A missense mutation in Caenorhabditis elegans prohibitin 2 confers

A missense mutation in Caenorhabditis elegans
prohibitin 2 confers an atypical multidrug resistance
Iryna Zubovych, Thomas Doundoulakis, Patrick G. Harran, and Michael G. Roth*
Department of Biochemistry, University of Texas Southwestern Medical Center, Dallas, TX 75390-9038
Communicated by Xiaodong Wang, University of Texas Southwestern Medical Center, Dallas, TX, August 24, 2006 (received for review July 1, 2006)
Me
Me
O
Me
N
CO2H
N
H
NHMe
N
Me
O
Pr-i
Me
1
Me
Me
O
Bu-t
Me
N
CO2H
N
H
NHMe
R
I
O
Pr-i
Me
2 R = H (HTI-286)
3 R = Br
4 R=A
antimitotic 兩 hemiasterlin
dentifying the molecular target(s) and mechanisms of biological activity of small molecules remains a major challenge.
Among the tactics available, the techniques of classical forward
genetics are particularly attractive. A genetic screen or selection
for organisms resistant to a drug requires no assumptions about
the drug target, mechanism of action, or biological pathway.
Such a screen can also reveal mechanisms that can inactivate a
compound or suppress its effects, issues of great importance for
the development of effective therapeutics. We chose the roundworm Caenorhabditis elegans as a genetic model system to
investigate possible pathways involved in resistance to the potent
antimitotic hemiasterlin.
Hemiasterlin is one of the naturally occurring toxins found to
act on tubulin (1). Hemiasterlin poisons spindle assembly in
cultured cells, leading to growth arrest in mitosis (2). Among
hemiasterlin analogs evaluated, HTI-286 (compound 2, Fig. 1)
has been characterized most extensively. This compound inhibits
microtubule assembly in vitro (3), and mutations in ␣- or
␤-tubulin that increase microtubule stability confer resistance to
HTI-286 in ovarian carcinoma cells (4). Because HTI-286 is in
clinical development as a promising cancer chemotherapeutic
(5), we initiated genetic experiments to identify additional
pathways associated with hemiasterlin toxicity. We have isolated
eight mutants resistant to a functional analog of hemiasterlin and
characterized the mutated gene in one recessive mutant as that
encoding prohibitin 2. We show that a single amino acid change
in prohibitin 2 protects C. elegans not only from hemiasterlin but
also from several other toxins that bind tubulin and from
camptothecin, an inhibitor of DNA replication. However, this
mutation confers no resistance to the tubulin polymerization
inhibitor nocodazole or the actin stabilizer phalloidin.
Bu-t
Me
Me
O
Bu-t
Me
CO2H
N
N
H
NHMe
R
O
Pr-i
5 R=A
O
A=
HN
NH
H
N
S
O
Fig. 1. Hemiasterlin (1), HTI-286 (2), and synthetic variants (3–5) used in this
study. Compounds 3 and 4 were designed and synthesized based on earlier
structure兾activity relationships established by Nieman et al. (7). Compound 5
is an inactive control.
(Table 1, which is published as supporting information on the
PNAS web site). Saturated compound 5, as anticipated by
precedent (7), was found to be inactive and served as a structurally related, biotinylated negative control. Compound 3 inhibited microtubule formation in vitro (Fig. 5, which is published
Results
Compound 3 Is Cytotoxic, Binds Tubulin, and Inhibits Tubulin Polymerization. We synthesized three variants of hemiasterlin (Fig. 1)
and compared their properties to those reported for natural
hemiasterlin. Compound 3 intoxicated cells at concentrations
similar to those reported for hemiasterlin (6, 7), and the biotinconjugate of compound 3, compound 4, was similarly toxic
www.pnas.org兾cgi兾doi兾10.1073兾pnas.0607338103
Author contributions: I.Z., P.G.H., and M.G.R. designed research; I.Z. performed research;
T.D. and P.G.H. contributed new reagents兾analytic tools; I.Z., P.G.H., and M.G.R. analyzed
data; and I.Z., P.G.H., and M.G.R. wrote the paper.
Conflict of interest statement: X.W. is part of the program grant that produced this work.
*To whom correspondence should be addressed. E-mail: michael.roth@utsouthwestern.edu.
© 2006 by The National Academy of Sciences of the USA
PNAS 兩 October 17, 2006 兩 vol. 103 兩 no. 42 兩 15523–15528
GENETICS
Hemiasterlin is a potent antimitotic peptide that interferes with
microtubule dynamics at picomolar concentrations in cell culture.
The molecule largely eludes P glycoprotein-mediated drug efflux,
and an analog is currently being evaluated in clinical trials as cancer
chemotherapy. From a nonclonal genetic screen in Caenorhabditis
elegans we isolated eight independent mutants resistant to a
synthetic hemiasterlin analog. In one recessive mutant, phb2(ad2154), a point mutation in prohibitin 2 (E130K) protects worms
from drug-induced injury. Data indicate that direct binding of
hemiasterlin to prohibitin 2 is unlikely. In fact, C. elegans phb2(ad2154) was also found to be resistant to numerous other drugs
that bind tubulin and to camptothecin, yet this mutant was
sensitive to nocodazole and phalloidin. Thus, prohibitin 2 is
implicated in a previously uncharacterized pathway of multidrug
resistance.
Fig. 2. Compound 4 binds tubulin but not PHB2. (A) Toxic 4 binds tubulin and
nontoxic 5 does not. Preparations of purified tubulin, tubulin enriched in microtubule-associated proteins, and a HeLa cell lysate were incubated overnight with
5 ␮M biotinylated hemiasterlin probes or an equal volume of the DMSO vehicle
and then precipitated with neutravidin beads. The precipitates were analyzed by
PAGE, and tubulin was detected by immunoblotting. A result typical of five
experiments is shown. (B) Compound 4 fails to pull down recombinant PHB2.
Purified recombinant PHB2 was prepared as described and subjected to a precipitation experiment with biotinylated compound 4 as described for tubulin in
Materials and Methods. Samples were resolved by PAGE, and PHB2 was detected
by immunoblotting. Lane 1, purified PHB2 from the starting material; lane 2, the
neutravidin biotin-binding matrix alone; lane 3, PHB2 incubated with DMSO and
biotin-binding matrix; lane 4, PHB2 incubated with 5 ␮M compound 4
and biotin-binding matrix; lane 5, PHB2 incubated with 5 ␮M compound 5 and
biotin-binding matrix.
as supporting information on the PNAS web site), and cells
overexpressing MDR1 displayed only 3-fold resistance to compound 3 but were 150-fold resistant to taxol and peloruside A
(data not shown), in agreement with a report that HTI-286 is a
poor substrate for the P glycoprotein pump (3). Cytotoxic,
biotinylated compound 4, but not the control biotinylated compound 5, was able to specifically affinity-purify tubulin (Fig. 2A)
from HeLa cell lysate. Thus, compound 3 retained the functional
properties of the natural product and was an appropriate probe
of hemiasterlin biology.
Compound 3 Is Highly Toxic for C. elegans. To determine the effect
of compound 3 on C. elegans, we examined development and
health of wild-type L1 larvae growing in liquid culture at
various concentrations of compound. Vehicle (DMSO) control
had no effect on worms. The lowest toxic dose of compound
3, 700 nM, produced normal sized, but paralyzed, animals.
Larvae growing at 1 ␮M compound 3 developed into fertile,
dumpy adults (Fig. 3; and see Movies 1 and 2, which are
published as supporting information on the PNAS web site). At
2 ␮M, the dumpy phenotype became more severe, and animals
laid eggs that never developed (embryonic lethality). All
larvae incubated with 5 ␮M compound 3 were paralyzed within
12 h, growth-arrested at the L1 stage, and died within 2– 4 days.
The significantly higher concentration of compound 3 required
to observe a phenotype in C. elegans compared with mamma15524 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0607338103
Fig. 3. When grown in 1 ␮M compound 3, wild-type worms and phb2(ad2154) mutants differ in morphology. (A) Wild-type worms grown for 4
days in normal culture conditions. (B) Wild-type animals grown for 4 days in 1
␮M compound 3 become short, paralyzed, and morphologically abnormal but
lay eggs and reproduce (middle two animals). (C) Phb-2(ad2154) animals
grown for 4 days in 1 ␮M compound 3 develop into healthy adults. Animals to
the right are laying eggs.
lian cells was also observed for other drugs tested (Fig. 4).
Some of the drugs tested, such as colchicines, cytochalasin B,
or dolastatin 15, were not toxic for worms at all.
Characteristics of Mutants Resistant to Compound 3. We next per-
formed a genetic screen for mutants that were resistant to
compound 3. After mutagenesis with ethyl methanesulfonate,
we isolated eight fully penetrant mutants: two dominant and
six recessive. We found 1 ␮M compound 3 to be the most
convenient concentration to assay, because adults reproduced
and offspring of a single animal could be characterized as
either sensitive (paralyzed fertile dumpy) or resistant (healthy
animals). Two animals with strong phenotypes were outcrossed three times and characterized. One carried a dominant
mutation (ad2155dm) that was mapped to a region on the left
arm of chromosome III, between cosmids C32A3 and W03A5.
The only tubulin gene (tbb-2) located in this area was sequenced and found to be wild type. The other animal, phb2(ad2154) carried a recessive mutation and appeared unaffected at 1 ␮M compound 3 (see Movies 3 and 4, which are
published as supporting information on the PNAS web site).
The mutation conferring resistance in phb-2(ad2154) was
identified by standard genetic methods and DNA sequencing,
as described in Materials and Methods. Injecting a PCR fragment containing the wild-type phb-2 gene with 1 kb of
Zubovych et al.
upstream and 369 bp of downstream sequence into phb2(ad2154) worms resensitized them to the drug, confirming
that the mutation in phb-2 was responsible for drug resistance.
This mutation was a G-to-A transversion, resulting in a change
of an evolutionarily conserved glutamic acid to lysine at amino
acid 130 (Fig. 6, which is published as supporting information
on the PNAS web site).
mnDf30 and mnDf96 worms bear a deletion covering the phb-2
gene. Crossing phb-2(ad2154) males with these deficient strains
produced resistant progeny in 1 ␮M compound 3 that were
resistant to the same drug concentrations as their phb-2(ad2154)
Zubovych et al.
parent. Thus, a single copy of the phb-2 gene conferred drug
resistance. Mating phb-2(ad2154) males with each of the other
recessive mutants isolated in our screen produced drug-sensitive
progeny. DNA sequencing revealed that the remaining mutants
were wild type for phb-2 and phb-1.
The phenotype reported for worms in which phb-2 was
targeted by RNAi was similar to that observed in wild-type
worms treated with compound 3 (8). To determine whether
PHB2 was a direct target of compound 3, we incubated the
biotinylated compound 4 with pure recombinant PHB2. Under
conditions where compound 4 was able to affinity-purify
PNAS 兩 October 17, 2006 兩 vol. 103 兩 no. 42 兩 15525
GENETICS
Fig. 4. phb-2(ad2154) worms are resistant to many toxins but not to nocodazole or phalloidin. Approximately 300 wild-type N2 (squares) or phb-2(ad2154)
(triangles) L1 larvae were incubated in drugs at indicated concentrations for 4 days, and the healthy, morphologically normal, mobile adult worms with eggs
were counted. For each concentration, three wells were scored in an experiment. Average values were graphed ⫾ SEM. Similar results were obtained in at least
five independent experiments.
tubulin in vitro (Fig. 2 A), we could detect no interaction with
PHB2 (Fig. 2B).
Phb-2(ad2154) Worms Are Resistant to Other Drugs. We next determined the sensitivity of N2 worms to other tubulin-binding toxins,
both microtubule stabilizers and destabilizers (Fig. 4) by treating L1
larvae with the drugs and scoring their ability to develop normally.
Interestingly, compound 3 and nocodazole exhibited visually similar
effects on wild-type C. elegans, causing them to develop into
paralyzed, dumpy worms (Fig. 3B). However, nocodazole, unlike
compound 3, did not affect fertility and viability; dumpy animals
survived and reproduced at concentrations as high as 200 ␮M.
Cryptophycin-1 acted similarly to compound 3, except that adults
were mobile. All other tested drugs arrested worm development at
larval stages, but animals escaping this arrest grew to healthy, fertile
adults. Thus, compounds having comparable inhibitory actions on
tubulin and mitosis affected C. elegans development distinctly.
To determine whether phb-2(ad2154) conferred resistance to
other drugs, we counted the number of N2 and phb-2(ad2154)
worms grown in the presence of compounds that grew to be
healthy gravid adults (Fig. 4). phb-2(ad2154) worms were resistant to compound 3 but not to nocodazole. phb-2(ad2154)
animals showed similar levels of resistance to compound 3,
dolastatin 10, cryptophycin-1, paclitaxel, peloruside A, vincristine, vinblastine, and phomopsin A (Fig. 4). Because certain of
these drugs are excellent substrates for the P glycoprotein efflux
pump and compound 3 is not, we conclude that this mechanism
of resistance is not contributing to the phenotype. We also tested
toxic compounds that do not target tubulin, camptothecin (a
DNA topoisomerase I inhibitor) and phalloidin (stabilizes actin).
The phb-2(ad2154) mutant was resistant to camptothecin and
fully sensitive to phalloidin, indicating that mutant PHB2 protects worms from drug-induced toxicity specifically. Importantly,
the dominant mutant (ad2155dm) displayed similar drug responses. ad2155dm was fully sensitive to nocodazole and phalloidin and cross-resistant to the remaining compounds (data not
shown), indicating that these two mutants are likely to be on the
same pathway.
Discussion
Determining the target of bioactive compounds and their mechanisms of action is a challenge and requires the combined use of
organic synthesis, biochemistry, and genetics. Even when a target
of a compound has been previously identified, genetics has the
potential to reveal unsuspected aspects of the biological response
and novel mechanisms of resistance. In the case of our derivatives of hemiasterlin, we have mapped resistance to multiple
toxins to PHB2. Several very different functions have been
proposed for PHB2, none of which directly explain the drug
resistance we observe in phb-2(ad2154). PHB2 (also called
BAP37 or REA) is a ubiquitously expressed, evolutionary
conserved protein with high similarity to prohibitin 1 (PHB1,
BAP32). PHB1 and PHB2 form a high-molecular-weight complex and may function together in the mitochondria (9, 10).
Prohibitins have been reported to be localized to the plasma
membrane and the nucleus and in mitochondria with the N
terminus anchored into the inner mitochondrial membrane and
C terminus protruding into the intermembrane space (9, 11–13).
Prohibitins have been suggested to have multiple cellular functions, including cell proliferation, senescence and aging, tumor
suppression, mitochondrial protein folding, and transcriptional
repression. PHB1 was overexpressed in certain carcinomas
(14–16). Although called a ‘‘tumor suppressor,’’ how PHB1
participates in carcinogenesis is unclear.
Genome-wide transcriptional profiling in C. elegans found
phb-2 (T24H7.1) to be expressed at all developmental stages
from egg to adult, with the highest prevalence in embryo and
larvae and lowest in 2-week-old worms (17). Depending on the
15526 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0607338103
degree to which protein expression was reduced, phb-2 RNAi or
phb-1 RNAi worms exhibited embryonic lethality, reduced body
size, and developmental-delay phenotypes (8). Interestingly,
partial down-regulation of prohibitin expression in petunia
resulted in dwarf plants. Cells in the silenced dwarf flowers were
larger in diameter but significantly reduced in number, suggesting an effect on cell proliferation (18). Homozygous PHB2knockout mouse embryos die in early development, whereas
heterozygous animals survive and can reproduce (19).
Our results indicate that a single amino acid change in PHB2
protects C. elegans from some but not all drug effects, because
phb-2(ad2154) animals are fully sensitive to nocodazole and
phalloidin. The phb-2 RNAi phenotype in C. elegans is dumpy
and embryonic lethal, very similar to the phenotype of wild-type
worms grown in compound 3. We could not detect the binding
of biotinylated compound 4 to PHB2, and multiple drug resistance also suggests that PHB2 is not a direct target of compound
3. If PHB2 acts downstream of certain toxins in a mechanism
responsible for cell death, and phb-2(ad2154) is defective in this
response, not all tubulin-binding antimitotics induce cell death
through a common pathway. Alternatively, PHB2 might function
to control the intracellular concentration of certain toxins. If the
recessive mutation in phb-2(ad2154) is a partial loss or change in
function, PHB2 might inhibit drug efflux or degradation mechanisms or activate drug influx. It is not known how the drugs used
in this study penetrate worms. We have not compared the drug
concentrations inside wild-type worms and phb-2(ad2154)
worms, because the differences in effective drug concentrations
are small, and drugs in the mutant may be retained in the
intestinal lumen or excreted into pseudocoelomic fluid. Further
analysis of other mutants, particularly the (ad2155dm) mutant
that shows a similar response, will provide a more comprehensive view of this pathway. We believe that finding the molecular
mechanisms involved may have clinical implications to combat
multidrug resistance.
Materials and Methods
Synthetic Reagents. Hemiasterlin analog compound 3 was synthesized according to published protocols (7). The route used
was unchanged, except that it was initiated with ␤ -(3bromophenyl) isovaleric acid instead of ␤-phenylisovaleric acid.
We synthesized compound 3 instead of compound 2 because
compound 3 was the precursor to biotinylated probe compound
4, which was prepared by coupling a protected form of compound 3 (see supporting information) with BocHN(CH2)6CCH
[2 mol% (t-Bu3P)2Pd, 3 mol% CuI, (i-Pr)2NEt, p-dioxane, rt].
The resulting product was saponified (LiOH, aq MeOH), exposed to excess trifluoroacetic acid (CH2Cl2, rt), and treated
with the N-hydroxysuccinimidyl ester of L-biotin (Sigma, St.
Louis, MO). Inactive control compound 5 was synthesized in a
directly analogous manner. Both reagents were purified to
homogeneity by preparative HPLC (5 ␮M C18, CH3CN兾H2O
gradient) and possessed analytical data in full accord with their
proposed structures.
Cell Lines, Reagents, and Cytotoxicity. All cell lines tested were
purchased from American Type Culture Collection (Manassas,
VA). SK-MEL-5 cells were grown in modified Eagle’s medium
(MEM; Invitrogen, Carlsbad, CA) with 1.5 g兾liter NaHCO3
supplemented with 10% FBS (GIBCO, Carlsbad, CA) and 1 mM
sodium pyruvate (Invitrogen). BSC-1 and MCF-7 cells were
maintained in MEM with 3.5 g兾liter NaHCO3 and 10% FBS,
CHO in Dulbecco’s modified Eagle’s medium (DMEM; Invitrogen) with 5% FBS, 10 mM Hepes, and 2 mM L-proline. HeLa
cells were cultured in DMEM supplemented with 10% FBS, 1
mM sodium pyruvate, and 10 mM Hepes. All culture media
contained 1% of penicillin–streptomycin solution (Invitrogen).
To characterize the antiproliferative effect of drugs, 30,000 cells
Zubovych et al.
Tubulin and in Vitro Polymerization Assays. MAP-enriched tubulin
was purified from four fresh bovine brains (Dallas City Packing,
Dallas, TX) by using a protocol from the Mitchison laboratory
with slight modifications. MAP-enriched tubulin aliquots were
stored in a ⫺80°C freezer until use. Pure tubulin was purchased
from Cytoskeleton (Denver, CO).
Binding to Biotinylated Compounds. After thawing, tubulin aliquots
were centrifuged at 14,000 ⫻ g for 10 min at 4°C to remove
aggregates, and then the supernatant was diluted to 0.5 mg兾ml
final concentration in PEM buffer containing 1% Nonidet P-40
(Calbiochem, San Diego, CA). Compounds 4 or 5 were added at
5 ␮M in a total reaction volume of 200 ␮l in 0.5-ml tubes and
incubated at 4°C overnight on a rotator. Fifty microliters of
Immobilized Neutravidin Biotin Binding Protein beads (Pierce,
Rockford, IL) was added to each tube for 30 min, the reactions
were carefully transferred to 1.5-ml Eppendorf tubes and washed
in incubation buffer seven times at 4°C, and precipitates were
analyzed by SDS兾PAGE. After electrophoresis, proteins were
electrotransferred to 0.45-␮m nitrocellulose membrane (Protran
BA85), immunoblotted with a 1:3,000 dilution of monoclonal
anti-␣-tubulin antibody (Sigma), followed by a 1:3,000 dilution of
HRP-conjugated goat anti-mouse (Bio-Rad, Hercules, CA), and
then detected with ECL reagent (PerkinElmer Life Sciences,
Boston, MA) according to the manufacturer’s instructions.
To prepare HeLa cell lysate, cells were lysed directly on 15-cm
tissue culture plate by adding 4 ml of PEM buffer supplemented
with 1% Nonidet P-40 and protease-inhibitor mixture (Roche,
Indianapolis, IN). The cell lysate was incubated for 1 h on a
rotator at 4°C and then centrifuged at 14,000 ⫻ g for 40 min at
4°C. The supernatant was collected and incubated with biotinconjugated compounds. The ability of the compounds to pull
down tubulin from the HeLa lysate was determined as for
purified tubulin.
Worm Culture. Worms were grown at 20°C by using standard
culture conditions (20) with slight modifications (21). Wild-type
C. elegans var. Bristol N2 was the parent of all mutant strains. For
genetic mapping experiments, strains used were Hawaiian strain
CB4856; CB768 bli-2 (e768) II; CB187 rol-6 (e187) II;
SP543(mnDf30) unc-4(e120)兾mnC1 dpy-10(e128) unc-52(e444)
II; and SP788 unc-4(e120) mnDf96兾mnC1 dpy-10(e128) unc52(e444) II. Worms were fed bacterial strain HB101 (22). To
obtain first-stage synchronous larvae (L1s), eggs were purified by
alkaline hypochlorite treatment (23), rinsed five times in M9
buffer (24), suspended in 4 ml of M9, and placed on a shaker for
13–16 h. Drug tests were performed in 150 ␮l of M9 buffer
containing 5 mg兾liter cholesterol, HB101, and compounds at
room temperature (⬇22°C) in 48-well plates (FALCON; Becton
Dickinson, Franklin Lakes, NJ).
Drugs. Paclitaxel (taxol), nocodazole, vinblastine, vincristine,
camptothecin, phomopsin A, and phalloidin are commercially
available (Sigma). Dolastatin 10 was a kind gift from Gregory
Kalemkerian (University of Michigan Medical Center, Ann
Arbor, MI), cryptophycin 1 was generously provided by Richard
Himes and Gunda Georg (University of Kansas, Lawrence, KS),
and peloruside A was obtained from Jef De Brabander (University of Texas Southwestern Medical Center). Stock solutions
were prepared in DMSO, aliquoted, and stored at ⫺80°C.
Working dilutions were prepared in M9 medium immediately
before use.
Zubovych et al.
Genetic Screen for Compound 3-Resistant Mutants. Late L4 N2
animals were mutagenized by exposure to 47 mM ethyl methanesulfonate for 4 h. Nine-hundred thousand L1 F2 progeny of
mutagenized parental animals were plated in 4 ␮M compound
3, and the worms were scored for body shape and ability to move.
The progeny of candidates were examined for the same criteria,
including the ability to reproduce in the presence of the drug.
Eight animals had a mutant phenotype.
Sequencing. We amplified all 65.7 kb of genomic DNA from the
phb-2(ad2154) mutant from snp uCE2–1658 to snp㛭C29F5[1]
and sequenced the PCR products. We identified the only
mutation present in that region in the T24H7.1 (phb-2) gene.
Complementation Experiments. Cosmids were prepared by using
the Plasmid Isolation Mini kit (Qiagen, Valencia, CA) and
injected, by using standard injection techniques, into phb2(ad2154) at a concentration 20 ng兾␮l with rol-6 DNA as a
marker for transformation at a concentration 50 ng兾␮l (25).
Genetic Mapping of phb-2(ad2154) and Rescue Experiments. By standard polymorphism mapping, a recessive drug-resistance mutation was linked to chromosome II. More precise mapping was
performed by flanking the mutation between bli-2(e768) and
rol-6(e187) visible markers. Briefly, drug-resistant mutant males
were mated with bli-2 hermaphrodites, and drug-resistant bli-2
males were selected and mated with rol-6 hermaphrodites to
obtain drug-resistant triple-mutant animals. CB4856 males were
mated with triple mutants and, from the F2 generation, recombinant Bli-nonRol worms were selected, made homozygous, and
tested for the compound 3-sensitive or -resistant genotype. After
SNP analysis of isolated recombinants, the mutation was mapped
between snp uCE2-1658 and snp㛭C29F5[1]. Mating with two
strains, mnDf30 and mnDf96, each having a deficiency spanning
the region, confirmed the location of the mutation. This 65-kb
region was covered by three overlapping clones, T24H7, F13H8,
and C29F5, and included 19 genes. We sequenced this region
simultaneously with rescue experiments. Cosmid T24H7 rescued
the mutant phenotype, and sequence data identified only a single
mutation present in the 65.7 kb of DNA in the T24H7.1 gene. For
rescue with a genomic fragment, a 2807-bp PCR product spanning the phb-2 locus with 1 kb of upstream sequence and 369 bp
of downstream sequence was amplified from T24H7 cosmid
DNA and injected into phb-2 (ad2154) mutants at a concentration 0.5 ng兾␮l with rol-6 (50 ng兾␮l). The following primers were
used for amplification: forward, CGTGAAGTAGATCGATTCATG; reverse, CTGCATGCTGTATTGATCGA. We isolated 13 independent stable transgenic lines, and 12 of them were
converted to drug sensitivity by the presence of wild-type phb-2;
80–100% transgenic worms of each line displayed dumpy phenotype in the presence of 1 ␮M compound 3.
Drug Toxicity Experiments with Worms. Escherichia coli HB101
from cultures grown overnight in 1 liter of LB broth supplemented with 0.05 g兾liter streptomycin were washed twice with
water, resuspended in 150 ml of M9 buffer supplemented with
cholesterol (5 mg兾liter), and dispensed into 48-well tissue culture
plates to provide food for worms. Compounds were added to the
wells to the desired concentrations. DMSO was always loaded as
a control, and DMSO concentrations were kept below 1.5%.
Approximately 300 synchronized L1 larvae were plated into each
well of 48-well tissue culture plates. After 3–4 days, when all
animals grown in DMSO had become adults, the content of each
well was transferred to an empty Petri dish, and the number of
healthy adult worms was counted.
Photography. Images of worms were taken with a MaxCam CM10
digital camera (Finger Lakes Instrumentation, Lima, NY) on an
PNAS 兩 October 17, 2006 兩 vol. 103 兩 no. 42 兩 15527
GENETICS
were plated per well in 12-well tissue culture plates. The next day,
cells were treated with varying concentrations of compound, and
cell numbers were determined for 4 days by using a Beckman
Cell Counter (Coulter, Miami, FL).
Axiophot microscope (Zeiss, Oberkochen, Germany) using Nomarsky optics and a 5⫻ objective. Supporting movies were
recorded with a KP-160 camera (Hitachi, Tokyo, Japan) at a rate
of 30 frames per second and digitized by using a Matrox RTX-10
video capture board (Matrox Electronic Systems, Dorval, QC,
Canada).
PHB2 Expression, Purification, and Pull-Down Experiments. PHB2
CA). The washed beads were incubated overnight with thrombin
(Amersham) at room temperature in 1⫻ PBS containing 2.5 mM
CaCl2, and then thrombin was inactivated with 1 mM PMSF. The
released PHB2 (0.2 mg兾ml in PBS, with 1% Nonidet P-40 added)
was incubated with biotinylated compounds and neutravidin
beads and analyzed as described for experiments with tubulin,
except that the nitrocellulose membrane was probed with a 1:500
dilution of PHB2 antibody (BioLegend, San Diego, CA).
cDNA in a pCMV6-XL5 vector was purchased from OriGene
Technologies (Rockville, MD). The cDNA was amplified by
PCR using the following primers: forward, containing an EcoRI
restriction site, GATCAGAATTCTAATGGCCCAGAACTTGAAG, and reverse, containing an XhoI restriction site, GATCACTCGAGTCATTTCTTACCCTTGATGA, and inserted
into the pGEX-KG GST-fusion vector. After DNA sequencing
to verify the GST-PHB2 fusion sequence, the plasmid was
transformed into BL21-CodonPlus-RIL cells (Stratagene, La
Jolla, CA). Protein expression was induced with IPTG for 4 h at
room temperature, and the GST-PHB2 from a cell lysate was
bound to glutathione Sepharose (Amersham, San Francisco,
We thank Leon Avery (University of Texas Southwestern Medical
Center) for sharing strains and reagents and for discussion, Anthony
Burgett for the tubulin purification, Audrey Fraser (Wellcome Trust
Sanger Institute, Hinxton, Cambridge, U.K.) for cosmid clones, Boris
Shtonda for help with photography, Gregory Kalemkerian for dolastatin
10, Gunda Georg and Richard Himes for providing cryptophycin-1, and
Jef De Brabander for peloruside A. Some strains used in this work were
supplied by the Caenorhabditis Genetics Center (University of Minnesota, Minneapolis, MN), which is funded by the National Institutes of
Health National Center for Research Resources. This work was supported by National Cancer Institute Grant P01CA95471. This investigation was conducted in a facility constructed with support from
Research Facilities Improvement Program Grant C06 RR-15437.
1. Janin YL (2003) Amino Acids 25:1–40.
2. Anderson HJ, Coleman JE, Andersen RJ, Roberge M (1997) Cancer Chemother Pharmacol 39:223–226.
3. Loganzo F, Discafani CM, Annable T, Beyer C, Musto S, Hari M, Tan X, Hardy
C, Hernandez R, Baxter M, et al. (2003) Cancer Res 63:1838–1845.
4. Poruchynsky MS, Kim JH, Nogales E, Annable T, Loganzo F, Greenberger
LM, Sackett DL, Fojo T (2004) Biochemistry 43:13944–13954.
5. Ayral-Kaloustian S, Zask A (2005) Drugs Future 30:254–260.
6. Gamble WR, Durso NA, Fuller RW, Westergaard CK, Johnson TR, Sackett
DL, Hamel E, Cardellina JH, II, Boyd MR (1999) Bioorg Med Chem 7:1611–
1615.
7. Nieman JA, Coleman JE, Wallace DJ, Piers E, Lim LY, Roberge M, Andersen
RJ (2003) J Nat Prod 66:183–199.
8. Artal-Sanz M, Tsang WY, Willems EM, Grivell LA, Lemire BD, van der Spek
H, Nijtmans LG (2003) J Biol Chem 278:32091–32099.
9. Nijtmans LG, de Jong L, Artal Sanz M, Coates PJ, Berden JA, Back JW,
Muijsers AO, van der Spek H, Grivell LA (2000) EMBO J 19:2444–2451.
10. Tatsuta T, Model K, Langer T (2005) Mol Biol Cell 16:248–259.
11. Fusaro G, Dasgupta P, Rastogi S, Joshi B, Chellappan S (2003) J Biol Chem
278:47853–47861.
12. Ikonen E, Fiedler K, Parton RG, Simons K (1995) FEBS Lett 358:273–277.
13. Terashima M, Kim KM, Adachi T, Nielsen PJ, Reth M, Kohler G, Lamers MC
(1994) EMBO J 13:3782–3792.
14. Asamoto M, Cohen SM (1994) Cancer Lett 83:201–207.
15. Srisomsap C, Subhasitanont P, Otto A, Mueller EC, Punyarit P, WittmannLiebold B, Svasti J (2002) Proteomics 2:706–712.
16. Wang KJ, Wang RT, Zhang JZ (2004) World J Gastroenterol 10:2179–2183.
17. Hill AA, Hunter CP, Tsung BT, Tucker-Kellogg G, Brown EL (2000) Science
290:809–812.
18. Chen JC, Jiang CZ, Reid MS (2005) Plant J 44:16–24.
19. Park SE, Xu J, Frolova A, Liao L, O’Malley BW, Katzenellenbogen BS (2005)
Mol Cell Biol 25:1989–1999.
20. Sulston JE, Hodgkin JA (1988) Methods, the Nematode Caenorhabditis elegans
(Cold Spring Harbor Lab Press, Cold Spring Harbor, NY).
21. Avery L (1993) Genetics 133:897–917.
22. Boyer HW, Roulland-Dussoix D (1969) J Mol Biol 41:459–472.
23. Emmons SW, Klass MR, Hirsh D (1979) Proc Natl Acad Sci USA 76:1333–1337.
24. Brenner S (1974) Genetics 77:71–94.
25. Mello C, Fire A (1995) Methods Cell Biol 48:451–482.
15528 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0607338103
Zubovych et al.
SupplFig1.pzf:Layout 1 - Thu Jun 1 11:18:09 2006
Tubulin Polymerization
absorbance
0.5
Peloruside A
Taxol
0.4
DMSO
3
0.3
500
1000
seconds
Supplementary Figure 1
1500
2000
YEERVLPSIVNEVLKSVVAKFN
YEERVLPSIVNEVLKSVVAKFN
YDERVLPSIVNEVLKSVVAKFN
WEERVLPSICNEVLKGVVAKFN
YDERVLPSIVNEVLKAVVAQFN
Figure 5
-
human
rat
zebrafish
C. elegans
yeast