PP1 in
the PP1:I-2 complex and are likely inaccesible to protein binding. However, in the PSI
complex these residues are flexibile and are accessible for interaction with other
proteins.
The release of the RVxF motif of I-2 also confirms the importance of secondary
interaction sites, such as the I-2 SILK motif for PP1 binding. Originally the RVxF motif
was the best and only defined interaction site for PP1, however with the determination of
PP1:regulatory protein structures, it is becoming clear, that most PP1 regulatory proteins
contain additional binding motifs43,62. These interaction sites can also play a direct role in
the function of the regulator, i.e. the helix of I-2 directly blocks the PP1 active site. Lastly,
our binding data demonstrates that, while the RVxF motif was originally thought to be the
most important interaction motif, secondary interaction sites such as the I-2 SILK motif
also play a significant role in binding.
6.4. Conclusions.
PP1 regulators contain little sequence identity and the majority only share a short
conserved interaction motif, the RVxF motif, which suggests that these proteins are
highly diverse regulators of PP1 function. However, it was previously unknown if the low
162
sequence
different
identity
binding
supports
interactions
between PP1 regulators and
PP1. We have characterized
three
PP1:regulatory
complexes
including
PP1:spinophilin,
Figure 63. Overlay of PP1:regulatory protein
structures.
PP1 (grey, surface representation) is shown with the
three known regulatory protein structures. Spinophilin
(red) I-2 (cyan) and MYPT1 (blue) are all shown as
cartoon representations. The 90 rotation highlights the
RVxF motif interaction which is the only conserved
interaction between all three proteins.
protein
PP1:I-2
PP1:spinophilin:I-2
the
and
complexes
and begun work on a fourth, the
PP1:GM complex. All of our
data,
together
with
the
PP1:MYPT1 structure supports the idea that regulatory proteins interact with PP1
through unique interaction sites which drive diverse functions through either altering the
substrate specificity or inhibiting the activity of PP143. A comparison of the
PP1:spinophilin, PP1:I-2 and PP1:MYPT1 structures, shows that these structures
contain no structural conservation outside of the RVxF motif, which is in good agreement
with the diversity suggested based on the sequence comparison34. Additionally, an
overlay of all three structures reveals that these regulators cover the majority of the
catalytic face of PP1. This, in combination with biochemical data for the interaction of
PP1 with another targeting protein, Sds22, begins to demonstrate that the majority of
PP1 surfaces serve as a protein interaction surface. However, due to the large number
(~200) of PP1 regulators it is impossible for each of the regulators to interact with a
completely novel interaction site on PP1. This suggests that there are only a limited set
of possible mechanisms for modifying the activity of PP1. This combinatorial problem
becomes more difficult because ~2/3 of PP1 regulatory proteins are predicted to interact
with PP1 via an unstructured domain21. One advantage of unstructured proteins is that
163
they can fold upon binding forming diverse and unique structures that a folded protein
might not be able to adopt. Therefore, we believe that this might allow PP1 regulators to
interact using identical interaction sites, but forming unique structures, which will modify
the substrate specificity and activity of PP1 in different ways. If the regulatory proteins
interact with PP1 in truly novel ways and alter the substrate specifity by unique
mechanisms, it seems plausible that PP1 could become ~200 seemingly different
holoenzymes through its interaction with its regualtory proteins.
6.5. Short term outlook.
The work presented here has led to numerous hypotheses about the role of I-2 in the
heterotrimeric PSI complex and the role of the CBM21 domain in the regulation of PP1
by GM. Our model of the PSI heterotrimeric complex suggests that phosphorylation of
Thr72 in I-2 may not result in the reactivation of PP1, as it does in the PP1:I-2 complex.
We hypothesize that phosphorylation of Thr72 may have either no effect on the PSI
complex or cause the dissociation of I-2 from PSI. To test these hypotheses, the
heterotrimeric PSI complex will be purified and GSK-3β (Invitrogen) will be used to
phosphorylate Thr72 on I-2. Subsequntly, the phosphorylated heterotrimeric PSI
complex will be subjected to SEC. We predict that if phosphorylation leads to
dissociation of I-2 that the PP1:spinophilin complex and I-2 will elute separately from the
column. Additionally, we will meaure the dissociation constant of phosphorylated and
non-phosphorylated I-2 for the PP1:spinophilin complex using ITC. We predict that if
phosphorylation causes the dissociation of I-2 from the PSI complex that the dissociation
constant of phosphorylated I-2 would be higher than the dissociation constant of nonphosphorylated I-2. If phosphorylation of I-2 does not cause the release of I-2 from the
PSI heterotrimeric complex, we propose to assay the activity of PP1 for pNPP in Thr72
phosphorylated and non-phosphorylated PSI complexes. In this situation we hypothesize
164
that phosphorylation of Thr72 will not effect the activity of PSI and that the complex will
remain inactive towards pNPP.
Currently, it is believed that the CBM21 domain of GM acts as both a targeting
domain and PP1 substrate interaction domain56. However, it is currently unclear how
both of these interactions are supported. Therefore, we propose to finish the structure
determination of the CBM21 domain and investigate the interactions of four possible
binding partners of the CBM21 domain including a glycogen analog (β-cyclodextrin),
glycogen synthase, phosphorylase kinase and phosphorylase a using ITC. All of these
substrates have been shown to interact with at least one G subunit family member
through the CBM21 domain. It is unclear if all of the G subunit family members interact
with all of the substrates of PP1 or if the different CBM21 domains are selective towards
specific PP1 substrates. Additionally, many of the binding sites for these substrates have
not been investigated using structural biology techniques. Therefore, we propose to use
NMR spectroscopy to map the binding interaction sites for any interacting partners that
are confirmed using ITC. Interestingly, our ITC data for the interaction of PP1 and GM
suggests that the CBM21 domain may play a role in PP1 binding. To further characterize
the interaction between PP1 and GM we plan to complete the sequence specific
backbone assignment of GM 63-237. After completion of the assignment, we plan to
purify the PP1:GM complex using 2H and 15N labeled GM 63-237 and record a 2D [1H,15N]
HSQC spectrum. Peaks that shift between the 2D [1H,15N] HSQC spectra of GM and
PP1:GM indicate residues in GM that are involved in binding. This will also help to define
the smallest construct of GM to use for the attempted crystallization of the PP1:GM. We
believe that these experiments will further our understanding of the complex regulation
of PP1 by its targeting and inhibitory proteins.
6.6. Long term outlook.
165
We believe that it is an exciting time to study the regulation of PP1. The data presented
here and elsewhere point towards a large amount of diversity of the interactions with
PP1 and therefore the regulatory mechanims43,62. We believe that we have only begun to
scratch the surface of these interactions. This hypothesis can only be confirmed by
determining additional structures of PP1:regulatory protein complexes. These structures
may reveal whether there are conserved mechanisms for altering the substrate
specificity. These structures may also reveal if the unstructured regulatory proteins bind
to PP1 forming similar secondary structural elements. Additionally, to fully understand
the regulation of PP1 by targeting proteins it will be important to measure the activity of
numerous PP1:regulatory protein complexes for different substrates. These assays may
reveal classes of substrates that are dephosphorylated by specific PP1:targeting protein
complexes that are inhibited by others. As more PP1:regulatory protein structures are
determined and more biochemical data become available it will be interesting to see
what and how many additional regulatory mechanisms exist for altering the activity of
PP1.
Appendix A. The effect of metal binding on the activity
of PP1.
166
167
A.1. Introduction.
The catalytic mechanism for the dephosphorylation of substrates by protein phosphatase
1 (PP1) relies on the specific coordination of two metal ions at its active site. Each metal
ion is coordinated by five ligands, including PP1 residues Asp64, His66, Asp92, His 248,
His 173, Asn124 and 2 water molecules (Figure 64)26. The metal ions act to stabilize
negative charges on the phosphate and support the nucleophilic attack of the phosphate
by a water molecule (Figure 64). For in vitro expression, PP1 is often expressed in
media that is supplemented with
MnCl2,
which is required to
produce
active
Therefore,
the
protein.
metal
ions
present at the active site in all
but one of the available crystal
structures is Mn2+
27,43
. Only the
crystal structure of PP1 in the
presence
Figure 64. Metal coordination of PP1.
The two metal ions bound at the active site are shown
as pink circles labeled M1 and M2. Amino acids
involved in metal coordination and catalysis are shown
with distances labeled between atoms in Å. This figure
is adapted from Goldberg et al. 1995.
of
tungstate,
a
phosphate analog, contained an
equivalent molar ratio of Fe2+
and
Mn2+70.
Currently,
the
identity of the metal ions at the active site of PP1 in vivo is an ongoing subject of debate.
The most common hypothesis is that these ions are Fe2+ and Zn2+. This assumption is
based on similarities between the metal ion coordination of PP1 and the purple acid
phosphatase, which contains 1 Fe2+ and 1 Zn2+ ion bound at its active site26,183. The
activity of PP1 is most likely dependent on the identity of the metal ions. Therefore, we
have tried to deplete the Mn2+ ions at the active site of PP1 using EDTA in attempt to
substitute Mn2+ for different metal ions. Additionally, we have expressed PP1 in the
168
presence of Zn2+, Cu2+, Ni2+, Fe2+ and Mn2+ and tested the activity of PP1 against the
model substrate pNPP.
A.2. Methods.
A.2.A. Protein expression and purification.
For PP1 expression in the presence of different metals, E. coli BL21 (DE3) cells
(Invitrogen) were co-transformed with rabbit PP1α 7-330 in RP1B and the vector pGro7
(Takara), which encodes the GroEL/ES chaperone. Colonies were used to inoculate LB
supplemented with 100 mM Mops pH 7.4 and 10 μM of either ZnCl2, CuCl2, FeCl2 or
NiCl2. Cells were grown at 30°C under vigorous shaking (250 rpm) until an OD600 of 0.5
was reached. Arabinose was added at 2 g/L of LB to induce the expression of the
chaperone. Cells were subsequently grown to an OD600 of 1.0 and then placed in ice
water baths. The incubator was cooled to 10°C, and expression of PP1 was induced by
the addition of 0.1 mM IPTG. Cells were grown under vigorous shaking (250 rpm) at
10°C overnight. After ~20 hours cells were harvested by centrifugation (6000 xg, 10 min,
4°C). Cells were resuspended in fresh LB containing 100 mM Mops pH 7.4, 10 μM of
either ZnCl2, CuCl2, FeCl2 or NiCl2 and 200 μg/ml of chloramphenicol. Cells were
incubated at 10°C under vigorous shaking (250 rpm) for two hours to allow for additional
refolding. Cells were harvested by centrifugation (6000 xg,10 min, 4°C) and stored at 80°C.
Spinophilin 417-583 was produced following the protocol in Chapter 2. The first
day of the PP1:spinophilin complex purification was identical to the protcol in Chapter 2
except that each purification buffer was supplemented with either 10 μM ZnCl2, FeCl2,
NiCl2, or CuCl2 instead of MnCl2. The activity of the PP1:spinophilin complex was
measured after the first day of purification.
169
A.2.B. Measuring the activity of PP1 against pNPP.
Dephosphorylation reactions were initiated by the addition of the PP1:spinophilin
complex to 0.5, 1, 2, 4, and 8 mM pNPP. Reactions were incubated for 30 min at 30°C
and quenched by addition of 1 M NaOH. The amount of p-nitrophenolate product was
determined using the absorbance at 405 nm with a molar extinction coefficient of 18,000
M−1 cm−1. Sigmaplot 8.0 was used to fit the data to a one site ligand binding model and
determine the kcat and KM.
A.3. Results.
EDTA is a small molecule chelator
that has an extremely high affinity for
metal ions. As such it has been used
to sequester the metal ions from
several
metalloenzymes
as
a
mechanism to inactivate them184. We
have attempted to use EDTA to
Figure 65. The effect of EDTA on the solubility
of PP1.
Chromatograms are shown for SEC purifications
prior to the addition of EDTA (blue) and after the
addition of EDTA (red).
chelate the Mn2+ ions from the active
site of PP1 to facilitate the substitution
with other metal ions. This will allow
for the investigation of the metal ion dependency of PP1. After purification of the
PP1:spinophilin complex, 90 mM EDTA was added to chelate the Mn2+ ions from the
active site, and incubated for ~12 hours at 4°C. Subsequently, the PP1:spinophilin
complex was subjected to SEC using a Superdex 75 (26/60) column in 20 mM Tris pH
7.5, 50 mM NaCl and 0.5 mM TCEP. Prior to the adition of EDTA the PP1:spinophilin
complex eluted at a peak at 150 ml, which is the expected elution volume for the
PP1:spinophilin complex. However after the addition of EDTA, PP1 eluted primarily in a
170
peak at the void volume demonstrating that chelation of Mn2+ ions from the active site of
PP1 induces aggregation of the enzyme likely through unfolding. Therefore the metal
ions are not only required for the activity of the enzyme, but they are also required for
proper folding.
To investigate the metal
dependency of PP1 we have
expressed PP1 in the presence
of 10 μM CuCl2, ZnCl2, FeCl2,
and NiCl2. None of the metal ions
inhbited the growth of the E. coli
cultures,
since
all
cultures
reached a final OD600 of 2.2.
Figure 66. Michelis-Menten plot for
PP1:spinophilin in 10μM CuCl2.
A plot of velocity (v) against pNPP concentration. The
data are fit to a one site ligand binding model using
Sigmaplot 8.0.
Since PP1 can not be readily
purified to more then 80% purity
in its apo form we have tested the
Metal Ion
-1
kcat (s )
-1 -1
KM (M)
kcat/KM (M s )
−3
Mn
0.46
4.03 x 10
Fe
0.15
3.48 x 10
Zn
0.13
2.61 x 10
Ni
Cu
0.12
0.10
−3
−3
−3
2.96 x 10
−3
3.32 x 10
activity
of
PP1
containing
114.1
43.1
different metal ions by purifying
49.8
the PP1:spinophilin complex and
40.5
30.1
Table 11. The depedence of PP1 catalytic activity
on metal ions.
measuring
its
activity
against
pNPP. Expression of PP1 in the
presence of different metals did
not negatively affect the purification of the PP1:spinophilin complex. The activity of the
PP1:spinophilin complex was measured against 5 different concentrations of pNPP
(Figure 66). PP1:spinophilin complex containing Mn2+ had the highest catalytic turnover
of pNPP with a kcat of 0.46 s−1. This was ~3 times higher than the second most active
171
PP1:spinophilin complex, the Fe2+ containing complex. The Fe2+, Zn2+, Ni2+, and Cu2+
complexes had similar kcat values ranging from 0.15 to 0.10 sec−1(Table 11).
Interestingly, the Mn2+ containing complex had the lowest binding affinity for pNPP with a
KM of 4.03 x 10−3 M. The catalytic efficiency of an enzyme for a particular substrate is
kcat/KM4. Despite the increase in KM the Mn2+ containing enzyme has the highest catalytic
efficiency (kcat/KM) at 114.1 M−1s−1, which takes into acount both the catalytic turnover
rate and the binding affinity of the enzyme for substrate.
A.4. Discussion.
PP1 expressed in E. coli is Mn2+ dependent and has different properties for some of its
inhibitor proteins185. Inhibitor-1 (I-1) is a pseudosubstrate inhibitor, which when
phosphorylated interacts direclty at the active site of PP1 to inihibit its activity. In vivo
PP1 is incapable of dephosphorylating I-1. However in vitro, Mn2+ containing PP1 rapidly
dephosphorylates Inhibitor-1. Our data demonstrate that PP1 expressed in the presence
of Mn2+ was more active towards pNPP than PP1 expressed in the presence of other
metals. This result may begin to explain how PP1 is able to dephosphorylate I-1 in vitro
but no in vivo.
Appendix B. Small angle X-ray scattering analysis of the
unbound forms of Spinophilin and Inhibitor-2.
172
173
B.1. Introduction.
Many PP1 regulatory proteins are predicted to interact with PP1 through an intrinsically
unstructured domain. As a result X-ray crystallography can not be used to investigate
the structure of the unbound forms of the PP1 regulators. Inhibitor-2 (I-2) and the PP1
binding domain of The PP1 binding domain of spinophilin, residues 417-494, is
completely unstructured when not bound to PP1. Both of these regulators have been
studied using NMR spectroscopy to investigate the presence of transient secondary and
tertiary structure57. Small angle X-ray scattering (SAXS) can also be used to investigate
the structural properties of unfolded proteins, including the radius of gyration (Rg) and
the foldedness. Here we report SAXS data for the unbound forms of both I-2 and
spinophilin and compare the results with both NMR data that has been previously
reported and ensemble modeling data from Chapter 457.
B.2. Methods.
B.2.A. Protein expression and purification.
I-29-164 was expressed and purified as previously described57. Spinophilin417-583 was
expressed and purified as described in Chapter 2.
B.2.B. SAXS data acquisition.
Spinophilin417-583 and I-29-164 were purified, frozen in liquid nitrogen and stored at−80 °C.
Prior to scattering experiments the sample was thawed on ice diluted to the desired
concentration and filtered through a 0.02 µm filter (Whatman). Synchrotron X-ray
scattering data were collected at the National Synchrotron Light Source (NSLS)
beamline X9. Small angle X-ray scattering (SAXS) data were collected using a MarCCD
165 located at 3.4 m distance from the sample. Wide angle X-ray scattering (WAXS)
data were collected simultaneously with SAXS data using a Photonic Science CCD
located at 0.47 m from the sample. 20 µl of sample was continuously pushed through a 1
174
mm diameter capillary for either 180 or 360 s of total measurement time and exposed to
a 400 x 200 microns X-ray beam. Scattering data for spinophilin was collected at 1.5
mg/ml and 2.4 mg/ml. Scattering data for I-2 was collected at 1.6 mg/ml and 2.5 mg/ml.
Normalization for beam intensity, buffer subtraction and merging of the data from both
detectors were carried out using PRIMUS168. A Guinier approximation, I(q) = I(0)exp(q2Rg2/3), where a plot of I(q) and q2 is linear for q<1.3/Rg, was performed on four
independent scattering trials and averaged to determine the radius of gyration115.
HYDROPRO was used to calculate the Rg for the spinophilin PDZ domain and for the
PP1:spinophilin holoenzyme for comparison169.
B.3. Results.
Scattering data for spinophilin417-583 have been recorded at 1.5 and 2.4 mg/ml (Figure
67). The Kratky plot for spinophilin417-583 demonstrates that the protein is partially folded.
This is in good agreement with NMR studies of spinophilin417-583 which demonstrates that
Figure 67. SAXS data for spinophilin.
A) Scattering data for spinophilin 417-583. Error bars are shown in grey B) A Kratky plot
demonstrating the folding state of spinophilin. C) The Guinier region of the scattering data for
2.4 and 1.5 mg/ml spinophilin.
the PP1 binding domain of spinophilin (residues 417-494) is unstructured (Chapter 2)
and the PDZ domain (residues 495-583) is folded52. We have determined that the radius
of gyration (Rg) using the Guinier approximation to 29.3 Å (Figure 67, Table 12). No Rg
has been determined previously for this construct and so there is no basis for direct
comparison. However, the radius of gyration of the PDZ domain is 13.6 Å as determined
175
using HYDROPRO and the radius of gyration of the PP1 binding domain Ensemble
model is 16.4 Å as calculated in Chapter 4169. Both of these Rgs are smaller than the Rg
of spinophilin 417-583 as expected. The Rg for the PP1:spinophilin crystal structure
determined using HYDROPRO is 25.7 Å. This is also smaller than the Rg of unbound
spinophilin, which further confirms that the
Protein
Spinophilin417-583
I-29-164
Rg (Å)
29.3 ± 0.8
34.5 ± 1.3
Table 12. Radius of Gyration for
spinophilin and I-2 determined by SAXS.
PP1 binding domain of spinophilin folds
upon binding to PP1.
.
SAXS data have been recorded for I-2 9-
164 (Figure 68). The Kratky plot for I-2 demonstrates that I-2 is unfolded in solution
(Figure 68 B). This is in good agreement with previously published NMR data, which
demonstrates
that
I-2
is
unstructured
Figure 68. SAXS data for I-2.
A) Scattering data for spinophilin 417-583. Error bars are shown in grey B) A Kratky plot
demonstrating the folding state of I-2. C) The Guinier region of the scattering data for 2.5 and
1.6 mg/ml I-2.
in solution57. Using the Guinier approximation we have determined that the Rg for I-2 is
34.5 Å (Figure 68 C, Table 12). This is in excellent agreement with the Rg
calculated for the ensemble model of I-2 in Chapter 4, which is 34.6 Å. Therefore
we have demonstrated that our calculated ensemble model agrees well with the
experimental SAXS data.
Appendix C. Supplementary information for “Spinophilin
directs protein phosphatase 1 specificity by blocking
substrate binding sites.”
The following appendix was published as Supporting Information for the manuscript of the
above name in the journal Nature Structural & Molecular Biology, Volume 17, Number 4, April
2010, pages 459-464.
176
177
Figure 69. Ragusa, Supplementary Figure 1.
Electron density for the spinophilin PP1 binding domain interacting with PP1α. Sigma-A 2mFo
– DFc map of the PP1 binding domain of spinophilin contoured at 1σ to 1.85Å (black mesh).
The PP1 binding domain of spinophilin is shown in red as a stick representation with the RVxF
sequence in green and the C-terminal groove binding motif in turquoise. PP1α and the PDZ
domain of spinophilin have been removed for clarity. Inset: The C-terminal groove binding
motif is shown in turquoise with residues 465 and 468 removed for clarity.
178
The neurabin426-592:PP1α7-330 complex. (a), Cartoon representation of the neurabin426-592 PP1
binding (red) and PDZ domains (purple) and PP1α7-330 (gray; surface representation) complex;
2+
centered on the active site of PP1α7-330. Two Mn -ions (pink spheres) mark the active site of
PP1.Neurabin residues interacting with the RVxF binding pocket are represented as sticks in
green and those interacting with the PP1 C-terminal groove are represented as sticks in cyan.
(b), Cartoon representation of the neurabin426-592(orange):PP1α7-330(gray) complex
superimposed with the spinophilin417-583(magenta):PP1α7-330(gray) complex; RVxF motifs
2+
shown in green, C-terminal groove binding motif shown in cyan, and Mn -ions shown as pink
spheres. The structures were superimposed using the mainchain coordinates of PP1α7-330,
spinophilin431-583, and neurabin440-592. The structures of the two complexes are virtually
identical (rmsd = 0.52 Å). (c), Sequence comparison of rat spinophilin residues 417-483
(upper) and rat neurabin residues 426-592 (lower). Identical residues are highlighted in dark
grey and similar residues are highlighted in light grey. The spinophilin PP1-binding and PDZ
domains are 76.6% identical and 86.2% similar to the PP1-binding and PDZ domains of
neurabin.
179
The interaction of neurabin with the PP1 RVxF binding pocket and the C-terminal
binding groove are identical to those of spinophilin. Enlarged views of the structural
comparison of neurabin and spinophilin for the PP1 RVxF binding site and the Cterminal groove. Neurabin residues (RVxF sequence (dark green); C-terminal groove
binding motif (dark teal)), spinophilin residues (RVxF sequence (light green); Cterminal groove binding motif (light teal)) and PP1 (gray, surface and stick
representation for interacting residues) are shown. Hydrogen bonds are represented
as black dashes. (a) The neurabin RVxF sequence (456RKIKF460) forms similar polar
contacts as the spinophilin RVxF sequence (447RKIHF451). (b) Neurabin residues 471478 interact with the C-terminal groove of PP1 in the same manner as spinophilin
residues 462-469. Neurabin residues 474 and 477 have been omitted.
180
Circular dichroism (CD) spectra of eleven spinophilin mutants of spinophilin. I449A,
H450A, F451A (RVxF motif; interaction region I), Q457A, F459D, T461A (β-sheet
area; interaction region II), Y467A, R469A, R469D (residues essential for the
interaction with the PP1 C-terminal groove; interaction region IV), as well as E482A
and E486A (α-helix; interaction region III) and wild-type spinophilin (WT). The lack of
differences shows that the introduced mutations do not disrupt the overall fold of
spinophilin.
181
The 2.0 Å crystal structure of the spinophilin417-583:PP1α7-330:Nodularin-R complex.
Cartoon representation of the spinophilin417-583 PP1 binding (red) and PDZ domains
(purple) and PP1α7-330 (grey; surface representation) complex. Two Mn2+-ions (pink
spheres) are bound at the active site of PP1. Nodularin-R is shown as a stick model
in orange. As illustrated in the manuscript in Figure 4, Nodularin-R binds identically to
both PP1 alone (PDBID 3E7A) and to the PP1:spinophilin holoenzyme (PDBID
3EGH; shown here), highlighting the vast accessibility of the PP1 active site even
upon spinophilin binding and accounting for why it is possible to purify the
PP1:spinophilin complex using microcystin-LR beads.
182
Different modes of directing PP1 substrate specificity by targeting proteins. (a),
Surface representation of the spinophilin417-583 (red) and PP1α7-330 (grey) complex.
Two Mn2+-ions (pink spheres) are bound at the active site of PP1. When
spinophilin417-583 is bound to PP1 the accessible substrate binding surface is
decreased. This ensures that substrates (shown as a purple line) are exclusively
selected that bind into the hydrophobic or acidic substrate binding grooves. (b)
Surface representation of the MYPT11-299 (blue) and PP1δ (grey) complex (PDB
1S70). Two Mn2+-ions (pink spheres) are bound at the active site of PP1. When
MYPT11-299 is bound to PP1 the substrate binding grooves of PP1 are significantly
extended, which requires substrates (shown as an orange line) to interact with these
elongated surface binding grooves and noticeably increases the substrate specificity.
(c) Surface representation of the MYPT11-299 (blue) and PP1δ (grey) complex (PDB
1S70) with a different potential substrate (shown as a black line).
183
Neurabin426-592:PP1α7-330
Data collection
Space group
Cell dimensions
a, b, c (Å)
α, β, γ (°)
Resolution (Å)*
Rmerge*
I / σI*
Completeness (%)*
Redundancy*
C2
120.9, 83.7, 108.8
90.0, 93.6, 90.0
20.0-2.20 (2.24-2.20)
0.103 (0.263)
10.4 (5.0)
97.9 (97.5)
3.3 (3.4)
Refinement
Resolution (Å)
20.0 - 2.20
No. reflections
53705/2704
Rwork / Rfree
0.161/0.221
No. atoms
Protein
6456
Ligand/ion
14
Waters
560
B-factors
Protein
23.0
Ligand/ion
21.0
Waters
32.0
R.m.s. deviations
Bond lengths (Å)
0.013
1.34
Bond angles (°)
Table 13. Ragusa, Supplemental Table 1.
Data collection and refinement statistics for the neurabin426-592:PP1α7330 complex.One crystal was used for each structure
*Values in parentheses are for highest-resolution shell.
Appendix D. Supplementary information for “Structural
diversity in free and bound states of intrinsically
disordered protein phosphatase 1 regulators.”
The following appendix is in press as Supporting Information for the manuscript of the above
name currently inpress in the journal Structure.
184
185
Figure 75. Marsh, Supplementary Figure 1.
Comparison of experimental I-2 secondary chemical shifts (black) to SHIFTX predicted values
from calculated ensembles (red) and TraDES ‘Coil’ ensembles (blue).
186
Comparison of experimental spinophilin secondary chemical shifts (black) to SHIFTX
predicted values from calculated ensembles (red) and TraDES ‘Coil’ ensembles (blue).
187
Comparison of experimental DARPP-32 secondary chemical shifts (black) to SHIFTX
predicted values from calculated ensembles (red) and TraDES ‘Coil’ ensembles (blue).
188
Comparison of experimental I-2 PRE intensity ratios (black) to predicted values from
calculated ensembles (red) and TraDES ‘Coil’ ensembles (blue).
189
Comparison of experimental spinophilin PRE intensity ratios (black) to predicted values from
Comparison of experimental DARPP-32 PRE intensity ratios (black) to predicted values from
190
15
Comparison of experimental (A) I-2, (B) spinophilin and (C) DARPP-32 N R2 values (black)
to the number of heavy atom contacts within an 8 Å radius for the calculated ensembles (red)
and TraDES ‘Coil’ ensembles (blue).
191
SAXS analysis of the PP1:I-2 complex. (a) Kratky plot demonstrating the folding state of
PP1:I-2. (b) Pair-distance distribution function for PP1:I-2 confirms the Rg determined by the
Guinier approximation. (c) Average fit of the ab initio models (red line) to the scattering data
(squares).
Protein
Cluster
Population
a
Rg (Å)
Average CαCα
distance matrix
RMSD (Å)
18.4
18.2
14.8
9.3
8.4
14.2
15.3
16.8
1
0.44 ± 0.07
25.5
I-29-164
2
0.17 ± 0.07
30.1
3
0.14 ± 0.03
42.5
1
0.68 ± 0.16
15.4
Spinophilin417-494
2
0.10 ± 0.01
16.9
1
0.50 ± 0.04
21.0
DARPP-321-118
2
0.29 ± 0.09
29.2
3
0.12 ± 0.07
37.1
Table 14. Marsh, Supplemental Table 1.
 
Significantly populated clusters based upon C C distance matrix RMSDs for the three
unbound ensembles.
a
Errors represent the standard deviation of each cluster population between independently
calculated ensembles.
Appendix E. Supplementary information for “Molecular
investigations of the structure and function of the
protein phosphatase 1:spinophilin:inhibitor-2
heterotrimeric complex.”
The following appendix is in preparation as Supporting Information for the manuscript of the
above name that is currently in preparation.
192
193
Figure 83. Dancheck, Supplementary Figure 1.
Overlay of the PS (PDB ID: 3EGG) and PI (PDB ID: 208A) crystal structures. PP1 is shown as
a surface representation in grey. Spinophilin and I-2 are shown as cartoon representations in
red and cyan, respectively. Red circles highlight two regions of overlap between the two
structures. (Left) Helices of spinophilin and I-2 share a common interaction site near the
o
hydrophobic substrate binding groove. (Right) A 90 rotation of the structure on the left shows
that spinophilin and I-2 both interact with PP1 in the RVxF binding pocket.
194
CD spectra of I-2 mutants show that the mutations do not affect transient structure in I-2. CD
spectra were recorded for I-2 C85S and the four I-2 mutants used in the pull-down assays and
are shown in A. % alpha helical character of each mutant was calculated from the molar
ellipticity at 222 nm and is shown in B. The SILK mutant is I13A/L14A/K15A, RVxF mutant 1 is
W46A, RVxF mutant 2 is K42A/Q44G, and the helix mutant is R157A/D163A.
195
SAXS data for the PSI complex. (A) The experimental scattering data for the PSI complex. (B)
A Kratky plot demonstrates that the PSI complex is folded. (C) A pair distance distribution plot
confirms the Rg determined using the Guinier approximation.
Appendix F. The extended PP1 toolkit: designed to
create specificity.
196
197
Reproduced from the journal Trends in Biochemical Sciences, Volume 35, Issue 8,
August 2010 pages 450-458.
This article appeared with the cover image for the August 2010 issue. The cover image
was designed by M.J.R. and is included on the next page.
Mathieu Bollen1, Wolfgang Peti2, Michael J. Ragusa2 and Monique Beullens1
1 Laboratory of Biosignaling & Therapeutics, Department of Molecular Cell Biology,
University of Leuven, B-3000 Leuven, Belgium
2 Department of Molecular Pharmacology, Physiology and Biotechnology, Brown
University, Providence, RI 02912, USA
Corresponding authors: Bollen, M. (Mathieu.Bollen@med.kuleuven.be); Peti, W.
(Wolfgang_Peti@Brown.edu).
198
Figure 86. The cover of Trends in Biochemical Sciences for August 2010.
PP1 is not a simple housekeeping enzyme. On pages 450-458 of the August 2010 issue,
Mathieu Bollen et al. discuss recent data that explains how PP1 uses a diverse regulatory
toolkit to direct highly specific enzymatic reactions. The cover is designed to highlight the
structural diversity of the regulatory proteins that interact with PP1. Design by Michael J.
Ragusa.
199
Protein Ser/Thr phosphatase-1 (PP1) catalyzes the majority of eukaryotic protein
dephosphorylation reactions in a highly regulated and selective manner. Recent studies
have identified an unusually diversified PP1 interactome with the properties of a
regulatory toolkit. PP1-interacting proteins (PIPs) function as targeting subunits,
substrates and/or inhibitors. As targeting subunits, PIPs contribute to substrate selection
by bringing PP1 into the vicinity of specific substrates and by modulating substrate
specificity via additional substrate docking sites or blocking substrate-binding channels.
Many of the nearly 200 established mammalian PIPs are predicted to be intrinsically
disordered, a property that facilitates their binding to a large surface area of PP1 via
multiple docking motifs. These novel insights offer perspectives for the therapeutic
targeting of PP1 by interfering with the binding of PIPs or substrates.
F.1. PP1 and the challenge of specificity.
Protein phosphorylation represents one of the most common post-translational
modifications in eukaryotes. It affects 30–70% of all cellular proteins and some cellular
processes, such as the entry into mitosis, are associated with thousands of
phosphorylation events
by
186
and 2. In mammals, phosphorylation reactions are catalyzed
500 protein kinases187. The majority of these phosphorylation events are highly
dynamic owing to their ability to be rapidly reversed by protein phosphatases. Whereas
the numbers of protein tyrosine kinases and phosphatases are well balanced ( 100
each), intriguingly, the mammalian genome encodes only
phosphatases to offset
40 protein Ser/Thr
400 Ser/Thr kinases1. This discrepancy raises the key question
of how do so few protein Ser/Thr phosphatases reverse the actions of this large number
of protein kinases in a specific and regulated manner? The emerging consensus is that
the diversity of protein Ser/Thr phosphatases is, in decisive contrast to Ser/Thr kinases,
not achieved primarily by gene duplication, but rather by their unparalleled ability to form
200
stable protein–protein complexes. This property results in the accumulation of an
abundant number of phosphatase holoenzymes, each with its own substrate and mode
of regulation. This concept has been well illustrated for protein phosphatases-1 (PP1)
and -2A (PP2A), which belong to the phosphoprotein phosphatase (PPP) superfamily of
protein Ser/Thr phosphatases, and together account for more than 90% of the protein
phosphatase activity in eukaryotes
1,32
. Recent data suggest that mammals contain as
many as 650 distinct PP1 complexes and approximately 70 PP2A holoenzymes188,
indicating that PP1 catalyzes the majority of protein dephosphorylation events in
eukaryotic cells.
In this review, we discuss recently acquired insights that help to explain how PP1
functions in a specific and regulated manner. First, we address the broad substrate
specificity of the free catalytic subunit and discuss how its action is controlled by a
substrate-targeting and inhibitory toolkit. Next, we discuss how these PP1-interacting
proteins (PIPs) form stable complexes with PP1 via degenerate docking motifs, often in
the context of a structurally disordered interaction domain. Finally, we highlight the
molecular mechanisms of substrate selection and holoenzyme regulation, and explore
how structural insights can be used to develop PP1 as a therapeutic target.
F.2. The substrate specificity of the catalytic subunit.
All members of the PPP superfamily (PP1, PP2A (PP2), PP2B (PP3) and PP4-7) have
catalytic cores that share the same structural fold and catalytic mechanism189.
Differences between these enzymes reside mainly in the solvent-exposed loops that
determine the shape and charge of the surface, and hence the affinity for ligands. For
example, the catalytic site of PP1 is conspicuously surrounded by acidic residues
26,70
.
This feature likely explains why PP1 dephosphorylates the β-subunit of phosphorylase
201
kinase much faster than the more acidic α-subunit, a property that has been widely used
to biochemically differentiate PP1 from other protein Ser/Thr phosphatases
190
. Another
unique feature of PP1 is that it poorly dephosphorylates short peptides modeled after its
physiological substrates, demonstrating that, unlike many protein kinases, PP1 does not
recognize a consensus sequence surrounding the phosphorylated residue. Instead,
efficient substrate binding depends on docking motifs for PP1 surface grooves that are
remote from the active site. Under controlled buffer conditions, the free PP1 catalytic
subunit has an exceptionally broad substrate specificity. Bacterially expressed
mammalian PP1 even acts as a protein tyrosine phosphatase and can dephosphorylate
small molecules such as p-nitrophenylphosphate
70
. However, the catalytic subunit that
is purified from mammalian tissues does not act on these atypical substrates. This
finding suggests that the native enzyme is more selective, possibly because the metals
that are incorporated in the
active site (Fe2+ and Zn2+)
differ from those of the
bacterially
expressed
enzyme (Mn2+) (Figure 87).
Mammalian
Figure 87. Surface representation of the structure of
PP1α (PDB ID 1FJM).
Surface representation of the structure of PP1a (PDB ID
1FJM). (a) The active site of PP1 (green) contains two metal
ions (pink spheres) and lies at the Y-shaped intersection of
three substrate-binding grooves; the hydrophobic (blue), the
acidic (orange), and the C-terminal (red). (b) A 130o rotation
of a, showing binding sites for the PP1-docking motifs RVxF
(purple), SILK (cyan) and MyPhoNE (wheat).
genomes
contain three PP1-encoding
genes that together encode
four
distinct
catalytic
subunits:
PP1α,
PP1β/δ
and
splice
variants
the
PP1γ1 and PP1γ21,25,34, which differ mainly in their extremities. However, the free PP1
isoforms exhibit a similarly broad substrate specificity.
202
Physiological substrates of PP1 can be classified into two groups on the basis of their
affinity for the catalytic subunit. Some substrates (e.g. the tumor suppressor BRCA1)
have high-affinity docking sites for PP1 and form stable heterodimeric complexes with
the phosphatase even when they are dephosphorylated (see Table 15). By contrast,
other substrates (e.g. glycogen phosphorylase) establish only weak interactions with the
catalytic
subunit,
as
suggested
by
the
nearly
complete
inhibition
of
their
dephosphorylation by physiological salt concentrations and their inability to form stable
complexes with PP1
80
. The efficient in vivo dephosphorylation of the latter substrates
requires PIPs that provide additional substrate docking sites or increase the local
substrate
concentration
by
tethering
the
phosphatase
to
substrate-containing
compartments. Thus, substrate selection by PP1 clearly depends on phosphatase
docking motifs and subcellular targeting subunits which, together, constitute the
substrate-targeting and -specifying toolkit of PP1.
F.3. The PP1 protein interactome.
Most proteins interact with a limited number of ligands, but a small proportion of proteins,
termed hubs, have many partners
191
. Party hubs interact with many of their ligands
simultaneously, whereas date hubs bind their distinct partners at different times or
locations. PP1 isozymes can be classified as date hubs because they form stable
complexes with numerous proteins but only a few proteins can interact simultaneously
(see Table 15). PP1-interacting proteins were originally identified using classical
biochemical approaches as well as yeast two-hybrid screens. More recently, in silico
screenings based on stringent definitions of the five-residue RVxF-type PP1-docking
motif, combined with a biochemical validation procedure, have led to a near doubling of
the PP1 interactome
33,177
. Novel PP1 complexes also have been identified by affinity
chromatography with covalently bound microcystin-LR, a potent small-molecule inhibitor
203
of PPP phosphatases, in combination with the selective elution of PP1-bound proteins
by competition with a synthetic RVxF-type docking peptide
complexes has been identified using antibody arrays
192
. Yet another set of PP1
193
. However, these latter two
approaches do not differentiate between direct and indirect interaction partners. At
present, approximately 180 mammalian genes are known to encode direct PP1
interactors (see Table 15), but the real number is almost certainly much higher. Indeed,
a bioinformatics-assisted screen recovered only about one-third of the previously known
mammalian PIPs with an RVxF motif
33
, indicating that about 450 genes, instead of the
currently validated 150, are likely to encode this type of PIP. In addition, unbiased
screens suggest that
30% of PIPs do not have a functional RVxF motif. Collectively,
these data suggest that PP1 forms stable complexes with as many as 650 mammalian
proteins.
F.4. Mutual control of PP1 and PIPs.
Although only a minority of PP1 complexes have been functionally analyzed, some
recurring themes have emerged. Some PIPs are PP1 substrates, and their controlled
dephosphorylation serves a regulatory function. Conversely, many PIPs control PP1 by
acting as substrate-targeting subunits or inhibitors. Interestingly, a subset of PIPs are
both substrates and regulators for associated PP1.
F.5. Substrates.
More than a dozen vertebrate PIPs have been identified as PP1 substrates (see Table
15). Some of these substrate PIPs are selectively dephosphorylated on a single site,
whereas others are dephosphorylated rather indiscriminately at multiple positions
and
194 195
,
196
. Nearly half of the known substrate PIPs are enzymes. They are often activated
by dephosphorylation, as is the case for BRCA1, an E3 ubiquitin ligase, focal adhesion
204
kinase (FAK), the protein phosphatase CDC25C and caspase 2
195-198
. By contrast, PP1
maintains the associated protein kinases NEK2 and Aurora-A in an inactive state
Dephosphorylation by PP1 stabilizes the transcription factor Ikaros
200
199
.
and regulates the
binding of ligands to various PIPs201-204.
Some substrate-PIPs also regulate PP1 function. Inhibitor-2 and the protein
kinase-C potentiated inhibitor (CPI-17) are both substrates and potent inhibitors of PP1
62,194
and protein kinase NEK2 as well as the membrane-targeting protein TIMAP are
dephosphorylated by PP1 but also target other proteins in the complex for PP1-mediated
dephosphorylation
205,206
. Thus, PP1 has diverse effects on substrate PIPs, with some of
these substrates functioning as PP1 regulators. One can envisage that the stable
association between PP1 and a subset of its substrates has facilitated the evolution of
such a reciprocal relationship, although it cannot be excluded that some PIPs originated
as PP1 regulators and became substrates only later in evolution.
F.6. Substrate-targeting proteins.
Many PIPs contain specific domains that mediate the binding of PP1 to specific cellular
compartments or macromolecular complexes (see Table 15). Indeed, PIPs can target
PP1 to such diverse structures as the plasma membrane (e.g. integrin αIIB),
mitochondria (e.g. URI), endoplasmic reticulum (e.g. the stress-induced protein
GADD34), glycogen particles (e.g. G-subunits), the actin cytoskeleton (e.g. spinophilin,
also called neurabin-2), chromatin (e.g. Repo-man) and nucleoli (e.g. NOM1). The
targeting by PIPs brings PP1 into close proximity to specific subsets of substrates; the
associated increased local substrate concentration is sufficient to increase the
dephosphorylation rate by up to several orders of magnitude 207.
205
The concept of multi-targeting
emerges from the ability of
some PIPs to target PP1 to
various
signaling
complexes,
and hence to function as signal
integrators. Multi-targeting can
be explained by the presence of
multiple targeting domains or by
competition between substrates
for
a
single,
multifunctional
targeting domain. Among the
best studied PIPs that harbor
Figure 88. PP1-interacting proteins with multiple
substrate-targeting domains.
The figure shows selected PIPs that can act as signal
integrators because they have multiple substratetargeting domains. (a) GADD34, (b) MYPT1, (c) PNUTS
and (d) spinophilin target PP1 (blue) either directly to its
substrates (pink squares) or via adaptor proteins (green
ovals). Some of the substrates (LCP1, TRF2,
GABACR1 and GPCRs) are still hypothetical.
Abbreviations: ER, endoplasmic reticulum; RB,
retinoblastoma protein; ActinBP, actin-binding proteins
(including myosin, merlin and moesin); GPCRs, G
protein-coupled receptors; Glu-Rs, glutamate receptors;
RYR, ryanodine receptors.
multiple targeting domains are
GADD34,
the
myosin
phosphatase targeting subunit
MYPT1,
spinophilin
PNUTS,
(Figure
and
88).
GADD34 promotes the PP1mediated dephosphorylation of
the translation regulator eIF2α, the GTP-regulatory proteins TSC1 and TSC2, TGFβ
receptor 1 as well as I-κB kinase (IKK)
208, 209, 210
and
211
. The respective targeting to the
latter two substrates is mediated by the GADD34-binding adaptor proteins SMAD7 and
CUEDC2. MYPT1 contains binding sites for the retinoblastoma protein, polo-like kinase1, and a number of actin-binding proteins, including myosin, merlin and moesin, which
are all PP1 substrates
. PNUTS targets PP1 to the γ-aminobutyric acid receptor,
212-214
transcription factor LCP1 and telomeric protein TRF2
215-217
. In addition to its actin-
206
binding domain, spinophilin contains three other substrate-targeting domains: an
interaction domain for G protein-coupled receptors, a PDZ domain that recruits soluble
(e.g. the ribosomal protein kinase p70S6K) and integral membrane proteins (e.g.
glutamate and ryanodine receptors) and a C-terminal coiled-coil domain that mediates
the binding to doublecortin
46,52,72,218
. NIPP1 is a targeting PIP with a single,
multifunctional substrate-targeting domain
219,220
. It possesses a phosphothreonine-
binding forkhead-associated domain that binds various putative PP1 substrates involved
in transcription, pre-mRNA splicing and cell-cycle regulation.
F.7. Substrate-specifiers and inhibitors.
More than half of all PIPs inhibit PP1 when glycogen phosphorylase is used as a
substrate
33
. Most of these PIPs are poor inhibitors, but some substrate and targeting
PIPs, including GADD34
66
, the neurabins
182
, PNUTS
221
and NIPP1
40
, are inhibitory in
the low nanomolar range. Nevertheless, these proteins are better defined as substrate
specifiers than as inhibitors, because they selectively inhibit the dephosphorylation of
only a subset of substrates, including glycogen phosphorylase. In addition, at least some
of these PIPs have an opposite effect, enhancing specific activity towards the PP1
physiological substrates, as has been best illustrated for the glycogen targeting Gsubunits and MYPT1 34,71.
Nonetheless, some PIPs are true PP1 inhibitors because they block access to
the active site and inhibit the dephosphorylation of all substrates. These PIPs constitute
the PP1 inhibitory toolkit. Some PIPs, including Inhibitor-1, CPI-17 and their paralogues
are inhibitory only when phosphorylated, functioning as pseudosubstrates
25,194
.
Similarly, a phosphorylated domain of the targeting PIP MYPT1 acts as a
pseudosubstrate inhibitor
222
. The inhibitory activity of other PIPs, including Inhibitor-2
207
and Inhibitor-3, does not require prior phosphorylation. These inhibitors often function as
a second or third non-catalytic subunit of PP1 complexes. However, there is some
variation in the architecture of these holoenzymes. In vivo, Inhibitor-3 and CPI-17
paralogues appear to be holoenzyme-specific, as they are present in PP1 complexes
containing SDS22 and MYPT1, respectively
67,194
. In these complexes, the targeting and
inhibitory subunits have non-overlapping PP1-docking sites. By contrast, Inhibitor-2
functions as an inhibitory subunit of various PP1 holoenzymes, including complexes
containing NEK2, spinophilin and Aurora-A
65,223
. Because each of these targeting PIPs,
as well as Inhibitor-2, has a functional RVxF motif as one of several PP1-docking sites,
they must compete for the RVxF-binding groove within the complex. Although a similar
reasoning applies to the complex of PP1 containing GADD34 and Inhibitor-1, this
complex is stabilized additionally by interactions between GADD34 and Inhibitor-1 66.
F.8. PP1-docking motifs.
Most PIPs contain four to eight residue docking sequences that combine to create a
large interaction surface for PP1. On average, a docking motif occupies
total PP1 surface, creating an
850 Å2 interaction surface
425 Å2 of the
35,43,58,62
. Assuming that the
entire surface is involved in the binding of PIPs, PP1 has up to 30 non-overlapping PIPbinding sites, much fewer than the number of distinct PIPs (see Table 15). Therefore,
PIPs must share PP1-docking motifs, a conclusion that is supported by substantial
experimental evidence. However, PIPs differ in the number and the type of their PP1docking sites, thus enabling a combinatorial control of PP1
71
. In addition, PP1-docking
motifs are degenerate and sequence variants show considerable differences in their
affinity for PP1. These structural insights suggest that small-molecule compounds that
compete with specific (combinations of) docking motifs for binding to PP1 can be used to
208
functionally disrupt subsets of PP1 holoenzymes and may have a therapeutic potential
(Box 1).
Box 1. The therapeutic potential of PP1.
In recent years, protein kinases have become highly successful drug targets, mainly for the
79
treatment of cancer . As most phosphorylations are reversible, protein phosphatases are
equally powerful drug targets to interfere with protein phosphorylation. This is impressively
illustrated by the PP2B inhibitors cyclosporin A and FK506, which are clinically used as potent
immunosuppressants. Clearly, PP1 inhibitors hold great promise for the treatment of various
human pathologies, including cancer, neurodegenerative diseases, type 2 diabetes, heart
48,74,80-82
failure and viral diseases
. However, highly specific cell-permeating inhibitors for the
PP1 catalytic subunit are not yet available and it is questionable whether such agents could
78
ever be used therapeutically, as they would be likely to inhibit all PP1 holoenzymes . A more
selective approach, inspired by recently acquired structural insights, involves the functional
disruption of subsets of PP1 holoenzymes with small-molecule compounds that bind to PIP
interaction sites on PP1, such as the hydrophobic binding grooves for the RVxF, SILK and
MyPhoNE sequences. Blocking the less prevalent SILK and MyPhoNE motifs will affect
smaller subsets of PP1 holoenzymes; however, even compounds that interfere with the
docking of RVxF sequences can provide greater selectivity than predicted from the
abundance of this motif, as its importance is holoenzyme-dependent. Moreover, RVxF
competing agents can be used at concentrations that disrupt only the binding of low-affinity
RVxF variants. Another PP1 targeting strategy aims to interfere with substrate recruitment at
extended docking sites of specific holoenzymes. At the very least, inhibitors of subsets of PP1
holoenzymes could be employed for functional studies.
The primary PP1-docking motif, commonly referred to as the RVxF motif, is
present in about 70% of all PIPs. It generally conforms with the consensus sequence
K/R K/R V/I x F/W, where x is any residue other than Phe, Ile, Met, Tyr, Asp, or Pro (see
Table 15)
33
. Often, this motif is flanked N-terminally by basic residues and C-terminally
by acidic residues. The RVxF sequence binds in an extended conformation to a PP1
hydrophobic groove that is 20 Å away from the active site (Figure 87). Binding of this
motif does not change the PP1 conformation and functions only to anchor the PIPs to
PP1
35,43,58,62,142
. However, this binding event is essential, as it brings PP1 into close
proximity with its PIPs and promotes secondary interactions that contribute to PP1
isoform selection and determines the activity and substrate specificity of the holoenzyme
58,62,71,182
. The contribution of the RVxF motif to the binding of PP1 is interactor-
209
dependent; although it is essential for the binding of many PIPs, some PIPs still interact
strongly with PP1 in the presence of an excess of a synthetic RVxF peptide or despite
alterations in their RVxF motif
71
. Surprisingly, some established RVxF variants, such as
the KSQKW sequence of Inhibitor-2, deviate considerably from the consensus RVxF
sequence 62. Thus, the true diversity of RVxF motifs remains elusive, suggesting that the
PP1 interactome could be even larger than currently estimated.
A PP1-docking sequence that is present in seven of the known vertebrate PP1
interactors is the so-called SILK-motif (see Table 15), which contains the consensus
sequence G/S I L R/K 33. Always positioned N-terminal to the RVxF sequence, it binds in
a hydrophobic groove on the opposite face of the PP1 active site (Figure 87). Identical
with the RVxF motif, it does not change the conformation of PP1, but instead fulfills an
anchoring function
33,62,142
. MYPT1 contains an N-terminal PP1 interaction motif that
adheres to the consensus sequence R x x Q V/I/L K/R x Y/W, where x can be any
residue
33,43
. This motif, referred to as the myosin phosphatase N-terminal element or
MyPhoNE, is also present in six other PIPs, again always N-terminal to the RVxF
sequence (see Table 15). The MyPhoNE motif of MYPT1 lies within a five-turn α-helix,
which faces hydrophobic residues in a shallow hydrophobic cleft on PP1 (Figure 87),
and contributes to substrate selection.
Some PIPs, including MYPT1
43
and the neurabins
182
, interact with PP1 in an
isoform-dependent manner, suggesting that they possess isoform-specific docking sites.
Because the PP1 isoforms differ mainly at the N- and C-termini, these represent obvious
binding places for specific docking sequences. This is certainly the case for MYPT1,
which contains eight tandem ankyrin repeats that interact with the PP1β/δ C-terminus.
The interaction centers around two tyrosine residues that are exclusive to the PP1β/δ
isoform
43
. However, recent mutagenesis studies suggest that the N-terminal MyPhoNE
210
motif might also contribute to isoform selection 224. Surprisingly, neurabins do not interact
with PP1 N- or C-termini; thus, the basis of their isoform selectivity is unclear
58
.
Although an RVxF flanking sequence in spinophilin has been identified as an auxiliary
PP1γ1-selectivity determinant
182,225
, the available structural data have not disclosed an
underlying mechanism 58.
F.9. Structural insights into PP1–PIP interactions.
In their unbound form, full-length PIPs or their PP1-interacting domains often do not fold
into
a
typical
three-dimensional
protein
structure,
hampering
their
structural
characterization. Nonetheless, the few available structures have yielded crucial insights
into substrate selection and inhibition.
F.10. PIPs as intrinsically disordered proteins.
Recent structural studies identified a subset of PIPs, including DARPP-32, Inhibitor-2
and spinophilin, that are highly disordered in their unbound state and do not assume a
well defined three-dimensional structure
57,58
. These findings place them in the group of
intrinsically unstructured (IUPs) or disordered (IDPs) proteins (Box 2). Their intrinsic
flexibility enables them to form extensive and unique interactions with PP1 through
several docking motifs, as illustrated by crystal structures of PP1 complexes
(43,58,62,Figure 89). However, there are distinct differences between the unbound forms of
these intrinsically disordered PIPs. Detailed analysis has shown that DARPP-32,
Inhibitor-2 and spinophilin display different degrees of unstructured character in solution.
Whereas the spinophilin PP1-interacting domain lacks any preferred secondary or
tertiary structure, distinct preferred conformations can be identified in DARPP-32 and
Inhibitor-2
57,58
. The structural preferences of PIPs in their unbound form could help to
drive the formation of complexes with PP1 through conformational selection. The
211
flexibility of these proteins might play a larger role than simply enabling extensive
interactions surfaces with PP1, as
70% of Inhibitor-2 remains flexible in the PP1-
bound state. In this manner, the residual flexible region might form additional or new
binding sites for other proteins.
Box 2. Intrinsically disordered proteins.
A classic dogma of biochemistry is that protein function is correlated directly to threedimensional structure. However, in recent years it has become apparent that
25% of all
83
eukaryotic proteins contain significant regions of disorder . These proteins are referred to as
IUPs or IDPs. The intrinsic disorder can include the entire protein, an unstructured region next
to one or multiple structured domains, or a long (>50 residues) flexible linker between
structured domains. Unstructured regions can be identified using bioinformatic tools, as they
are enriched in certain amino acids, particularly Asn, Gln, Ser, Thr, Gly, Ala and Pro.
Furthermore, they typically contain few hydrophobic residues, e.g. Trp, Val, Leu, Ile, Phe, and
Tyr, which usually make up the core of folded proteins. The exceedingly dynamic nature of the
IDP prevents its description as a single, rigid structure.
84
IDPs have multiple functional advantages over structured proteins . First, they can interact
with their binding partners, often structured proteins, in highly extended conformations. These
binding events can result in novel, often unexpected interaction surfaces that are two- to
fourfold larger than commonly seen for interactions between two folded proteins. Therefore,
IDPs require fewer residues than structured proteins to bind their targets with an identical
surface area. This has been proposed as a mechanism to reduce cell crowding. Second, their
intrinsic flexibility enables a single IDP to bind multiple target proteins, as the flexibility allows
for adaptation to different protein surfaces. Third, IDPs are thought to have extensive capture
radii for their target proteins, allowing them to initiate protein–protein interactions more
efficiently than their folded counterparts.
IDPs can interact with structured binding partners in different modes. In some instances the
85
binding reaction is coupled to IDP folding . Alternatively, the binding reaction can proceed
through conformational selection, where certain IDP conformers that contain structural
preferences resembling the bound state are selected during the binding interaction. Lastly,
IDPs can retain flexibility, even when bound to one or multiple target proteins, allowing them
to have crucial biological roles as structural scaffolds.
So, how many PIPs have an intrinsically disordered PP1-interaction domain? A
bioinformatics analysis using the IUPRED program
100
, which scores for the occurrence
of disorder-inducing amino acids that are typically enriched in IDPs (Box 2), predicts that
the PP1 interaction domain of about two-thirds of all RVxF-type PIPs is disordered in a
212
Figure 89. Crystal structures of protein-PP1 complexes.
The regulatory proteins are shown as pink ribbons and PP1 as a blue surface representation.
Top images are centered around the active site of PP1 whereas the bottom images have
been rotated 100o to highlight the interaction within the RVxF docking motif. (a) The crystal
structure of spinophilin417-583–PP1a7-330 (PDB ID 3EGG). (b) The crystal structure of
MYPT11-299–PP1d (PDB ID 1S70). (c) The crystal structure of Inhibitor-2–PP1g (PDB ID
2O8A). These PP1 interacting proteins share only the common RVxF motif interaction. All
other interactions are unique for each holoenzyme.
region of at least 100 residues (see Table 15). In nearly half of these IDPs, the
unstructured region flanks both sides of the RVxF sequence. It seems likely that the high
percentage of intrinsically disordered PIPs is crucial for the creation of extensive
interaction areas upon PP1 binding and thus for the formation of unique PP1
holoenzymes.
F.11. Structural basis of substrate selection.
213
PP1 contains three potential substrate-binding grooves: the hydrophobic, the acidic and
the C-terminal groove. These grooves form a Y-shape surface that intersects at the PP1
active site (Figure 87). The recently described structure of the spinophilin–PP1 complex
revealed that spinophilin blocks the C-terminal substrate-binding groove (58, Figure 89a).
Follow-up mutational and biochemical analysis showed that this interaction plays a
crucial role in substrate selection by the spinophilin–PP1 complex. Owing to the
obstruction of the C-terminal substrate groove, the activity of the spinophilin–PP1
complex is restricted to specific substrates. Thus, this mode of regulation allows PP1 to
dephosphorylate substrates that exclusively bind the acidic and the hydrophobic
grooves, while blocking the dephosphorylation of those that require interaction with
residues in the C-terminal groove.
At first glance, a similar substrate-binding groove modification cannot be
identified in the MYPT1–PP1 structure (Figure 89b). However, C-terminal to the RVxF
interaction site, MYPT1 has eight tandem ankyrin repeats that interact with the PP1 Cterminus, extending the PP1 acidic substrate-binding groove, which might play a crucial
role in the recruitment of MYPT1–PP1 holoenzyme substrates. However, a biochemical
analysis suggests that the MYPT1 N-terminal domain, comprising the MyPhoNE motif,
participates in the positive selection of substrates 43.
F.12. Structural basis of PP1 inhibition.
PP1 is potently inhibited by some small-molecule toxins, including microcystin-LR,
nodularin-R and tautomycin, which all block its active site
26
and
27
. CPI-17, an inhibitor
of the MYPT1–PP1 complex, is a structured protein and must be phosphorylated to
become a potent inhibitor
194
. Recent findings show that CPI-17 undergoes a significant
structural modification upon phosphorylation, which allows the phosphorylated residue to
become exposed to the surface and primed for inhibition of the holoenzyme 226. Inhibitor-
214
1 and its paralogues are also phosphorylation-dependent inhibitors and, although
structural information is lacking, it is apparent that its inhibitory activity requires the
phosphorylated residue to be pushed into the active site of PP1. By contrast, Inhibitor-2
does not require phosphorylation to inhibit PP1. The structure of the PP1–Inhibitor-2
complex revealed three crucial interaction sites
62
, involving an RVxF motif, a SILK-
sequence and a long, kinked α-helix of Inhibitor-2 (Figure 89c). This α-helix binds the
acidic and hydrophobic substrate-binding grooves and, in doing so, covers the active
site, preventing all PP1 activity. In addition, although Inhibitor-2 binding does not induce
a conformational change of PP1, it does trigger the release of one of two metals that are
essential for catalysis.
F.12. Enhanced selectivity through regulation of PIPs.
The exquisite specificity of PP1 in vivo is explained by the structural design and diversity
of its toolkit, and by various regulatory mechanisms that impinge on PIPs (Figure 90).
Some PIPs are expressed in a cell type-dependent manner, accounting for cell typespecific PP1 activity1,
32
and 25. Recent data show that the concentration of several PIPs
is controlled by regulated proteolysis
227
,
228
,
229
and
230
. Moreover, many signaling
pathways interfere with the affinity of specific PIPs for PP1. For example,
phosphorylation of Ser/Thr residues in or near RVxF-type docking sequences is often
associated with a reduced binding affinity for the RVxF-binding channel 71. Signaling can
also result in the recruitment or release of inhibitory PIPs
25,194,231
. Another PIP control
mechanism involves positive or negative allosteric regulation by metabolites or other
proteins
232
and
233
. The PP1-mediated dephosphorylation of some substrate-PIPs is
restrained through regulated masking of the phosphorylated residues by 14-3-3 proteins
195 198 234 235
,
,
,
and 236. Finally, PP1–PIP complexes are highly dynamic and different PIPs
215
compete for the same PP1 binding sites
235
and
236
. Ultimately, the concentration and
PP1-binding affinities of PIPs determines which PP1 holoenzymes are formed.
F.13. Concluding remarks and future perspectives.
The view of PP1 as an imprecise housekeeping enzyme, which was based largely on
assays with the purified catalytic subunit, is no longer tenable. Indeed, PP1 is a highly
specific and regulated phosphatase, owing to the unusual diversity and structural design
of its regulatory toolkit. There could be as many distinct PP1 complexes as there are
Figure 90. PIP-mediated regulation of PP1 holoenzymes.
The figure gives an overview of regulatory mechanisms that affect PIPs, and hence PP1
function. (i) Controlled proteolysis of PIPs. (ii) Dissociation of holoenzymes by the
phosphorylation of PIPs. (iii) Recruitment or dissociation of inhibitory PIPs. (iv) Allosteric
regulation by thebinding of metabolites or proteins to PIPs. (v) Substrate masking by 14-33 protein binding. (vi) Competition between PIPs for the same binding sites on PP1.
Abbreviations: AR, allosteric regulator; I, inhibitor; P, phosphate.
protein Ser/Thr kinases, suggesting that both types of enzymes have a similarly
restricted substrate specificity at the holoenzyme level. Clearly, much more work is
required to uncover the true diversity of the PP1 interactome. In the coming years, the
216
determination of additional atomic resolution three-dimensional structures of PP1
complexes should yield much needed novel insight into the PIP–PP1 interaction and
substrate selection mechanisms. This will be a daunting task because many PIPs are
intrinsically disordered, rendering crystallization of these holoenzymes extremely difficult.
However, as only structural data will reveal how PP1 can be developed as a therapeutic
target by interfering with its binding to PIPs and substrates, it is a goal worth pursuing.
217
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+6
+
+
+
+
+
+
+
+
+
+
+
+
+
-
+
+
+
+
+ T
+ T
- T
+ ?
+ S
- ?
S
- T
- T
- T
+ S
- ?
+ ?
S
S
- ?
+ ?
+ ?
+ ?
S
- ?
+ ?
+ ?
+ ?
+ ?
+ ?
I
+ ?
7
+ I
+ ?
+ S
+ ?
- ?
+ ?
+ ?
+ ?
- ?
- ?
+ T
- ?
S
+ ?
+ ?
- ?
+ ?
+ ?
- ?
?
+ T
+ I
- T
+ T
+ ?
- ?
+ ?
I
237
238
239
177
240
241
202
242
243
243
244
33
33
245
198
33
33
33
33
195
33
33
33
33
177
33
246
33
158
33
247
33
33
33
248
33
33
33
249
250
197
33
33
251
33
33
143
252
253
254
255
256
33
33
33
257
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+7
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
T
?
?
?
?
S
I
I, S
I
T
?
?
I
S
?
?
?
I
?
T?
?
?
?
T
?
?
?
?
?
?
?
?
?
?
?
T
T
T
?
T
S
S,T
T
T?
?
T
?
S
T
S
?
?
?
?
T
S
Reference
Function4
PPP1R3F
HCF1
HDAC6
HYDIN
C11orf66
IKZF1
PPP1R1A
PPP1R2
PPP1R11
ITGA2B
ITPR1
ITPR3
PPP1R1C
AHCYL1
JARID1B
KCNA6
KCNK10
PPP1R14C
KIAA1244
PPP1R3E
LMTK2
RPL5
KIAA0430
AATK
LMTK3
C7orf47
C14orf50
LRRC68
MAP1B
MCM7
GRM1
GRM5
GRM7
MKI67
MPHOSPH10
PPP1R12A
PPP1R12B
PPP1R16A
MYO1D
MYO16
NCOR1
NEK2
PPP1R9A
NEFL
KIAA1543
SLC9A1
PPP1R8
NOC2L
SLC12A2
NOM1
OCLN
OPN3
ORC5L
PPP1R13A
PPP1R13B
PPP1R12C
IDP3
HB2E
HCF1
HDAC6
HYDIN
IIIG9
Ikaros
Inhibitor-1
Inhibitor-2
Inhibitor-3
integrin αIIB
IP3R1
IP3R3
IPP5
IRBIT
JARID1B
KCNA6
KCNK10
KEPI
KIAA1244
KIAA1443
KPI-2
L5
LIMKAIN b1
LMTK1
LMTK3
LOC221908
LOC145376
LRRC68
MAP1B
MCM7
mGlu1
mGlu5
mGlu7
MKI67
MPHOSPH10
MYPT 1
MYPT 2
MYPT 3
myosin1D
MYR 8
N-Cor
NEK2a
neurabin-I
neurofilament L
NEZHa2
NHE1
NIPP1
NIR
NKCCl
NOM1
occludin
opsin 3
ORC5L
p53BP2
p53BP2like
p84
MyPhoNE
Gene
SILK
Protein
RVxF
Reference
Function4
AKAP1
AKAP11
AKAP9
APC
AURKA
AURKB
AXIN1
BCL2
BCL2L2
BCL2L1
BRCA1
CASC1
CASC5
CASP9
CASP2
DLG2
CCDC8
CCDC128
CD2BP2
CDC25C
CENPE
CEP192
CHCHD3
CHCHD6
CLCN7
CNST
PPP1R14A
CSMD1
PPP1R1B
DDX31
POLD3
KIAA0649
DYSFIP1
DZIP3
EIF2S2
ELFN1
ELFN2
ELL
ZFYVE16
SH3GLB1
PTK2
FAM130A1
FAM130A2
FER
FARP1
FKBP15
PPP1R15B
YWHAG
PPP1R15A
PPP1R14D
PPP1R3B
PPP1R3A
GPATCH2
GPR12
GRXCR1
C7orf16
Docking motif
IDP3
AKAP149
AKAP220
AKAP450
APC
Aurora-A
Aurora-B
AXIN
BCL2
BCL-w
BCL-x
BRCA1
CASC1
CASC5
caspase 9
caspase 25
chapsyn110
CCDC8
CCDC128
CD2BP2
CDC25C5
CENPE
CEP192
CHCHD3
CHCHD6
CLC7
Consortin
CPI-17
CSMD1
DARPP32
dead box31
DNApolIIp68
DRIM BP
DYSFIP1
DZIP3
eIF2β
ELFN1
ELFN2
ELL1
Endofin
endophilin B1t
FAK
FAM130A1
FAM130A2
FER kinase
FERM
FK506BP15
FLJ14744
14-3-3gamma
GADD34
GBPI-1
GL
GM
GPATCH2
GPR12
glutaredoxin
G-substrate
MyPhoNE
Gene
RVxF
Protein
SILK
Docking motif
143
258
259
33
33
200
260
62
261
262
177
177
263
201
33
177
33
264
33
143
265
266
33
267
33
33
33
33
33
33
268
268
268
33
33
43
269
270
33
271
272
205
273
274
33
275
40
258
276
37
277
177
33
278
143
279
218
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
280
281
282
283
33
284
33
285
33
286
33
287
288
33
289
290
291
292
177
33
293
33
33
33
294
33
210
295
281
296
33
33
33
+
+6
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+7
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
-
?
?
?
?
T
?
?
?
S
?
?
?
S
?
T
?
?
?
?
?
?
?
?
?
T
?
?
?
?
T
?
?
?
?
Reference5
WBP11
SMARCB1
SLC7A14
SPATA2
PPPR9B
SPOCD1
SPRED1
SPZ1
SFRS13
STAU
DLG3
SYTL2
MAPT
TNS1
PPP1R16B
TMEM132C
TMEM132D
TRA2B
TRIM42
TRPC4AP
TRPC5
TSC2
TSKS
UBN1
C19orf2
VDR
VPS54
WDR81
WNK1
YLPM1
ZBTB38
ZCCHC9
ZFYVE1
ZSWIM3
Function4
SIPP1
SNF5
solute carrier 7-14
SPATA2
spinophilin
SPOCD1
SPRED1
SPZ1
SRp38
staufen
SAP102
SYTL2
TAU
tensin 1
TIMAP
TMEM132C
TMEM132D
TRA-2BETA
TRIM42
TRPC4AP
TRP5
TSC2
TSKS
ubinuclein 1
URI
vitamin D receptor
VPS54
WDR81
WNK1
ZAP3
ZBTB38
ZCCHC9
ZFYVE1
ZSWIM3
33
IDP3
203
MyPhoNE
Gene
SILK
+
+
+
+
S
?
?
?
I
T
?
T
?
S
?
T
?
?
?
?
?
T
T
S
?
?
T
?
?
?
S
?
T
?
?
?
?
?
?
Protein
RVxF
+
-
Reference5
+
+
+6
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
Function4
PARD3
PCIF1
PFKM
PHACTR4
PPP1R14B
KIAA1949
PHRF1
ANKRD28
PKMYT1
EIF2AK2
PMP22CD
PPP1R10
DEPDC2
PLCL1
PCDH7
PCDH11X
SFPQ
PPP1R3C
PPP1R3D
RB1
RB1CC1
RBM26
CDCA2
RIMBP2
RPGRIP1L
RRP1B
RYR1
SACS
ZFYVE9
ANKDR42
PHACTR3
PPP1R7
SFI1
SH2D4A
SH3RF2
Docking motif
IDP3
PAR-3
PCIF1
PFK-1
PHACTR1-4
PHI-1
phostensin
PHRF1
PITK
PKMYT1
PKR
PMP22cd
PNUTS
pREX2
PRIP-1
protocadherin 7
protocadherin11X
PSF
PTG
R6
RB
RB1CC1
RBM26
Repo-man
RIMBP2
RPGRIP1L
RRP1B
ryanodine recept.
SACSIN
SARA
SARP
scapinin
SDS22
SFI1
SH2D4A
SH3RF2
MyPhoNE
Gene
RVxF
Protein
SILK
Docking motif
297
298
33
33
42
33
33
299
234
300
33
33
301
302
303
33
33
304
33
33
177
33
33
33
305
306
33
33
33
307
33
33
33
33
Table 15. Bollen, Table S1.
Validated vertebrate PIPs1,2.
1
Several additional vertebrate proteins have been found to co-purify or co-immunoprecipitate with
PP1 but the available data do not allow distinguishing between direct and indirect interactors.
These include the proteins ALK1, androgen receptor, C9orf75, caveolin, clathrin light chain b,
EWS, GRP78, HDAC1, HDAC8, HOX11, integrin α3A, LCP1, MEF2A, myopalladin, NCAM,
PDE6B, RIF1, SUR8, SRC3, TERA and TMEM113.
2
The list excludes paralogues of known PP1 interactors with a conserved RVxF motif that have
308
not yet been validated as PIPs. A list of these paralogues is given in .
3
Taking the available structures and biophysical data of unbound and bound PP1 regulatory
proteins as a boundary blueprint, an average score from IUPRED (ASI) for ± 100 residues
surrounding the F/W of the RVxF/W motif was calculated. PIPs with ASI values > 0.45 were
considered to be disordered (+); PIPs with ASI values < 0.45 were marked as structured (-).
4
Abbreviations. I, inhibitory PIP; IDP, intrinsically disordered protein; PIP, PP1-interacting protein;
S, substrate PIP; T, targeting PIP; ?, unknown
5
The validated PP1-docking motif present in the Xenopus or mouse isoforms is not conserved in
humans.
6
PIP isoforms with a conserved RVxF motif that have not yet been validated.
7
PIPs that were experimentally confirmed as IDPs.
219
Appendix G. Structural characterization of the neurabin
sterile alpha motif domain.
220
221
Reproduced from the Proteins, Volume 69, Number 1, October 2007 pages 192-198.
The expression, purification and characterization of the Neurabin coiled-coil domain was
performed by M.J.R.
Tingting Ju1, Michael J. Ragusa2, Jebecka Hudak1, Angus C Nairn3, & Wolfgang Peti1
1
Department of Molecular Pharmacology, Physiology and Biotechnology and
2
Department of Molecular Biology, Cell Biology and Biochemistry, Brown University,
Providence, Rhode Island, USA. 3Department of Psychiatry, Yale University School of
Medicine, New Haven, Connecticut, USA.
Correspondence should be addressed to W.P. (wolfgang_peti@brown.edu).
222
G.1. Introduction.
Neurabin36 (SWISS-PROT/TrEMBL O35867) and its isoform spinophilin14,42 (SWISSPROT/TrEMBL O35274) are neuronal scaffolding proteins that are critical for synaptic
transmission and synaptic plasticity47,171. They are highly enriched in dendritic spines,
the site of excitatory neurotransmission, because of their ability to interact with F-actin in
this subcellular compartment38. Neurabin is expressed exclusively in neuronal cells,
whereas spinophilin is expressed ubiquitously, although it is also highly enriched in
neurons. Spinophilin is sometimes referred to as neurabinII14. Neurabin consists of 1095
residues (MW 122,730 Da), whereas spinophilin is smaller consisting of only 817
residues (MW 89,640 Da).
A recent report by Allen et al.74 showed, using neurabin and spinophilin knockout
mice, that both proteins are necessary for monoamineric signal transduction. Thus,
despite their similarity in domain structure (Figure 91), neurabin and spinophilin have
distinct signaling and regulatory roles. Therefore, it is essential to understand the
structural differences upon which this functional difference is based. As is typical for
scaffolding proteins, both proteins contain multiple common protein interaction domains
(Figure 91). Both neurabin and spinophilin contain an F-actin binding, a functionally
critical Protein Phosphatase 1 (PP1)-binding, a PDZ,52,309 and a C-terminal coiled-coil
domain; sequence similarity within these functional domains can be as high as 86%
(PDZ domains). The F-actin binding domain is important for targeting these proteins to
dendritic spines38,310. The PP1-binding domain of both isoforms targets PP1 towards its
223
substrates, one of which is Ser845 of the GluR1 AMPA receptor subunit. The PDZ
domains of both proteins have
also been reported to bind the
C-termini of p70S6 kinase311
and kalirin-7312. Finally, the
coiled-coil domains have been
implicated in playing a central
role in the homo- and heteromultimerization of these two
proteins with themselves or
Figure 91. Ju, Proteins Figure1.
A) Domain structure of spinophilin (top) and neurabin
(bottom). Sequence identity between the domains is
annotated in the figure. B) Primary sequence alignment
of the neurabin SAM and shank3 SAM domains. The
zinc binding residues in shank3 and the corresponding
residues in neurabin are highlighted by boxes (see text).
The shank3 (M1803E) mutation that is necessary to
achieve high enough protein solubility and reduced
aggregation is underlined. Secondary structure for the
neurabin SAM domain is included above the sequence
alignment.
with coiled-coil domains from
other proteins313,314. The most
significant
domain
these
difference
structure
two
proteins
in
the
between
is
that
neurabin, but not spinophilin,
contains a sterile alpha motif (SAM) domain at its C-terminus (Pfam PF07647). Only in
invertebrates spinophilin can have a SAM domain315. Currently, no functional role has
been identified for the neurabin SAM domain.
In 1995, Ponting316 identified 14 proteins that each contained a conserved
domain of
70 residues. This domain was named SAM because of its presence in yeast
proteins that are essential for sexual differentiation. SAM domains have since been
224
found in more than 1300 proteins317, comparing well with
1600 proteins that have SH2
domains (Src Homology 2), underlying its importance as a functional protein motif. But
unlike SH2 domains, which nearly always bind to phosphorylated tyrosine residues,
there has so far been no common functional theme identified for SAM domains. Instead
SAM domains appear to have a wide diversity of functions. SAM domains can either
form homodimers with themselves or heterodimers with other SAM domains318.
Recently, reports have shown that SAM domains are not only protein:protein interaction
sites but also protein:RNA recognition motifs319,320.
Baron et al.321 demonstrated that the SAM domain of the neuronal scaffolding
protein shank3 also forms homo-polymeric
-helical sheet-like structures and suggested
that the shank3 SAM domain may be essential for the organization of the post synaptic
density (PSD). This polymerization and organization-like behavior of the shank3 SAM
domain is promoted in vitro by the addition of zinc, which is present in the PSD (4.1
nmol/mg of protein)322. In the present study, we have determined the structure of the
neurabin SAM domain by solution state nuclear magnetic resonance (NMR)
spectroscopy. Notably, our results show that the shank3 SAM domain is the closest
structural neighbor of the neurabin SAM domain. We have also examined the
polymerization state of the neurabin SAM domain and find, in contrast to the shank3
SAM domain, that it is a monomer.
G.2. Materials and methods.
G.2.A. Expression and purification of the neurabin SAM domain.
225
After optimal construct screening99, the neurabin SAM domain (residues 986-1056) was
cloned into a his-tagged bacterial expression vector (Thio6-His6-TEV-). This construct
was transformed into the Escherichia coli strain BL21-CodonPlus (DE3)-RIL
(Stratagene) and a single colony was used to inoculate a 100-mL culture of LB
containing kanamycin (50 μg/mL) and chloramphenicol (34 μg/mL). The culture was
grown overnight at 37°C with shaking at 250 rpm. The next morning the cells were
diluted 1:50 into fresh LB medium with appropriate antibiotics. Cells were grown at 37°C
until they reached an OD600 of 0.6-0.9. The cultures were then placed at 4°C and the
shaker temperatures adjusted to 18°C. Cultures were induced with 1 mM isopropyl β-D1-thiogalactopyranoside and allowed to express overnight ( 18 h) at 18°C under
vigorous shaking (250 rpm). Cultures were harvested by centrifugation and the pellets
stored at -80°C until purification. The expression of uniformly
13
C/15N-labeled and
15
N-
labeled protein was carried out by growing freshly transformed cells in M9 minimal
medium containing 4 g/L [13C]-D-glucose and/or 1 g/L
15
NH4Cl as the sole carbon and
nitrogen sources, respectively.
For purification, the pellets were resuspended in lysis buffer (50 mM Tris pH 8.0,
5 mM imidazole, 500 mM NaCl, and 0.1% Triton-X, Complete tabs-EDTA free (Roche)).
The cells were lysed by three passes through a C3 Emulsiflex cell cracker (Avestin) and
the cell debris removed by centrifugation (40,000g/30 min/4°C). The clarified lysate was
filtered through a 0.22-μm membrane (Millipore) and loaded onto a HisTrap HP column
(GE Healthcare) equilibrated with 50 mM Tris pH 8.0, 5 mM imidazole, and 500 mM
NaCl. The protein was eluted with a 5-500 mM imidazole gradient. The fractions
226
containing the target protein were identified by SDS-PAGE gel electrophoresis and were
pooled and dialyzed against 50 mM Tris pH 7.5 and 50 mM NaCl. Tobacco etch virus
(TEV) NIa protease was used to cleave the purification tag. Cleavage, verified by SDSPAGE gel electrophoresis, was achieved overnight at room temperature under steady
rocking. The cleaved sample was then dialyzed against 50 mM Tris pH 7.5 and 250 mM
NaCl for 5 h and further purified by a second his-tag purification step. The cleaved, now
untagged SAM domain was collected in the flow through. The purity of the final sample
was checked by SDS-PAGE gel analysis. The sample was concentrated and finally
exchanged into a buffer containing 20 mM sodium phosphate buffer pH 6.5 and 50 mM
NaCl using an Amicon Ultra centrifugation filter device with a 5000 MWCO (Millipore).
Ten percent D2O and 0.02% NaN3 were added into the sample that reached a final
concentration of
2 mM.
All NMR measurements were performed at 298 K on a Bruker AvanceII 500 MHz
spectrometer using a TCI HCN-z cryoprobe. Proton chemical shifts were referenced to
internal 3-(trimethyl-silyl)-1-propanesulfonic acid, sodium salt (DSS). Using the absolute
frequency ratios, the 13C and 15N chemical shifts were referenced indirectly to DSS.
G.2.B. Chemical shift assignment and structure calculation.
The following spectra were used to achieve the sequence-specific backbone and
sidechain assignments of all aliphatic residues: 2D [1H,15N]-HSQC, 2D [1H,13C]-HSQC,
3D HNCACB, 3D CBCA(CO)NH, 3D CC(CO)NH, 3D HNCO, 3D HNCA, 3D
HBHA(CO)NH, 3D 15N-resolved [1H,1H]-TOCSY, 3D HC(C)H-TOCSY323. The 2D
[1H,1H]-NOESY, 2D [1H,1H]-TOCSY, and 2D [1H,1H]-COSY spectra of the neurabin986-1056
227
samples in D2O solution after complete H/D exchange of the labile protons were used for
the assignment of the aromatic side chains. The NMR spectra were processed with
Topspin1.3 (Bruker, Billerica, MA) and analyzed with the CARA software package
(http://www.nmr.ch).
The following spectra were used for structure calculation: 3D 15N-resolved
[1H,1H]-NOESY (mixing time of 85 ms), 3D
13
C-resolved [1H,1H]-NOESY (mixing time of
85 ms), and 2D [1H,1H]-NOESY (mixing time 85 ms, D2O solution). The amino acid
sequence, the chemical shift assignment, and the NOESY spectra were input for the
automated
NOESY
peak
picking
and
NOE
assignment
method
of
ATNOS/CANDID/CYANA324-326. The results of ATNOS/CANDID/CYANA were refined by
manual peak adjustment and additional calculations in CYANA. The 20 conformers from
the final CYANA cycle with the lowest residual CYANA target function values (of 100
calculated) were energy-minimized in a water shell with the program CNS327 using the
RECOORD328 script package. The quality of the structures was assessed by the
programs WHATCHECK329, AQUA330, NMR-PROCHECK330, and MOLMOL331.
Chemical shift assignments of neurabin986-1056 were deposited in the BMRB under
accession number 7118 and coordinates were submitted to the PDB under accession
code 2GLE.
G.2.C. Zinc titration of the neurabin SAM domain.
Purified neurabin986-1056 was exchanged from the NMR based phosphate buffer into the
following buffer, 10 mM Tris pH 7.5 and 50 mM NaCl, using a HiTrap desalting column
228
(GE Healthcare). This extra step is necessary to prevent precipitation of Zn3(PO4)2. A 2D
[1H,15N] HSQC spectrum was used to monitor perturbations in 1H and 15N chemical shifts
upon zinc titration. A 50-μM neurabin986-1056 sample was used to obtain the reference
spectrum and 100 μM of Zn2+ was subsequently added.
G.2.D. Neurabin coiled-coil domain (neurabin627-822) expression and
purification.
DNA encompassing the neurabin coiled-coil domain (neurabin627-822) was subcloned into
a modified pETM41 (His6-MBP-TEV-, EMBL) vector. The plasmid was transformed into
the E. coli strain BL21-CodonPlus (DE3)-RIL (Stratagene). Cultures were grown under
vigorous shaking until they reached an OD of
0.7. Expression was induced by 1 mM
IPTG and cultures were incubated overnight ( 18 h) at 18°C under vigorous shaking.
Cells were collected by centrifugation and stored at -80°C until purification. For
purification, cells were resuspended in lysis buffer (50 mM Tris HCl pH 8.0, 5 mM
imidazole, 500 mM NaCl, and 0.1% Triton X-100) containing one Complete EDTA-free
Protease Inhibitor Cocktail Tablet (Roche) and lysed by three passes through a C3
Emulsiflex cell cracker (Avestin). Cellular debris was pelleted by centrifugation at
45,000g/35 min/4°C. The filtered soluble fraction was loaded onto a Histrap HP column
(GE Healthcare) pre-equilibrated with 50 mM Tris-HCl pH 8.0, 5 mM imidazole, and 500
mM NaCl. Protein was eluted using a gradient of 5-500 mM imidazole in 36 column
volumes. Fractions containing MBP-neurabin627-822 were pooled, TEV NIa protease was
added and the sample was dialyzed overnight ( 18 h, 4°C) against 25 mM Tris pH 8.0
and 500 mM NaCl. Successfully cleaved protein was subsequently loaded onto a
229
second his-tag column that had been pre-equilibrated with 50 mM Tris-HCl pH 8.0, 5 mM
imidazole, and 500 mM NaCl and incubated at 4°C with shaking for 1.5 h. All fractions
containing untagged neurabin627-822 were collected and pooled. The sample was
concentrated to
400
M and exchanged into the following buffer: 20 mM sodium
phosphate buffer pH 6.5 and 50 mM NaCl.
G.2.E. Biophysical characterization of neurabin627-822.
CD spectroscopy was used to determine the folded state of neurabin627-822 (Jasco J810). The data were recorded using the following parameters: 20 mM sodium phosphate
buffer, pH 6.5, 500 mM NaCl, and 0.02% NaN3, room temperature, continuous scanning
mode, scanning speed of 20
m/min, 3 scans/experiment, and a cell length of 1 cm.
The CD spectrum of neurabin627-822 confirmed its expected
-helical state. To test the
multimeric state of neurabin627-822, dynamic light scattering (DLS, Viscotec 802), native
gel electrophoresis, and size exclusion chromatography (Superdex-200, GE Healthcare)
were performed.
G.3. Results and discussion.
G.3.A. The neurabin SAM domain has a typical SAM domain fold.
The 3D structure of the neurabin986-1056 has been solved by solution state NMR
spectroscopy. The structure resembles the typical SAM domain fold consisting of five
helices (α-helix1: 9-19; α-helix2: 21-29; α-helix3: 34-40; α-helix4: 42-48; and α-helix5:
53-74). In other SAM domains, such as TEL or ETS-1, the loops that link the helical core
are highly flexible332,333. However, this is not the case for the neurabin SAM domain
230
where all loops are highly structured. Helices 1-4 are relatively short (2-3 helical turns)
and surround the considerably longer (5 turns) C-terminal Helix 5 (Figure 92 C). The
most striking characteristic of the neurabin986-1056 SAM domain is that all five helices form
a compact bundle. The four aromatic residues in the neurabin SAM domain (2 Trp, 1
Tyr, and 1 Phe) are buried in the hydrophobic core of the structure (Figure 92 A,
orange). The surface of the neurabin SAM domain is predominately hydrophilic, with a
positively charged patch along the extended Helix 5 and a more hydrophobic region
interspersed with negative charge on the opposite side, composed of Helices 1 and 2
(Figure 92 B). This mixed charged and hydrophobic surface is surprising, because most
known SAM domain interactions are mediated by hydrophobic surface residues. In
addition, it excludes clearly the possibility that the neurabin SAM domain is an RNA
interacting SAM domain, such as the Smaug and VTS1 SAM domains,319,334-336which
have large positively charged surfaces critical for RNA interaction.
Tyr24 of the neurabin SAM domain is a proposed phosphorylation site that is
conserved in numerous SAM domains. The specific tyrosine kinase for the neurabin
SAM domain is currently unknown. Tyr24 seems to be placed in a protected pocket
formed by Helices 2, 3, and 4 and not accessible for direct phosphorylation. Therefore,
the SAM domain needs to be structurally rearranged to become phosphorylated. This
phosphorylation might be necessary for neurabin SAM domain to interact with its
potential protein interaction partner.
A total of 2121 NOESY derived distance constraints ( 29 NOE constraints per
residue) were used for the structure calculation of neurabin986-1056 (Table 16). The
231
neurabin986-1056
model
has
excellent
stereochemistry,
as
tested
with
NMR-
PROCHECK90,337, with 83.8% of the residues in the most favored region, 16.2% in the
additionally allowed region of the Ramachandran diagram, and no residues in the
generously allowed region and the disallowed region (Table 16). Figure 92 (A) shows the
local RMSD of the side chain heavy atoms. Flexible residues (green) are located on the
surface of the structure whereas the hydrophobic core residues (cyan) are well defined
and show very low mobility.
Neurabin SAM
Number of restrains
Unambiguous distance restrains (all)
2121
Short range |i–j| ≤ 1
982
Medium range 1 < |i–j| < 5
612
Long range |i–j| ≥ 5
527
Unambiguous distance restrains per residue
28.7
Deviations from idealized covalent geometry
Bonds (Å)
0.0108 ± 0.0002
Angles (°)
1.27 ± 0.035
Impropers (°)
1.43 ± 0.07
No. of violation NOE (>0.5 Å)
0±0
No. of violation dihedral (>5°)
0±0
Structural quality
305
Ramachandran plot (NMR-PROCHECK )
Most favored region (%)
83.8
Additionally allowed region (%)
16.2
Generously allowed region (%)
0
Disallowed region (%)
0
RMSD within the bundle (Å)
Backbone (N, Ca, C, and O) (4–70)
0.31 ± 0.05
All heavy atoms (4–70)
0.63 ± 0.05
Table 16. Ju, Proteins Table 1.
Structural and CNS Refinement Statistics for the Neurabin
SAM Domain
232
Figure 92. Ju, Proteins Figure 2.
A) Stereo view of the bundle of 20 energy-minimized CYANA conformers representing the
structure of neurabin986–1056 (PDB ID 2GLE). The superposition is the best fit of the
backbone atoms N, Ca, C0 of residues 4–70. Side chains with low local RMSD (<0.7) within
the bundle are shown in cyan. Side chains of residues with high localRMSD (>0.7) are shown
in green. Backbones are shown in dark blue. Aromatic heavy side chains are colored in
orange. The local RMSD of Trp7, Trp15, Tyr24, and Phe28 are 0.31, 0.11, 0.87, and 0.62,
respectively. B) Electrostatic surface of the neurabin SAM domain (left) and 1808 rotation
around x-axes (right). Red represents negative charges and blue positive charges. C) Ribbon
presentation of the closest conformer to the mean coordinates of the bundle of 20 conformers
used to represent the NMR structure of the neurabin SAM (left) and the shank3 SAM domain
(right). The regular secondary structure elements are identified.
233
G.3.B. The neurabin SAM domain is a monomer in solution.
SAM domains are versatile protein:protein interaction domains. They are known to selfassociate to form homo- and hetero-oligomers338, long helical homopolymers333, or in
some cases even multihelical sheets, as recently reported for the shank3 SAM domain
in the PSD321. To determine the role of the neurabin SAM domain in the multimerization
of neurabin, it is necessary to first determine the oligomeric state of isolated neurabin
SAM domain in solution. Three techniques were employed. First, measurements of the
transverse relaxation times (T2) of the amide protons were carried out using a 1D 1H
one-one spin echo experiments339,340. At different lengths of the echo delay (100 and 2.9
ms), the intensity of the amide proton signals was differentially modulated so as to
enable a determination of the transverse relaxation time. The intensities of
10
separated amide resonances in the 1D 1H NMR spectrum of the neurabin SAM domain
(MW 8.29 kDa) between 8.0 and 10 ppm were used to extract an average amide proton
T2 relaxation time of 30.6 ± 4.2 ms. For comparison, lysozyme, which is known to be a
monomer in solution with a molecular weight of 14.4 kDa, was dissolved in the identical
buffer used for the neurabin SAM measurements and its averaged amide proton T2
relaxation time was 18.2 ± 1.7 ms. Therefore, the significantly longer average amide
proton transverse relaxation time of the neurabin SAM domain when compared with
lysozyme confirms directly that the neurabin SAM is monomeric in solution. Secondly,
this result is also supported by analytical size exclusion chromatography data (Superdex
200, GE Healthcare) and DLS (Viscotec 802) data, which yielded a molecular weight of
about 8 kDa for the neurabin SAM domain (data not shown). This observation
234
demonstrates that the neurabin SAM domain is not a polymerization domain, rather it
must function via protein:protein interaction with other proteins.
G.3.C. The coiled-coil domain, and not the SAM domain, is the solely
responsible multimerization domain of neurabin.
To understand the function of the neurabin and how it differs from spinophilin, it is
important to understand the function of each individual neurabin domain and their
intramolecular interactions. In neurabin, domains fulfill joint functions, that is, the PP1binding and PDZ domain are critical for targeting and anchoring of PP1 close to postsynaptic substrates. Since it has been shown that in vivo neurabin can form homo- or
hetero-multimers36,314 and deletion of the coiled-coil and SAM domains in neurabin
abolishes its dimerization38, it was of interest to see if these domains function as one
entity or are functionally distinct. Although the neurabin SAM domain is clearly a
monomer in solution, the SAM domain could play an indirect dimerization role through
influencing the interaction of the coiled-coil domain. Recombinantly expressed neurabin
coiled-coil domain is clearly a multimer, as tested using DLS, SEC, and native gel
analysis (see Methods). Although it is possible to investigate its multimerization state, it
is more difficult to distinguish between its exact multimerization status, because all
techniques are calibrated for globular proteins and not anisotropic proteins such as the
more extended neurabin coiled-coil domain. However, all of our data suggest that the
neurabin coiled-coil domain is at least a trimer in solution. To test for neurabin coiledcoil:SAM domain interaction, purified 15N labeled neurabin SAM domain was titrated
with 1 equiv. of purified, unlabeled neurabin coiled-coil domain. No changes in the 2D
235
[1H,15N] HSQC spectrum of the neurabin SAM domain were observed (extreme line
broadening beyond detect-ability was expected in case of binding, because of the high
molecular weight of the neurabin coiled-coil multimer) upon the addition of the coiled-coil
domain (data not shown). Thus, the neurabin SAM domain does not interact intra or
intermolecularly with the neurabin coiled-coil domain.
G.3.D. The shank3 SAM domain is the closest structural neighbor of the
neurabin SAM domain.
The 3D structure of the neurabin SAM domain was used to further guide the
identification of its functional role in the PSD. As expected for SAM domains, the
structural similarity between the neurabin SAM domain and other SAM domains is high,
even though the sequence identity is low (between 12 and 28%). A pair-wise
comparison between the neurabin SAM domain structure and other SAM domain
structures was performed using DALI341 and Superpose342, respectively. Both 3D
structure-based methods identified that the closest structural neighbor of the neurabin
SAM domain is the shank3 SAM domain (z-score 10.4; PDBid shank3 SAM domain
PDBid: 2F3N).
The shank3 SAM domain was recently hypothesized to play a critical
organizational scaffolding role in the PSD. Namely, in vitro, it can form 2D multihelical
sheets, which have been proposed to form the underlying architectural framework for the
organization of the PSD. In contrast to the shank3 SAM domain, which forms
homopolymeric sheets, the neurabin SAM domain is a monomer is solution. However, it
was experimentally shown that Zn2+, which is present in vivo in the PSD, can bind to the
236
shank3 SAM domain and thereby dramatically foster this multihelical sheet formation.
The Zn2+ binding pocket of the shank3 SAM domain (E1768, H1769, H1801) is located
in inter and intrapolymer interfaces343, which enhances the organization of the sheet
structure. Given the structural similarity of the shank3 and neurabin SAM domains, we
examined whether Zn2+ might bind to the neurabin SAM domain. Specifically, a 2D
[1H,15N] HSQC spectrum of 15N-labeled neurabin SAM was used to monitor chemical
shift perturbations upon titration with Zn2+. It was possible to identify a potential Zn2+
binding pocket in the neurabin SAM domain formed by His2, His987, Glu988, and
Glu1054 (changes in chemical shifts monitored by 2D [1H,15N] HSQC measurements).
Importantly, His2 is a cloning artifact following TEV HIS-tag cleavage. In the native
neurabin SAM sequence this residue is an arginine (Arg984). Moreover, this zinc binding
pocket is not located in the corresponding shank3 SAM domain structure. Finally, the
neurabin SAM domain remains monomeric in solution upon zinc binding. Therefore, it is
apparent that the native neurabin SAM domain does not have a preformed Zn2+ binding
site, other than a cloning artifact based one, and does not interact with Zn2+ ions. Thus,
despite their similarity in structure, the scaffolding behavior of the neurabin SAM domain
is clearly distinct from that of its closest structural neighbor, the shank3 SAM domain.
G.4. Acknowledgements.
We thank R. Page for careful reading of the manuscript. The DLS instrument was
purchased using Brown University Seed Funds. ITC measurements were performed in
the NSF EPSCoR supported proteomics facility at Brown University. CD measurements
were performed in the RI-INBRE Research Core Facility. We thank Dr. John Marshall
237
(Brown University) for the shank3 plasmid and Dr. James Bowie (UCLA) for discussion
and the shank3 (L47E) protein.
List of References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
Moorhead, G.B., Trinkle-Mulcahy, L. & Ulke-Lemee, A. Emerging roles of nuclear
protein phosphatases. Nat Rev Mol Cell Biol 8, 234-44 (2007).
Ubersax, J.A. & Ferrell, J.E., Jr. Mechanisms of specificity in protein
phosphorylation. Nat Rev Mol Cell Biol 8, 530-41 (2007).
Fersht, A. Structure and mechanism in protein science: a guide to enzyme
catalysis and protein folding, (W.H. Freeman and Company, 1999).
Voet, D. & Voet, J.G. Biochemistry, 1591 (John Wiley & Sons, Inc., Hoboken, NJ,
2004).
Canagarajah, B.J., Khokhlatchev, A., Cobb, M.H. & Goldsmith, E.J. Activation
mechanism of the MAP kinase ERK2 by dual phosphorylation. Cell 90, 859-69
(1997).
Waksman, G., Kumaran, S. & Lubman, O. SH2 domains: role, structure and
implications for molecular medicine. Expert Rev Mol Med 6, 1-18 (2004).
Durocher, D. & Jackson, S.P. The FHA domain. FEBS Lett 513, 58-66 (2002).
Hsieh-Wilson, L.C. et al. Phosphorylation of spinophilin modulates its interaction
with actin filaments. J Biol Chem 278, 1186-94 (2003).
Robbins, D.J. et al. Regulation and properties of extracellular signal-regulated
protein kinases 1 and 2 in vitro. J Biol Chem 268, 5097-106 (1993).
Yaffe, M.B. & Elia, A.E. Phosphoserine/threonine-binding domains. Curr Opin
Cell Biol 13, 131-8 (2001).
Yaffe, M.B. Phosphotyrosine-binding domains in signal transduction. Nat Rev
Mol Cell Biol 3, 177-86 (2002).
Kaneko, T. et al. Loops govern SH2 domain specificity by controlling access to
binding pockets. Sci Signal 3, ra34 (2010).
Mehle, A., Goncalves, J., Santa-Marta, M., McPike, M. & Gabuzda, D.
Phosphorylation of a novel SOCS-box regulates assembly of the HIV-1 Vif-Cul5
complex that promotes APOBEC3G degradation. Genes Dev 18, 2861-6 (2004).
Satoh, A. et al. Neurabin-II/spinophilin. An actin filament-binding protein with one
pdz domain localized at cadherin-based cell-cell adhesion sites. J Biol Chem
273, 3470-5 (1998).
Gong, C.X., Liu, F., Grundke-Iqbal, I. & Iqbal, K. Dysregulation of protein
phosphorylation/dephosphorylation in Alzheimer's disease: a therapeutic target. J
Biomed Biotechnol 2006, 31825 (2006).
Danilkovitch-Miagkova, A. & Zbar, B. Dysregulation of Met receptor tyrosine
kinase activity in invasive tumors. J Clin Invest 109, 863-7 (2002).
Arad, M. et al. Constitutively active AMP kinase mutations cause glycogen
storage disease mimicking hypertrophic cardiomyopathy. J Clin Invest 109, 35762 (2002).
Iqbal, K., Liu, F., Gong, C.X., Alonso Adel, C. & Grundke-Iqbal, I. Mechanisms of
tau-induced neurodegeneration. Acta Neuropathol 118, 53-69 (2009).
Arias-Vasquez, A. et al. Cyclin-dependent kinase 5 is associated with risk for
Alzheimer's diseasein a Dutch population-based study. Journal of Neurology 255,
655-662 (2008).
238
239
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
Brown, A.M., Baucum, A.J., Bass, M.A. & Colbran, R.J. Association of protein
phosphatase 1 gamma 1 with spinophilin suppresses phosphatase activity in a
Parkinson disease model. J Biol Chem 283, 14286-94 (2008).
Bollen, M., Peti, W., Ragusa, M.J. & Beullens, M. The extended PP1 toolkit:
designed to create specificity. Trends Biochem Sci (2010).
Madhusudan et al. cAMP-dependent protein kinase: crystallographic insights into
substrate recognition and phosphotransfer. Protein Sci 3, 176-87 (1994).
Brown, N.R., Noble, M.E., Endicott, J.A. & Johnson, L.N. The structural basis for
specificity of substrate and recruitment peptides for cyclin-dependent kinases.
Nat Cell Biol 1, 438-43 (1999).
Alonso, A. et al. Protein tyrosine phosphatases in the human genome. Cell 117,
699-711 (2004).
Ceulemans, H. & Bollen, M. Functional diversity of protein phosphatase-1, a
cellular economizer and reset button. Physiol Rev 84, 1-39 (2004).
Goldberg, J. et al. Three-dimensional structure of the catalytic subunit of protein
serine/threonine phosphatase-1. Nature 376, 745-53 (1995).
Kelker, M.S., Page, R. & Peti, W. Crystal structures of protein phosphatase-1
bound to nodularin-R and tautomycin: a novel scaffold for structure-based drug
design of serine/threonine phosphatase inhibitors. J Mol Biol 385, 11-21 (2009).
Kita, A. et al. Crystal structure of the complex between calyculin A and the
catalytic subunit of protein phosphatase 1. Structure 10, 715-24 (2002).
Maynes, J.T. et al. Crystal structure of the tumor-promoter okadaic acid bound to
protein phosphatase-1. J Biol Chem 276, 44078-82 (2001).
Maynes, J.T. et al. Crystal structures of protein phosphatase-1 bound to
motuporin and dihydromicrocystin-LA: elucidation of the mechanism of enzyme
inhibition by cyanobacterial toxins. J Mol Biol 356, 111-20 (2006).
Barford, D. Molecular mechanisms of the protein serine/threonine phosphatases.
Trends Biochem Sci 21, 407-12 (1996).
Virshup, D.M. & Shenolikar, S. From promiscuity to precision: protein
phosphatases get a makeover. Mol Cell 33, 537-45 (2009).
Hendrickx, A. et al. Docking motif-guided mapping of the interactome of protein
phosphatase-1. Chem Biol 16, 365-71 (2009).
Cohen, P.T. Protein phosphatase 1--targeted in many directions. J Cell Sci 115,
241-56 (2002).
Egloff, M.P. et al. Structural basis for the recognition of regulatory subunits by the
catalytic subunit of protein phosphatase 1. EMBO J 16, 1876-87 (1997).
Nakanishi, H. et al. Neurabin: a novel neural tissue-specific actin filament-binding
protein involved in neurite formation. J Cell Biol 139, 951-61 (1997).
Gunawardena, S.R., Ruis, B.L., Meyer, J.A., Kapoor, M. & Conklin, K.F. NOM1
targets protein phosphatase I to the nucleolus. J Biol Chem 283, 398-404 (2008).
Oliver, C.J. et al. Targeting protein phosphatase 1 (PP1) to the actin
cytoskeleton: the neurabin I/PP1 complex regulates cell morphology. Mol Cell
Biol 22, 4690-701 (2002).
240
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
Allen, P.B., Kwon, Y.G., Nairn, A.C. & Greengard, P. Isolation and
characterization of PNUTS, a putative protein phosphatase 1 nuclear targeting
subunit. J Biol Chem 273, 4089-95 (1998).
Beullens, M. et al. The C-terminus of NIPP1 (nuclear inhibitor of protein
phosphatase-1) contains a novel binding site for protein phosphatase-1 that is
controlled by tyrosine phosphorylation and RNA binding. Biochem J 352 Pt 3,
651-8 (2000).
Toth, A. et al. Study of the subunit interactions in myosin phosphatase by surface
plasmon resonance. Eur J Biochem 267, 1687-97 (2000).
Allen, P.B., Ouimet, C.C. & Greengard, P. Spinophilin, a novel protein
phosphatase 1 binding protein localized to dendritic spines. Proc Natl Acad Sci U
S A 94, 9956-61 (1997).
Terrak, M., Kerff, F., Langsetmo, K., Tao, T. & Dominguez, R. Structural basis of
protein phosphatase 1 regulation. Nature 429, 780-4 (2004).
Lincoln, T.M. Myosin phosphatase regulatory pathways: different functions or
redundant functions? Circ Res 100, 10-2 (2007).
Suzuki, Y. et al. Insulin control of glycogen metabolism in knockout mice lacking
the muscle-specific protein phosphatase PP1G/RGL. Mol Cell Biol 21, 2683-94
(2001).
Sarrouilhe, D., di Tommaso, A., Metaye, T. & Ladeveze, V. Spinophilin: from
partners to functions. Biochimie 88, 1099-113 (2006).
Feng, J. et al. Spinophilin regulates the formation and function of dendritic
spines. Proc Natl Acad Sci U S A 97, 9287-92 (2000).
Bers, D.M. Macromolecular complexes regulating cardiac ryanodine receptor
function. J Mol Cell Cardiol 37, 417-29 (2004).
Marx, S.O. et al. Phosphorylation-dependent regulation of ryanodine receptors: a
novel role for leucine/isoleucine zippers. J Cell Biol 153, 699-708 (2001).
Yan, Z. et al. Protein phosphatase 1 modulation of neostriatal AMPA channels:
regulation by DARPP-32 and spinophilin. Nat Neurosci 2, 13-7 (1999).
Morishita, W. et al. Regulation of synaptic strength by protein phosphatase 1.
Neuron 32, 1133-48 (2001).
Kelker, M.S. et al. Structural basis for spinophilin-neurabin receptor interaction.
Biochemistry 46, 2333-44 (2007).
MacMillan, L.B. et al. Brain actin-associated protein phosphatase 1 holoenzymes
containing spinophilin, neurabin, and selected catalytic subunit isoforms. J Biol
Chem 274, 35845-54 (1999).
Stralfors, P., Hiraga, A. & Cohen, P. The protein phosphatases involved in
cellular regulation. Purification and characterisation of the glycogen-bound form
of protein phosphatase-1 from rabbit skeletal muscle. Eur J Biochem 149, 295303 (1985).
Delibegovic, M. et al. Disruption of the striated muscle glycogen targeting subunit
PPP1R3A of protein phosphatase 1 leads to increased weight gain, fat
deposition, and development of insulin resistance. Diabetes 52, 596-604 (2003).
241
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
Liu, J. & Brautigan, D.L. Glycogen synthase association with the striated muscle
glycogen-targeting subunit of protein phosphatase-1. Synthase activation
involves scaffolding regulated by beta-adrenergic signaling. J Biol Chem 275,
26074-81 (2000).
Dancheck, B., Nairn, A.C. & Peti, W. Detailed structural characterization of
unbound protein phosphatase 1 inhibitors. Biochemistry 47, 12346-56 (2008).
Ragusa, M.J. et al. Spinophilin directs protein phosphatase 1 specificity by
blocking substrate binding sites. Nat Struct Mol Biol 17, 459-64 (2010).
Gsponer, J. & Babu, M.M. The rules of disorder or why disorder rules. Prog
Biophys Mol Biol 99, 94-103 (2009).
Marsh, J.A. et al. Structural diversity in free and bound states of intrinsically
disordered protein phosphatase 1 regulators. Structure In Print(2010).
Dunker, A.K. et al. Intrinsically disordered protein. J Mol Graph Model 19, 26-59
(2001).
Hurley, T.D. et al. Structural basis for regulation of protein phosphatase 1 by
inhibitor-2. J Biol Chem 282, 28874-83 (2007).
Jurgensen, S. et al. On the mechanism of activation of the ATP X Mg(II)dependent phosphoprotein phosphatase by kinase FA. J Biol Chem 259, 586470 (1984).
Picking, W.D. et al. Fluorescence studies on the interaction of inhibitor 2 and
okadaic acid with the catalytic subunit of type 1 phosphoprotein phosphatases.
Biochemistry 30, 10280-7 (1991).
Terry-Lorenzo, R.T. et al. Neurabins recruit protein phosphatase-1 and inhibitor-2
to the actin cytoskeleton. J Biol Chem 277, 46535-43 (2002).
Connor, J.H., Weiser, D.C., Li, S., Hallenbeck, J.M. & Shenolikar, S. Growth
arrest and DNA damage-inducible protein GADD34 assembles a novel signaling
complex containing protein phosphatase 1 and inhibitor 1. Mol Cell Biol 21, 684150 (2001).
Lesage, B. et al. A complex of catalytically inactive protein phosphatase-1
sandwiched between Sds22 and inhibitor-3. Biochemistry 46, 8909-19 (2007).
Hsieh-Wilson, L.C., Allen, P.B., Watanabe, T., Nairn, A.C. & Greengard, P.
Characterization of the neuronal targeting protein spinophilin and its interactions
with protein phosphatase-1. Biochemistry 38, 4365-73 (1999).
Cohen, P. The regulation of protein function by multisite phosphorylation--a 25
year update. Trends Biochem Sci 25, 596-601 (2000).
Egloff, M.P., Cohen, P.T., Reinemer, P. & Barford, D. Crystal structure of the
catalytic subunit of human protein phosphatase 1 and its complex with tungstate.
J Mol Biol 254, 942-59 (1995).
Bollen, M. Combinatorial control of protein phosphatase-1. Trends Biochem Sci
26, 426-31 (2001).
Bielas, S.L. et al. Spinophilin facilitates dephosphorylation of doublecortin by PP1
to mediate microtubule bundling at the axonal wrist. Cell 129, 579-91 (2007).
Charlton, J.J. et al. Multiple actions of spinophilin regulate mu opioid receptor
function. Neuron 58, 238-47 (2008).
242
74.
75.
76.
77.
78.
79.
80.
81.
82.
83.
84.
85.
86.
87.
88.
89.
90.
91.
Allen, P.B. et al. Distinct roles for spinophilin and neurabin in dopamine-mediated
plasticity. Neuroscience 140, 897-911 (2006).
Dyson, H.J. & Wright, P.E. Intrinsically unstructured proteins and their functions.
Nat Rev Mol Cell Biol 6, 197-208 (2005).
Fraczkiewicz, R. & Braun, W. Exact and efficient analytical calculation of the
accessible surface areas and their gradients for macromolecules. Journal of
Computational Chemistry 19, 319-333 (1998).
Lo Conte, L., Chothia, C. & Janin, J. The atomic structure of protein-protein
recognition sites. J Mol Biol 285, 2177-98 (1999).
Dyson, H.J. & Wright, P.E. Coupling of folding and binding for unstructured
proteins. Curr Opin Struct Biol 12, 54-60 (2002).
Tanaka, J. et al. Interaction of myosin phosphatase target subunit 1 with the
catalytic subunit of type 1 protein phosphatase. Biochemistry 37, 16697-703
(1998).
Hubbard, M.J. & Cohen, P. Regulation of protein phosphatase-1G from rabbit
skeletal muscle. 2. Catalytic subunit translocation is a mechanism for reversible
inhibition of activity toward glycogen-bound substrates. Eur J Biochem 186, 7116 (1989).
Zhang, J., Zhang, Z., Brew, K. & Lee, E.Y. Mutational analysis of the catalytic
subunit of muscle protein phosphatase-1. Biochemistry 35, 6276-82 (1996).
Ceulemans, H. et al. Binding of the concave surface of the Sds22 superhelix to
the alpha 4/alpha 5/alpha 6-triangle of protein phosphatase-1. J Biol Chem 277,
47331-7 (2002).
Otwinowski, Z. & Minor, W. Processing of X-ray diffraction data collected in
oscillation mode. Macromolecular Crystallography, Pt A 276, 307-326 (1997).
McCoy, A.J., Grosse-Kunstleve, R.W., Storoni, L.C. & Read, R.J. Likelihoodenhanced fast translation functions. Acta Crystallogr D Biol Crystallogr 61, 45864 (2005).
Emsley, P. & Cowtan, K. Coot: model-building tools for molecular graphics. Acta
Crystallogr D Biol Crystallogr 60, 2126-32 (2004).
Murshudov, G.N., Vagin, A.A. & Dodson, E.J. Refinement of macromolecular
structures by the maximum-likelihood method. Acta Crystallographica Section DBiological Crystallography 53, 240-255 (1997).
Lamzin, V.S. & Wilson, K.S. Automated refinement of protein models. Acta
Crystallogr D Biol Crystallogr 49, 129-47 (1993).
Painter, J. & Merritt, E.A. TLSMD web server for the generation of multi-group
TLS models. Journal of Applied Crystallography 39, 109-111 (2006).
Davis, I.W. et al. MolProbity: all-atom contacts and structure validation for
proteins and nucleic acids. Nucleic Acids Res 35, W375-83 (2007).
Laskowski, R.A., Macarthur, M.W., Moss, D.S. & Thornton, J.M. Procheck - a
Program to Check the Stereochemical Quality of Protein Structures. Journal of
Applied Crystallography 26, 283-291 (1993).
Cohen, P. et al. Protein phosphatase-1 and protein phosphatase-2A from rabbit
skeletal muscle. Methods Enzymol 159, 390-408 (1988).
243
92.
93.
94.
95.
96.
97.
98.
99.
100.
101.
102.
103.
104.
105.
106.
107.
108.
Jacques, A.M. et al. Myosin binding protein C phosphorylation in normal,
hypertrophic and failing human heart muscle. J Mol Cell Cardiol 45, 209-16
(2008).
Jiang, G. & Zhang, B.B. Glucagon and regulation of glucose metabolism. Am J
Physiol Endocrinol Metab 284, E671-8 (2003).
Rybicka, K.K. Glycosomes--the organelles of glycogen metabolism. Tissue Cell
28, 253-65 (1996).
Nielsen, J.N. & Wojtaszewski, J.F. Regulation of glycogen synthase activity and
phosphorylation by exercise. Proc Nutr Soc 63, 233-7 (2004).
Tang, P.M., Bondor, J.A., Swiderek, K.M. & DePaoli-Roach, A.A. Molecular
cloning and expression of the regulatory (RG1) subunit of the glycogenassociated protein phosphatase. J Biol Chem 266, 15782-9 (1991).
Moorhead, G., MacKintosh, C., Morrice, N. & Cohen, P. Purification of the
hepatic glycogen-associated form of protein phosphatase-1 by microcystinSepharose affinity chromatography. FEBS Lett 362, 101-5 (1995).
Fong, N.M. et al. Identification of binding sites on protein targeting to glycogen for
enzymes of glycogen metabolism. J Biol Chem 275, 35034-9 (2000).
Peti, W. & Page, R. Strategies to maximize heterologous protein expression in
Escherichia coli with minimal cost. Protein Expr Purif 51, 1-10 (2007).
Dosztanyi, Z., Csizmok, V., Tompa, P. & Simon, I. IUPred: web server for the
prediction of intrinsically unstructured regions of proteins based on estimated
energy content. Bioinformatics 21, 3433-4 (2005).
Lerin, C. et al. Regulation and function of the muscle glycogen-targeting subunit
of protein phosphatase 1 (GM) in human muscle cells depends on the COOHterminal region and glycogen content. Diabetes 52, 2221-6 (2003).
Yao, J., Dyson, H.J. & Wright, P.E. Chemical shift dispersion and secondary
structure prediction in unfolded and partly folded proteins. FEBS Lett 419, 285-9
(1997).
van Nuland, N.A., Dobson, C.M. & Regan, L. Characterization of folding the fourhelix bundle protein Rop by real-time NMR. Protein Eng Des Sel 21, 165-70
(2008).
Machovic, M. & Janecek, S. Starch-binding domains in the post-genome era. Cell
Mol Life Sci 63, 2710-24 (2006).
Liu, Y.N., Lai, Y.T., Chou, W.I., Chang, M.D. & Lyu, P.C. Solution structure of
family 21 carbohydrate-binding module from Rhizopus oryzae glucoamylase.
Biochem J 403, 21-30 (2007).
Tung, J.Y. et al. Crystal structures of the starch-binding domain from Rhizopus
oryzae glucoamylase reveal a polysaccharide-binding path. Biochem J 416, 2736 (2008).
Zhang, H.Y., Neal, S. & Wishart, D.S. RefDB: A database of uniformly referenced
protein chemical shifts. Journal of Biomolecular Nmr 25, 173-195 (2003).
Wang, Y.J. & Jardetzky, O. Probability-based protein secondary structure
identification using combined NMR chemical-shift data. Protein Science 11, 852861 (2002).
244
109.
110.
111.
112.
113.
114.
115.
116.
117.
118.
119.
120.
121.
122.
123.
124.
Wishart, D.S., Bigam, C.G., Holm, A., Hodges, R.S. & Sykes, B.D. H-1, C-13 and
N-15 Random Coil Nmr Chemical-Shifts of the Common Amino-Acids .1.
Investigations of Nearest-Neighbor Effects. Journal of Biomolecular Nmr 5, 67-81
(1995).
Wishart, D.S., Sykes, B.D. & Richards, F.M. The Chemical-Shift Index - a Fast
and Simple Method for the Assignment of Protein Secondary Structure through
Nmr-Spectroscopy. Biochemistry 31, 1647-1651 (1992).
Koch, M.H., Vachette, P. & Svergun, D.I. Small-angle scattering: a view on the
properties, structures and structural changes of biological macromolecules in
solution. Q Rev Biophys 36, 147-227 (2003).
Putnam, C.D., Hammel, M., Hura, G.L. & Tainer, J.A. X-ray solution scattering
(SAXS) combined with crystallography and computation: defining accurate
macromolecular structures, conformations and assemblies in solution. Q Rev
Biophys 40, 191-285 (2007).
Svergun, D.I. & Koch, M.H.J. Small-angle scattering studies of biological
macromolecules in solution. Reports on Progress in Physics 66, 1735-1782
(2003).
Wignall, G.D., Lin, J.S. & Spooner, S. Reduction of Parasitic Scattering in SmallAngle X-Ray-Scattering by a 3-Pinhole Collimating System. Journal of Applied
Crystallography 23, 241-245 (1990).
Guinier, A. La diffraction des rayons X aus tres petits angles: application a l'etude
de phenomenes ultramicroscopiques. Ann Phys 12, 161-237 (1939).
Hura, G.L. et al. Robust, high-throughput solution structural analyses by small
angle X-ray scattering (SAXS). Nat Methods 6, 606-12 (2009).
Glatter, O. New Method for Evaluation of Small-Angle Scattering Data. Journal of
Applied Crystallography 10, 415-421 (1977).
Semenyuk, A.V. & Svergun, D.I. Gnom - a Program Package for Small-Angle
Scattering Data-Processing. Journal of Applied Crystallography 24, 537-540
(1991).
Svergun, D.I. Determination of the regularization parameter in indirect transform
methods using perceptual criteria. J Appl Crystallogr 26, 495-503 (1992).
Svergun, D.I. Restoring low resolution structure of biological macromolecules
from solution scattering using simulated annealing. Biophys J 76, 2879-86
(1999).
Svergun, D.I., Petoukhov, M.V. & Koch, M.H. Determination of domain structure
of proteins from X-ray solution scattering. Biophys J 80, 2946-53 (2001).
Volkov, V.V. & Svergun, D.I. Uniqueness of ab initio shape determination in
small-angle scattering. J Appl Crystallogr 36, 860-864 (2003).
Svergun, D.I., Barberato, C. & Koch, M.H.J. CRYSOL - a program to evaluate Xray solution scattering of biological macromolecules from atomic coordinates. J
Appl Crystallogr 28, 768-773 (1995).
Schneidman-Duhovny, D., Hammel, M. & Sali, A. FoXS: a web server for rapid
computation and fitting of SAXS profiles. Nucleic Acids Res (2010).
245
125.
126.
127.
128.
129.
130.
131.
132.
133.
134.
135.
136.
137.
138.
139.
140.
141.
142.
Pelikan, M., Hura, G.L. & Hammel, M. Structure and flexibility within proteins as
identified through small angle X-ray scattering. Gen Physiol Biophys 28, 174-89
(2009).
Zuo, X. et al. Global molecular structure and interfaces: refining an RNA:RNA
complex structure using solution X-ray scattering data. J Am Chem Soc 130,
3292-3 (2008).
Akiyama, S. et al. Conformational landscape of cytochrome c folding studied by
microsecond-resolved small-angle x-ray scattering. Proc Natl Acad Sci U S A 99,
1329-34 (2002).
Tompa, P. Intrinsically unstructured proteins. Trends Biochem Sci 27, 527-33
(2002).
Wright, P.E. & Dyson, H.J. Intrinsically unstructured proteins: re-assessing the
protein structure-function paradigm. J Mol Biol 293, 321-31 (1999).
Dunker, A.K., Brown, C.J., Lawson, J.D., Iakoucheva, L.M. & Obradovic, Z.
Intrinsic disorder and protein function. Biochemistry 41, 6573-82 (2002).
Dunker, A.K. & Obradovic, Z. The protein trinity--linking function and disorder.
Nat Biotechnol 19, 805-6 (2001).
Iakoucheva, L.M., Brown, C.J., Lawson, J.D., Obradovic, Z. & Dunker, A.K.
Intrinsic disorder in cell-signaling and cancer-associated proteins. J Mol Biol 323,
573-84 (2002).
Uversky, V.N. Natively unfolded proteins: a point where biology waits for physics.
Protein Sci 11, 739-56 (2002).
Uversky, V.N. What does it mean to be natively unfolded? Eur J Biochem 269, 212 (2002).
Uversky, V.N., Gillespie, J.R. & Fink, A.L. Why are "natively unfolded" proteins
unstructured under physiologic conditions? Proteins 41, 415-27 (2000).
Cheng, Y. et al. Mining alpha-helix-forming molecular recognition features with
cross species sequence alignments. Biochemistry 46, 13468-77 (2007).
Csizmok, V. et al. Primary contact sites in intrinsically unstructured proteins: the
case of calpastatin and microtubule-associated protein 2. Biochemistry 44, 395564 (2005).
Mohan, A. et al. Analysis of molecular recognition features (MoRFs). J Mol Biol
362, 1043-59 (2006).
Oldfield, C.J. et al. Coupled folding and binding with alpha-helix-forming
molecular recognition elements. Biochemistry 44, 12454-70 (2005).
Tompa, P. The interplay between structure and function in intrinsically
unstructured proteins. Febs Letters 579, 3346-3354 (2005).
Vacic, V. et al. Characterization of molecular recognition features, MoRFs, and
their binding partners. J Proteome Res 6, 2351-66 (2007).
Wakula, P., Beullens, M., Ceulemans, H., Stalmans, W. & Bollen, M. Degeneracy
and function of the ubiquitous RVXF motif that mediates binding to protein
phosphatase-1. J Biol Chem 278, 18817-23 (2003).
246
143.
144.
145.
146.
147.
148.
149.
150.
151.
152.
153.
154.
155.
156.
157.
158.
159.
Ceulemans, H., Stalmans, W. & Bollen, M. Regulator-driven functional
diversification of protein phosphatase-1 in eukaryotic evolution. Bioessays 24,
371-81 (2002).
Li, M., Satinover, D.L. & Brautigan, D.L. Phosphorylation and functions of
inhibitor-2 family of proteins. Biochemistry 46, 2380-9 (2007).
Wang, W., Stukenberg, P.T. & Brautigan, D.L. Phosphatase inhibitor-2 balances
protein phosphatase 1 and aurora B kinase for chromosome segregation and
cytokinesis in human retinal epithelial cells. Mol Biol Cell 19, 4852-62 (2008).
Nairn, A.C. et al. The role of DARPP-32 in the actions of drugs of abuse.
Neuropharmacology 47 Suppl 1, 14-23 (2004).
Watanabe, T. et al. Protein phosphatase 1 regulation by inhibitors and targeting
subunits. Proc Natl Acad Sci U S A 98, 3080-5 (2001).
Choy, W.Y. & Forman-Kay, J.D. Calculation of ensembles of structures
representing the unfolded state of an SH3 domain. J Mol Biol 308, 1011-32
(2001).
Marsh, J.A. & Forman-Kay, J.D. Structure and disorder in an unfolded state
under nondenaturing conditions from ensemble models consistent with a large
number of experimental restraints. J Mol Biol 391, 359-74 (2009).
Marsh, J.A. et al. Improved structural characterizations of the drkN SH3 domain
unfolded state suggest a compact ensemble with native-like and non-native
structure. J Mol Biol 367, 1494-510 (2007).
Feldman, H.J. & Hogue, C.W. A fast method to sample real protein
conformational space. Proteins 39, 112-31 (2000).
Frishman, D. & Argos, P. Knowledge-based protein secondary structure
assignment. Proteins 23, 566-79 (1995).
Chan, H.S. & Dill, K.A. Intrachain Loops in Polymers - Effects of Excluded
Volume. Journal of Chemical Physics 90, 492-509 (1989).
Kozin, M.B. & Svergun, D.I. Automated matching of high- and low-resolution
structural models. J Appl Crystallogr 34, 33-41 (2001).
Weber, G. Ligand binding and internal equilibria in proteins. Biochemistry 11,
864-78 (1972).
Koshland, D.E. Application of a Theory of Enzyme Specificity to Protein
Synthesis. Proc Natl Acad Sci U S A 44, 98-104 (1958).
Sugase, K., Dyson, H.J. & Wright, P.E. Mechanism of coupled folding and
binding of an intrinsically disordered protein. Nature 447, 1021-5 (2007).
Hemmings, H.C., Jr., Greengard, P., Tung, H.Y. & Cohen, P. DARPP-32, a
dopamine-regulated neuronal phosphoprotein, is a potent inhibitor of protein
phosphatase-1. Nature 310, 503-5 (1984).
Park, I.K., Roach, P., Bondor, J., Fox, S.P. & DePaoli-Roach, A.A. Molecular
mechanism of the synergistic phosphorylation of phosphatase inhibitor-2.
Cloning, expression, and site-directed mutagenesis of inhibitor-2. J Biol Chem
269, 944-54 (1994).
247
160.
161.
162.
163.
164.
165.
166.
167.
168.
169.
170.
171.
172.
173.
174.
Park, I.K. & DePaoli-Roach, A.A. Domains of phosphatase inhibitor-2 involved in
the control of the ATP-Mg-dependent protein phosphatase. J Biol Chem 269,
28919-28 (1994).
Satinover, D.L., Leach, C.A., Stukenberg, P.T. & Brautigan, D.L. Activation of
Aurora-A kinase by protein phosphatase inhibitor-2, a bifunctional signaling
protein. Proc Natl Acad Sci U S A 101, 8625-30 (2004).
Li, M., Stukenberg, P.T. & Brautigan, D.L. Binding of phosphatase inhibitor-2 to
prolyl isomerase Pin1 modifies specificity for mitotic phosphoproteins.
Biochemistry 47, 292-300 (2008).
Mittag, T., Kay, L.E. & Forman-Kay, J.D. Protein dynamics and conformational
disorder in molecular recognition. J Mol Recognit 23, 105-16 (2010).
Permyakov, S.E., Millett, I.S., Doniach, S., Permyakov, E.A. & Uversky, V.N.
Natively unfolded C-terminal domain of caldesmon remains substantially
unstructured after the effective binding to calmodulin. Proteins 53, 855-62 (2003).
Sigalov, A.B., Zhuravleva, A.V. & Orekhov, V.Y. Binding of intrinsically
disordered proteins is not necessarily accompanied by a structural transition to a
folded form. Biochimie 89, 419-21 (2007).
Sigalov, A.B., Kim, W.M., Saline, M. & Stern, L.J. The intrinsically disordered
cytoplasmic domain of the T cell receptor zeta chain binds to the nef protein of
simian immunodeficiency virus without a disorder-to-order transition.
Biochemistry 47, 12942-4 (2008).
Tompa, P. & Fuxreiter, M. Fuzzy complexes: polymorphism and structural
disorder in protein-protein interactions. Trends Biochem Sci 33, 2-8 (2008).
Konarev, P.V., Volkov, V.V., Sokolova, A.V., Koch, M.H.J. & Svergun, D.I.
PRIMUS: a Windows-PC based system for small-angle scattering data analysis.
J Appl Crystallogr 36, 1277-1282 (2003).
Garcia De La Torre, J., Huertas, M.L. & Carrasco, B. Calculation of
hydrodynamic properties of globular proteins from their atomic-level structure.
Biophys J 78, 719-30 (2000).
Aggen, J.B., Nairn, A.C. & Chamberlin, R. Regulation of protein phosphatase-1.
Chem Biol 7, R13-23 (2000).
Greengard, P. The neurobiology of slow synaptic transmission. Science 294,
1024-30 (2001).
Oliver, C.J. & Shenolikar, S. Physiologic importance of protein phosphatase
inhibitors. Front Biosci 3, D961-72 (1998).
Kakinoki, Y., Somers, J. & Brautigan, D.L. Multisite phosphorylation and the
nuclear localization of phosphatase inhibitor 2-green fluorescent protein fusion
protein during S phase of the cell growth cycle. J Biol Chem 272, 32308-14
(1997).
Brautigan, D.L., Sunwoo, J., Labbe, J.C., Fernandez, A. & Lamb, N.J. Cell cycle
oscillation of phosphatase inhibitor-2 in rat fibroblasts coincident with p34cdc2
restriction. Nature 344, 74-8 (1990).
248
175.
176.
177.
178.
179.
180.
181.
182.
183.
184.
185.
186.
187.
188.
189.
190.
191.
Senba, S., Eto, M. & Yazawa, M. Identification of trimeric myosin phosphatase
(PP1M) as a target for a novel PKC-potentiated protein phosphatase-1 inhibitory
protein (CPI17) in porcine aorta smooth muscle. J Biochem 125, 354-62 (1999).
Svergun, D.I., Petoukhov, M.V. & Koch, M.H.J. Determination of domain
structure of proteins from X-ray solution scattering. Biophysical Journal 80, 29462953 (2001).
Meiselbach, H., Sticht, H. & Enz, R. Structural analysis of the protein
phosphatase 1 docking motif: molecular description of binding specificities
identifies interacting proteins. Chem Biol 13, 49-59 (2006).
Bollen, M., Peti, W., Ragusa, M.J. & Beullens, M. The extended PP1 toolkit:
designed to create specificity. Trends Biochem Sci 35, 450-458 (2010).
Shoemaker, B.A., Portman, J.J. & Wolynes, P.G. Speeding molecular recognition
by using the folding funnel: the fly-casting mechanism. Proc Natl Acad Sci U S A
97, 8868-73 (2000).
Tonks, N.K. Protein tyrosine phosphatases: from genes, to function, to disease.
Nat Rev Mol Cell Biol 7, 833-46 (2006).
Sheng, M. & Hoogenraad, C.C. The postsynaptic architecture of excitatory
synapses: a more quantitative view. Annu Rev Biochem 76, 823-47 (2007).
Carmody, L.C., Baucum, A.J., 2nd, Bass, M.A. & Colbran, R.J. Selective
targeting of the gamma1 isoform of protein phosphatase 1 to F-actin in intact
cells requires multiple domains in spinophilin and neurabin. FASEB J 22, 166071 (2008).
Strater, N., Klabunde, T., Tucker, P., Witzel, H. & Krebs, B. Crystal-Structure of a
Purple Acid-Phosphatase Containing a Dinuclear Fe(Iii)-Zn(Ii) Active-Site.
Science 268, 1489-1492 (1995).
Peng, R., Lin, J. & Wei, D. Purification and characterization of an organic solventtolerant lipase from Pseudomonas aeruginosa CS-2. Appl Biochem Biotechnol
162, 733-43 (2010).
Endo, S. et al. Conversion of protein phosphatase 1 catalytic subunit to a
Mn(2+)-dependent enzyme impairs its regulation by inhibitor 1. Biochemistry 36,
6986-92 (1997).
Olsen, J.V. et al. Quantitative phosphoproteomics reveals widespread full
phosphorylation site occupancy during mitosis. Sci Signal 3, ra3 (2010).
Manning, G., Plowman, G.D., Hunter, T. & Sudarsanam, S. Evolution of protein
kinase signaling from yeast to man. Trends Biochem Sci 27, 514-20 (2002).
Janssens, V., Longin, S. & Goris, J. PP2A holoenzyme assembly: in cauda
venenum (the sting is in the tail). Trends Biochem Sci 33, 113-21 (2008).
Shi, Y. Serine/threonine phosphatases: mechanism through structure. Cell 139,
468-84 (2009).
Bollen, M. & Stalmans, W. The structure, role, and regulation of type 1 protein
phosphatases. Crit Rev Biochem Mol Biol 27, 227-81 (1992).
Han, J.D. et al. Evidence for dynamically organized modularity in the yeast
protein-protein interaction network. Nature 430, 88-93 (2004).
249
192.
193.
194.
195.
196.
197.
198.
199.
200.
201.
202.
203.
204.
205.
206.
207.
208.
Moorhead, G.B. et al. Displacement affinity chromatography of protein
phosphatase one (PP1) complexes. BMC Biochem 9, 28 (2008).
Flores-Delgado, G., Liu, C.W., Sposto, R. & Berndt, N. A limited screen for
protein interactions reveals new roles for protein phosphatase 1 in cell cycle
control and apoptosis. J Proteome Res 6, 1165-75 (2007).
Eto, M. Regulation of cellular protein phosphatase-1 (PP1) by phosphorylation of
the CPI-17 family, C-kinase-activated PP1 inhibitors. J Biol Chem 284, 35273-7
(2009).
Margolis, S.S. et al. A role for PP1 in the Cdc2/Cyclin B-mediated positive
feedback activation of Cdc25. Mol Biol Cell 17, 1779-89 (2006).
Sankaran, S., Crone, D.E., Palazzo, R.E. & Parvin, J.D. Aurora-A kinase
regulates breast cancer associated gene 1 inhibition of centrosome-dependent
microtubule nucleation. Cancer Res 67, 11186-94 (2007).
Bianchi, M. et al. Regulation of FAK Ser-722 phosphorylation and kinase activity
by GSK3 and PP1 during cell spreading and migration. Biochem J 391, 359-70
(2005).
Nutt, L.K. et al. Metabolic control of oocyte apoptosis mediated by 14-3-3zetaregulated dephosphorylation of caspase-2. Dev Cell 16, 856-66 (2009).
Mi, J., Guo, C., Brautigan, D.L. & Larner, J.M. Protein phosphatase-1alpha
regulates centrosome splitting through Nek2. Cancer Res 67, 1082-9 (2007).
Popescu, M. et al. Ikaros stability and pericentromeric localization are regulated
by protein phosphatase 1. J Biol Chem 284, 13869-80 (2009).
Devogelaere, B. et al. Protein phosphatase-1 is a novel regulator of the
interaction between IRBIT and the inositol 1,4,5-trisphosphate receptor. Biochem
J 407, 303-11 (2007).
Luo, W. et al. Protein phosphatase 1 regulates assembly and function of the
beta-catenin degradation complex. EMBO J 26, 1511-21 (2007).
Traweger, A. et al. Protein phosphatase 1 regulates the phosphorylation state of
the polarity scaffold Par-3. Proc Natl Acad Sci U S A 105, 10402-7 (2008).
Vietri, M., Bianchi, M., Ludlow, J.W., Mittnacht, S. & Villa-Moruzzi, E. Direct
interaction between the catalytic subunit of Protein Phosphatase 1 and pRb.
Cancer Cell Int 6, 3 (2006).
Helps, N.R., Luo, X., Barker, H.M. & Cohen, P.T. NIMA-related kinase 2 (Nek2),
a cell-cycle-regulated protein kinase localized to centrosomes, is complexed to
protein phosphatase 1. Biochem J 349, 509-18 (2000).
Li, L. et al. Phosphorylation of TIMAP by glycogen synthase kinase-3beta
activates its associated protein phosphatase 1. J Biol Chem 282, 25960-9
(2007).
Zeke, A., Lukacs, M., Lim, W.A. & Remenyi, A. Scaffolds: interaction platforms
for cellular signalling circuits. Trends Cell Biol 19, 364-74 (2009).
Harding, H.P. et al. Ppp1r15 gene knockout reveals an essential role for
translation initiation factor 2 alpha (eIF2alpha) dephosphorylation in mammalian
development. Proc Natl Acad Sci U S A 106, 1832-7 (2009).
250
209.
210.
211.
212.
213.
214.
215.
216.
217.
218.
219.
220.
221.
222.
223.
224.
Li, H.Y. et al. Deactivation of the kinase IKK by CUEDC2 through recruitment of
the phosphatase PP1. Nat Immunol 9, 533-41 (2008).
Shi, W. et al. GADD34-PP1c recruited by Smad7 dephosphorylates TGFbeta
type I receptor. J Cell Biol 164, 291-300 (2004).
Watanabe, R. et al. GADD34 inhibits mammalian target of rapamycin signaling
via tuberous sclerosis complex and controls cell survival under bioenergetic
stress. Int J Mol Med 19, 475-83 (2007).
Jin, H., Sperka, T., Herrlich, P. & Morrison, H. Tumorigenic transformation by
CPI-17 through inhibition of a merlin phosphatase. Nature 442, 576-9 (2006).
Kiss, A. et al. Myosin phosphatase interacts with and dephosphorylates the
retinoblastoma protein in THP-1 leukemic cells: its inhibition is involved in the
attenuation of daunorubicin-induced cell death by calyculin-A. Cell Signal 20,
2059-70 (2008).
Yamashiro, S. et al. Myosin phosphatase-targeting subunit 1 regulates mitosis by
antagonizing polo-like kinase 1. Dev Cell 14, 787-97 (2008).
Kim, H. et al. TRF2 functions as a protein hub and regulates telomere
maintenance by recognizing specific peptide motifs. Nat Struct Mol Biol 16, 372-9
(2009).
Lee, S.J., Lee, J.K., Maeng, Y.S., Kim, Y.M. & Kwon, Y.G. Langerhans cell
protein 1 (LCP1) binds to PNUTS in the nucleus: implications for this complex in
transcriptional regulation. Exp Mol Med 41, 189-200 (2009).
Rose, M. et al. PNUTS forms a trimeric protein complex with GABA(C) receptors
and protein phosphatase 1. Mol Cell Neurosci 37, 808-19 (2008).
Wang, X. et al. Spinophilin/neurabin reciprocally regulate signaling intensity by G
protein-coupled receptors. EMBO J 26, 2768-76 (2007).
Nuytten, M. et al. The transcriptional repressor NIPP1 is an essential player in
EZH2-mediated gene silencing. Oncogene 27, 1449-60 (2008).
Tanuma, N. et al. Nuclear inhibitor of protein phosphatase-1 (NIPP1) directs
protein phosphatase-1 (PP1) to dephosphorylate the U2 small nuclear
ribonucleoprotein particle (snRNP) component, spliceosome-associated protein
155 (Sap155). J Biol Chem 283, 35805-14 (2008).
Kim, Y.M. et al. PNUTS, a protein phosphatase 1 (PP1) nuclear targeting
subunit. Characterization of its PP1- and RNA-binding domains and regulation by
phosphorylation. J Biol Chem 278, 13819-28 (2003).
Khromov, A., Choudhury, N., Stevenson, A.S., Somlyo, A.V. & Eto, M.
Phosphorylation-dependent autoinhibition of myosin light chain phosphatase
accounts for Ca2+ sensitization force of smooth muscle contraction. J Biol Chem
284, 21569-79 (2009).
Bollen, M., Gerlich, D.W. & Lesage, B. Mitotic phosphatases: from entry guards
to exit guides. Trends Cell Biol 19, 531-41 (2009).
Scotto-Lavino, E., Garcia-Diaz, M., Du, G. & Frohman, M.A. Basis for the
isoform-specific interaction of myosin phosphatase subunits protein phosphatase
1c beta and myosin phosphatase targeting subunit 1. J Biol Chem 285, 6419-24
(2010).
251
225.
226.
227.
228.
229.
230.
231.
232.
233.
234.
235.
236.
237.
238.
239.
240.
Carmody, L.C. et al. A protein phosphatase-1gamma1 isoform selectivity
determinant in dendritic spine-associated neurabin. J Biol Chem 279, 21714-23
(2004).
Eto, M. et al. Phosphorylation-induced conformational switching of CPI-17
produces a potent myosin phosphatase inhibitor. Structure 15, 1591-602 (2007).
Brush, M.H. & Shenolikar, S. Control of cellular GADD34 levels by the 26S
proteasome. Mol Cell Biol 28, 6989-7000 (2008).
Huang, H.S. & Lee, E.Y. Protein phosphatase-1 inhibitor-3 is an in vivo target of
caspase-3 and participates in the apoptotic response. J Biol Chem 283, 1813546 (2008).
Twomey, E. et al. Regulation of MYPT1 stability by the E3 ubiquitin ligase SIAH2.
Exp Cell Res 316, 68-77 (2010).
Vernia, S. et al. AMP-activated protein kinase phosphorylates R5/PTG, the
glycogen targeting subunit of the R5/PTG-protein phosphatase 1 holoenzyme,
and accelerates its down-regulation by the laforin-malin complex. J Biol Chem
284, 8247-55 (2009).
Wu, J.Q. et al. PP1-mediated dephosphorylation of phosphoproteins at mitotic
exit is controlled by inhibitor-1 and PP1 phosphorylation. Nat Cell Biol 11, 644-51
(2009).
Kelsall, I.R., Rosenzweig, D. & Cohen, P.T. Disruption of the allosteric
phosphorylase a regulation of the hepatic glycogen-targeted protein phosphatase
1 improves glucose tolerance in vivo. Cell Signal 21, 1123-34 (2009).
Rubenstein, E.M. et al. Access denied: Snf1 activation loop phosphorylation is
controlled by availability of the phosphorylated threonine 210 to the PP1
phosphatase. J Biol Chem 283, 222-30 (2008).
Shi, Y. & Manley, J.L. A complex signaling pathway regulates SRp38
phosphorylation and pre-mRNA splicing in response to heat shock. Mol Cell 28,
79-90 (2007).
Lesage, B. et al. Interactor-mediated nuclear translocation and retention of
protein phosphatase-1. J Biol Chem 279, 55978-84 (2004).
Trinkle-Mulcahy, L., Sleeman, J.E. & Lamond, A.I. Dynamic targeting of protein
phosphatase 1 within the nuclei of living mammalian cells. J Cell Sci 114, 421928 (2001).
Huang, L.J. et al. NH2-Terminal targeting motifs direct dual specificity A-kinaseanchoring protein 1 (D-AKAP1) to either mitochondria or endoplasmic reticulum.
J Cell Biol 145, 951-9 (1999).
Schillace, R.V. & Scott, J.D. Association of the type 1 protein phosphatase PP1
with the A-kinase anchoring protein AKAP220. Curr Biol 9, 321-4 (1999).
Takahashi, M. et al. Characterization of a novel giant scaffolding protein, CGNAP, that anchors multiple signaling enzymes to centrosome and the golgi
apparatus. J Biol Chem 274, 17267-74 (1999).
Katayama, H., Zhou, H., Li, Q., Tatsuka, M. & Sen, S. Interaction and feedback
regulation between STK15/BTAK/Aurora-A kinase and protein phosphatase 1
through mitotic cell division cycle. J Biol Chem 276, 46219-24 (2001).
252
241.
242.
243.
244.
245.
246.
247.
248.
249.
250.
251.
252.
253.
254.
255.
256.
Sugiyama, K. et al. Aurora-B associated protein phosphatases as negative
regulators of kinase activation. Oncogene 21, 3103-11 (2002).
Ayllon, V. et al. Bcl-2 targets protein phosphatase 1 alpha to Bad. J Immunol
166, 7345-52 (2001).
Ayllon, V., Cayla, X., Garcia, A., Fleischer, A. & Rebollo, A. The anti-apoptotic
molecules Bcl-xL and Bcl-w target protein phosphatase 1alpha to Bad. Eur J
Immunol 32, 1847-55 (2002).
Hsu, L.C. Identification and functional characterization of a PP1-binding site in
BRCA1. Biochem Biophys Res Commun 360, 507-12 (2007).
Dessauge, F. et al. Identification of PP1alpha as a caspase-9 regulator in IL-2
deprivation-induced apoptosis. J Immunol 177, 2441-51 (2006).
Eto, M., Senba, S., Morita, F. & Yazawa, M. Molecular cloning of a novel
phosphorylation-dependent inhibitory protein of protein phosphatase-1 (CPI17) in
smooth muscle: its specific localization in smooth muscle. FEBS Lett 410, 356-60
(1997).
Gao, Y. et al. Protein phosphatase-1 is targeted to DNA polymerase delta via an
interaction with the p68 subunit. Biochemistry 47, 11367-76 (2008).
Wakula, P. et al. The translation initiation factor eIF2beta is an interactor of
protein phosphatase-1. Biochem J 400, 377-83 (2006).
Shi, W. et al. Endofin acts as a Smad anchor for receptor activation in BMP
signaling. J Cell Sci 120, 1216-24 (2007).
Hrabchak, C., Henderson, H. & Varmuza, S. A testis specific isoform of
endophilin B1, endophilin B1t, interacts specifically with protein phosphatase-1c
gamma2 in mouse testis and is abnormally expressed in PP1c gamma null mice.
Biochemistry 46, 4635-44 (2007).
Pasder, O. et al. Downregulation of Fer induces PP1 activation and cell-cycle
arrest in malignant cells. Oncogene 25, 4194-206 (2006).
Platholi, J., Heerdt, P.M., Lim Tung, H.Y. & Hemmings, H.C., Jr. Activation of
brain protein phosphatase-1(I) following cardiac arrest and resuscitation involving
an interaction with 14-3-3 gamma. J Neurochem 105, 2029-38 (2008).
Novoa, I., Zeng, H., Harding, H.P. & Ron, D. Feedback inhibition of the unfolded
protein response by GADD34-mediated dephosphorylation of eIF2alpha. J Cell
Biol 153, 1011-22 (2001).
Liu, Q.R. et al. GBPI, a novel gastrointestinal- and brain-specific PP1-inhibitory
protein, is activated by PKC and inactivated by PKA. Biochem J 377, 171-81
(2004).
Doherty, M.J., Moorhead, G., Morrice, N., Cohen, P. & Cohen, P.T. Amino acid
sequence and expression of the hepatic glycogen-binding (GL)-subunit of protein
phosphatase-1. FEBS Lett 375, 294-8 (1995).
Johnson, D.F. et al. Identification of protein-phosphatase-1-binding domains on
the glycogen and myofibrillar targetting subunits. Eur J Biochem 239, 317-25
(1996).
253
257.
258.
259.
260.
261.
262.
263.
264.
265.
266.
267.
268.
269.
270.
271.
Hall, K.U. et al. Phosphorylation-dependent inhibition of protein phosphatase-1
by G-substrate. A Purkinje cell substrate of the cyclic GMP-dependent protein
kinase. J Biol Chem 274, 3485-95 (1999).
Ajuh, P.M. et al. Association of a protein phosphatase 1 activity with the human
factor C1 (HCF) complex. Nucleic Acids Res 28, 678-86 (2000).
Brush, M.H., Guardiola, A., Connor, J.H., Yao, T.P. & Shenolikar, S. Deactylase
inhibitors disrupt cellular complexes containing protein phosphatases and
deacetylases. J Biol Chem 279, 7685-91 (2004).
Endo, S., Zhou, X., Connor, J., Wang, B. & Shenolikar, S. Multiple structural
elements define the specificity of recombinant human inhibitor-1 as a protein
phosphatase-1 inhibitor. Biochemistry 35, 5220-8 (1996).
Zhang, J., Zhang, L., Zhao, S. & Lee, E.Y. Identification and characterization of
the human HCG V gene product as a novel inhibitor of protein phosphatase-1.
Biochemistry 37, 16728-34 (1998).
Gushiken, F.C. et al. The catalytic subunit of protein phosphatase 1 gamma
regulates thrombin-induced murine platelet alpha(IIb)beta(3) function. PLoS One
4, e8304 (2009).
Wang, X. et al. IPP5, a novel protein inhibitor of protein phosphatase 1, promotes
G1/S progression in a Thr-40-dependent manner. J Biol Chem 283, 12076-84
(2008).
Liu, Q.R. et al. KEPI, a PKC-dependent protein phosphatase 1 inhibitor regulated
by morphine. J Biol Chem 277, 13312-20 (2002).
Wang, H. & Brautigan, D.L. A novel transmembrane Ser/Thr kinase complexes
with protein phosphatase-1 and inhibitor-2. J Biol Chem 277, 49605-12 (2002).
Hirano, K., Ito, M. & Hartshorne, D.J. Interaction of the ribosomal protein, L5, with
protein phosphatase type 1. J Biol Chem 270, 19786-90 (1995).
Gagnon, K.B., England, R., Diehl, L. & Delpire, E. Apoptosis-associated tyrosine
kinase scaffolding of protein phosphatase 1 and SPAK reveals a novel pathway
for Na-K-2C1 cotransporter regulation. Am J Physiol Cell Physiol 292, C1809-15
(2007).
Croci, C., Sticht, H., Brandstatter, J.H. & Enz, R. Group I metabotropic glutamate
receptors bind to protein phosphatase 1C. Mapping and modeling of interacting
sequences. J Biol Chem 278, 50682-90 (2003).
Moorhead, G., Johnson, D., Morrice, N. & Cohen, P. The major myosin
phosphatase in skeletal muscle is a complex between the beta-isoform of protein
phosphatase 1 and the MYPT2 gene product. FEBS Lett 438, 141-4 (1998).
Skinner, J.A. & Saltiel, A.R. Cloning and identification of MYPT3: a prenylatable
myosin targetting subunit of protein phosphatase 1. Biochem J 356, 257-67
(2001).
Patel, K.G., Liu, C., Cameron, P.L. & Cameron, R.S. Myr 8, a novel
unconventional myosin expressed during brain development associates with the
protein phosphatase catalytic subunits 1alpha and 1gamma1. J Neurosci 21,
7954-68 (2001).
254
272.
273.
274.
275.
276.
277.
278.
279.
280.
281.
282.
283.
284.
285.
286.
287.
Hermanson, O., Jepsen, K. & Rosenfeld, M.G. N-CoR controls differentiation of
neural stem cells into astrocytes. Nature 419, 934-9 (2002).
McAvoy, T. et al. Regulation of neurabin I interaction with protein phosphatase 1
by phosphorylation. Biochemistry 38, 12943-9 (1999).
Terry-Lorenzo, R.T. et al. Neurofilament-L is a protein phosphatase-1-binding
protein associated with neuronal plasma membrane and post-synaptic density. J
Biol Chem 275, 2439-46 (2000).
Misik, A.J., Perreault, K., Holmes, C.F. & Fliegel, L. Protein phosphatase
regulation of Na+/H+ exchanger isoform I. Biochemistry 44, 5842-52 (2005).
Darman, R.B., Flemmer, A. & Forbush, B. Modulation of ion transport by direct
targeting of protein phosphatase type 1 to the Na-K-Cl cotransporter. J Biol
Chem 276, 34359-62 (2001).
Seth, A., Sheth, P., Elias, B.C. & Rao, R. Protein phosphatases 2A and 1 interact
with occludin and negatively regulate the assembly of tight junctions in the
CACO-2 cell monolayer. J Biol Chem 282, 11487-98 (2007).
Helps, N.R., Barker, H.M., Elledge, S.J. & Cohen, P.T. Protein phosphatase 1
interacts with p53BP2, a protein which binds to the tumour suppressor p53.
FEBS Lett 377, 295-300 (1995).
Tan, I., Ng, C.H., Lim, L. & Leung, T. Phosphorylation of a novel myosin binding
subunit of protein phosphatase 1 reveals a conserved mechanism in the
regulation of actin cytoskeleton. J Biol Chem 276, 21209-16 (2001).
Zhao, S. & Lee, E.Y. Targeting of the catalytic subunit of protein phosphatase-1
to the glycolytic enzyme phosphofructokinase. Biochemistry 36, 8318-24 (1997).
Sagara, J. et al. Scapinin, a putative protein phosphatase-1 regulatory subunit
associated with the nuclear nonchromatin structure. J Biol Chem 278, 45611-9
(2003).
Eto, M., Karginov, A. & Brautigan, D.L. A novel phosphoprotein inhibitor of
protein type-1 phosphatase holoenzymes. Biochemistry 38, 16952-7 (1999).
Kao, S.C. et al. Identification of phostensin, a PP1 F-actin cytoskeleton targeting
subunit. Biochem Biophys Res Commun 356, 594-8 (2007).
Kwiek, N.C., Thacker, D.F., Datto, M.B., Megosh, H.B. & Haystead, T.A. PITK, a
PP1 targeting subunit that modulates the phosphorylation of the transcriptional
regulator hnRNP K. Cell Signal 18, 1769-78 (2006).
Tan, S.L., Tareen, S.U., Melville, M.W., Blakely, C.M. & Katze, M.G. The direct
binding of the catalytic subunit of protein phosphatase 1 to the PKR protein
kinase is necessary but not sufficient for inactivation and disruption of enzyme
dimer formation. J Biol Chem 277, 36109-17 (2002).
Kreivi, J.P. et al. Purification and characterisation of p99, a nuclear modulator of
protein phosphatase 1 activity. FEBS Lett 420, 57-62 (1997).
Yoshimura, K. et al. Interaction of p130 with, and consequent inhibition of, the
catalytic subunit of protein phosphatase 1alpha. J Biol Chem 276, 17908-13
(2001).
255
288.
289.
290.
291.
292.
293.
294.
295.
296.
297.
298.
299.
300.
301.
302.
Yoshida, K., Watanabe, M., Kato, H., Dutta, A. & Sugano, S. BH-protocadherin-c,
a member of the cadherin superfamily, interacts with protein phosphatase 1
alpha through its intracellular domain. FEBS Lett 460, 93-8 (1999).
Hirano, K., Erdodi, F., Patton, J.G. & Hartshorne, D.J. Interaction of protein
phosphatase type 1 with a splicing factor. FEBS Lett 389, 191-4 (1996).
Brady, M.J., Printen, J.A., Mastick, C.C. & Saltiel, A.R. Role of protein targeting
to glycogen (PTG) in the regulation of protein phosphatase-1 activity. J Biol
Chem 272, 20198-204 (1997).
Armstrong, C.G., Browne, G.J., Cohen, P. & Cohen, P.T. PPP1R6, a novel
member of the family of glycogen-targetting subunits of protein phosphatase 1.
FEBS Lett 418, 210-4 (1997).
Durfee, T. et al. The retinoblastoma protein associates with the protein
phosphatase type 1 catalytic subunit. Genes Dev 7, 555-69 (1993).
Trinkle-Mulcahy, L. et al. Repo-Man recruits PP1 gamma to chromatin and is
essential for cell viability. J Cell Biol 172, 679-92 (2006).
Zhao, S., Brandt, N.R., Caswell, A.H. & Lee, E.Y. Binding of the catalytic subunit
of protein phosphatase-1 to the ryanodine-sensitive calcium release channel
protein. Biochemistry 37, 18102-9 (1998).
Browne, G.J. et al. SARP, a new alternatively spliced protein phosphatase 1 and
DNA interacting protein. Biochem J 402, 187-96 (2007).
Ohkura, H. & Yanagida, M. S. pombe gene sds22+ essential for a midmitotic
transition encodes a leucine-rich repeat protein that positively modulates protein
phosphatase-1. Cell 64, 149-57 (1991).
Llorian, M., Beullens, M., Andres, I., Ortiz, J.M. & Bollen, M. SIPP1, a novel premRNA splicing factor and interactor of protein phosphatase-1. Biochem J 378,
229-38 (2004).
Wu, D.Y., Tkachuck, D.C., Roberson, R.S. & Schubach, W.H. The human
SNF5/INI1 protein facilitates the function of the growth arrest and DNA damageinducible protein (GADD34) and modulates GADD34-bound protein
phosphatase-1 activity. J Biol Chem 277, 27706-15 (2002).
Hrabchak, C. & Varmuza, S. Identification of the spermatogenic zip protein Spz1
as a putative protein phosphatase-1 (PP1) regulatory protein that specifically
binds the PP1cgamma2 splice variant in mouse testis. J Biol Chem 279, 3707986 (2004).
Monshausen, M., Rehbein, M., Richter, D. & Kindler, S. The RNA-binding protein
Staufen from rat brain interacts with protein phosphatase-1. J Neurochem 81,
557-64 (2002).
Liao, H., Li, Y., Brautigan, D.L. & Gundersen, G.G. Protein phosphatase 1 is
targeted to microtubules by the microtubule-associated protein Tau. J Biol Chem
273, 21901-8 (1998).
Eto, M., Kirkbride, J., Elliott, E., Lo, S.H. & Brautigan, D.L. Association of the
tensin N-terminal protein-tyrosine phosphatase domain with the alpha isoform of
protein phosphatase-1 in focal adhesions. J Biol Chem 282, 17806-15 (2007).
256
303.
304.
305.
306.
307.
308.
309.
310.
311.
312.
313.
314.
315.
316.
317.
318.
319.
320.
321.
Kim, K. et al. The protein phosphatase-1 targeting subunit TIMAP regulates
LAMR1 phosphorylation. Biochem Biophys Res Commun 338, 1327-34 (2005).
Novoyatleva, T. et al. Protein phosphatase 1 binds to the RNA recognition motif
of several splicing factors and regulates alternative pre-mRNA processing. Hum
Mol Genet 17, 52-70 (2008).
Djouder, N. et al. S6K1-mediated disassembly of mitochondrial URI/PP1gamma
complexes activates a negative feedback program that counters S6K1 survival
signaling. Mol Cell 28, 28-40 (2007).
Bettoun, D.J. et al. A vitamin D receptor-Ser/Thr phosphatase-p70 S6 kinase
complex and modulation of its enzymatic activities by the ligand. J Biol Chem
277, 24847-50 (2002).
Ulke-Lemee, A. et al. The nuclear PP1 interacting protein ZAP3 (ZAP) is a
putative nucleoside kinase that complexes with SAM68, CIA, NF110/45, and
HNRNP-G. Biochim Biophys Acta 1774, 1339-50 (2007).
Hendrickx, A. et al. Docking motif-guided mapping of the interactome of protein
phosphatase-1. Chem. Biol. 16, 365-71 (2009).
Kelker, M.S. & Peti, W. NMR assignment of the spinophilin PDZ domain (493602). J Biomol NMR 36 Suppl 1, 24 (2006).
Grossman, S.D., Hsieh-Wilson, L.C., Allen, P.B., Nairn, A.C. & Greengard, P.
The actin-binding domain of spinophilin is necessary and sufficient for targeting
to dendritic spines. Neuromolecular Med 2, 61-9 (2002).
Burnett, P.E. et al. Neurabin is a synaptic protein linking p70 S6 kinase and the
neuronal cytoskeleton. Proc Natl Acad Sci U S A 95, 8351-6 (1998).
Penzes, P. et al. The neuronal Rho-GEF Kalirin-7 interacts with PDZ domaincontaining proteins and regulates dendritic morphogenesis. Neuron 29, 229-42
(2001).
Shmueli, A. et al. Site-specific dephosphorylation of doublecortin (DCX) by
protein phosphatase 1 (PP1). Mol Cell Neurosci 32, 15-26 (2006).
Ryan, X.P. et al. The Rho-specific GEF Lfc interacts with neurabin and
spinophilin to regulate dendritic spine morphology. Neuron 47, 85-100 (2005).
Keegan, J., Schmerer, M., Ring, B. & Garza, D. The 62E early-late puff of
Drosophila contains D-spinophilin, an ecdysone-inducible PDZ-domain protein
dynamically expressed during metamorphosis. Genet Res 77, 27-39 (2001).
Ponting, C.P. SAM: a novel motif in yeast sterile and Drosophila polyhomeotic
proteins. Protein Sci 4, 1928-30 (1995).
Kim, C.A. & Bowie, J.U. SAM domains: uniform structure, diversity of function.
Trends Biochem Sci 28, 625-8 (2003).
Qiao, F. & Bowie, J.U. The many faces of SAM. Sci STKE 2005, re7 (2005).
Green, J.B., Gardner, C.D., Wharton, R.P. & Aggarwal, A.K. RNA recognition via
the SAM domain of Smaug. Mol Cell 11, 1537-48 (2003).
Edwards, T.A. et al. Solution structure of the Vts1 SAM domain in the presence
of RNA. J Mol Biol 356, 1065-72 (2006).
Baron, M.K. et al. An architectural framework that may lie at the core of the
postsynaptic density. Science 311, 531-5 (2006).
257
322.
323.
324.
325.
326.
327.
328.
329.
330.
331.
332.
333.
334.
335.
336.
337.
338.
Jan, H.H., Chen, I.T., Tsai, Y.Y. & Chang, Y.C. Structural role of zinc ions bound
to postsynaptic densities. J Neurochem 83, 525-34 (2002).
Sattler, M., Schleucher, J. & Griesinger, C. Heteronuclear multidimensional NMR
experiments for the structure determination of proteins in solution employing
pulsed field gradients. Progress in Nuclear Magnetic Resonance Spectroscopy
34, 93-158 (1999).
Herrmann, T., Guntert, P. & Wuthrich, K. Protein NMR structure determination
with automated NOE assignment using the new software CANDID and the
torsion angle dynamics algorithm DYANA. Journal of Molecular Biology 319, 209227 (2002).
Herrmann, T., Guntert, P. & Wuthrich, K. Protein NMR structure determination
with automated NOE assignment using the new software CANDID and the
torsion angle dynamics algorithm DYANA. J Mol Biol 319, 209-27 (2002).
Guntert, P. Automated NMR structure calculation with CYANA. Methods Mol Biol
278, 353-78 (2004).
Brunger, A.T. et al. Crystallography & NMR system: A new software suite for
macromolecular structure determination. Acta Crystallogr D Biol Crystallogr 54,
905-21 (1998).
Nederveen, A.J. et al. RECOORD: a recalculated coordinate database of 500+
proteins from the PDB using restraints from the BioMagResBank. Proteins 59,
662-72 (2005).
Hooft, R.W., Vriend, G., Sander, C. & Abola, E.E. Errors in protein structures.
Nature 381, 272 (1996).
Laskowski, R.A., Rullmannn, J.A., MacArthur, M.W., Kaptein, R. & Thornton, J.M.
AQUA and PROCHECK-NMR: programs for checking the quality of protein
structures solved by NMR. J Biomol NMR 8, 477-86 (1996).
Koradi, R., Billeter, M. & Wuthrich, K. MOLMOL: a program for display and
analysis of macromolecular structures. J Mol Graph 14, 51-5, 29-32 (1996).
Slupsky, C.M. et al. Structure of the Ets-1 pointed domain and mitogen-activated
protein kinase phosphorylation site. Proc Natl Acad Sci U S A 95, 12129-34
(1998).
Kim, C.A. et al. Polymerization of the SAM domain of TEL in leukemogenesis
and transcriptional repression. EMBO J 20, 4173-82 (2001).
Aviv, T. et al. The RNA-binding SAM domain of Smaug defines a new family of
post-transcriptional regulators. Nat Struct Biol 10, 614-21 (2003).
Johnson, P.E. & Donaldson, L.W. RNA recognition by the Vts1p SAM domain.
Nat Struct Mol Biol 13, 177-8 (2006).
Oberstrass, F.C. et al. Shape-specific recognition in the structure of the Vts1p
SAM domain with RNA. Nat Struct Mol Biol 13, 160-7 (2006).
Morris, A.L., MacArthur, M.W., Hutchinson, E.G. & Thornton, J.M.
Stereochemical quality of protein structure coordinates. Proteins 12, 345-64
(1992).
Thanos, C.D., Goodwill, K.E. & Bowie, J.U. Oligomeric structure of the human
EphB2 receptor SAM domain. Science 283, 833-6 (1999).
258
339.
340.
341.
342.
343.
Kelker, M.S. et al. Crystal structure of human triggering receptor expressed on
myeloid cells 1 (TREM-1) at 1.47 A. J Mol Biol 342, 1237-48 (2004).
Sklenar, V. & Bax, A. Spin-Echo Water Suppression for the Generation of PurePhase Two-Dimensional Nmr-Spectra. Journal of Magnetic Resonance 74, 469479 (1987).
Holm, L. & Sander, C. Searching protein structure databases has come of age.
Proteins 19, 165-73 (1994).
Maiti, R., Van Domselaar, G.H., Zhang, H. & Wishart, D.S. SuperPose: a simple
server for sophisticated structural superposition. Nucleic Acids Res 32, W590-4
(2004).
Gundelfinger, E.D., Boeckers, T.M., Baron, M.K. & Bowie, J.U. A role for zinc in
postsynaptic density asSAMbly and plasticity? Trends Biochem Sci 31, 366-73
(2006).

The Regulation of Protein Phosphatase 1 by Targeting Proteins

Transcription

Similar documents

LISTENING REVOLUTION

Untitled - Eurobride

2013 collection

Tegra Linux Driver Package R21.3 Release Notes

Release Notes - NVIDIA Developer

print version

The CALIPSO/Cloud SAT Dataset by Graeme Stephens

Je Me Souviens - American-French Genealogical Society

Hypnology Magazine: Issue 3