Kubagawa et al. Supplementary Data.

SUPPLEMENTARY INFORMATION, FIGURE S1.
Figure S1. Sperm location and loss in wild-type and mutant hermaphrodite gonads. The
fluorescent stain, 4',6'-diamidino-2-phenylindole (DAPI), was used to visualize sperm
DNA, which is punctate (arrowheads) in appearance. (a, b) In wild-type gonads, sperm
are located within or adjacent to the spermatheca (SP). Sperm are rarely observed near
the vulva (VU). (c-f) In fat-2(wa17) (c and d) and fat-3(wa22) (e, f) mutant
hermaphrodite gonads, sperm are often found throughout the uterus and near the vulva.
(g, h) In hermaphrodites with defects in the sperm targeting mechanism, Mitotrackerlabeled sperm located near the vulva are often ejected from the uterus into the external
environment during egg-laying. An rme-2 RNAi hermaphrodite is shown (g, h). Scale
bars, 50µm.
SUPPLEMENTARY INFORMATION, FIGURE S2.
Figure S2.
PUFA addition to oocytes promotes sperm recruitment in fat-2(wa17)
hermaphrodites. (a) Transgenic animals expressing vitellogenin::GFP (YP170) in control
(wild-type) hermaphrodites show yolk distribution in the proximal gonad13.
Yolk
complexes are endocytosed by the oocytes closest to the spermatheca (yellow outline).
YP170 persists in yolk endosomes within oocytes and embryos until it is degraded
during embryogenesis. Bodipy-FAs are found throughout the membranes of oocytes
and their precursors (Fig. 2a), suggesting that yolk fats are released from YP170
complexes during or following endocytosis. This observation raises the possibility that
yolk endosomes are not the site where PUFA-derived signals are synthesized. YP170
distribution in fat-2 RNAi hermaphrodites resembles control hermaphrodites, indicating
that PUFA depletion does not prevent yolk endocytosis. (b) In wild-type hermaphrodites,
nearly all MT sperm have migrated to the spermatheca one hour after mating. By
contrast, most sperm (typically greater than 80%) fail to reach the spermatheca during
the same time period in fat-2(wa17) hermaphrodites or fat-2(wa17) hermaphrodites
microinjected with a buffer control. Microinjecting arachidonic acid (20:4n6) through the
vulva into the reproductive tract, where it spreads throughout the uterus, spermatheca,
and proximal gonad, rescues the sperm recruitment defect of fat-2(wa17)
hermaphrodites. Identical results are observed when purified PUFA-containing yolk is
microinjected into the fat-2(wa17) hermaphrodite tract. Because oocytes are the only
cell type that expresses the RME-2 yolk receptor13, we conclude that PUFAs are
sufficient in oocytes to promote sperm recruitment. Gonads are oriented as shown in
Fig. 1a. VU, vulva. Scale bars, 50µm.
SUPPLEMENTARY INFORMATION
MATERIALS AND METHODS
Strains and RNA-mediated interference
Worms were cultured as previously described1, except that E. coli NA22 bacteria2 was
used instead of OP50. NA22 bacteria grow to a higher density than OP50 bacteria on
NGM plates. C. elegans variety Bristol, strain N2 is the wild-type strain. Males were
generated from N2 crosses or by using the fog-2(q71) strain, which segregates “female”
and normal male progeny. fog-2(q71) male sperm are indistinguishable from N2 male
sperm.
The following strains were also used:
SS104 [glp-4(bn2)I], JK816 [fem-
3(q20)IV ], JK2321 [mog-5(q449) unc-4(e120)/mIn1[dpy-10(e128)]II], JK1466 [ g l d 1(q485)/dpy-5(e61) unc-13(e51) I], BX26 [fat-2(wa17)IV ], BX30 [fat-3(wa22)IV], BX24
[ fat-1(wa9) IV ], BX17 [fat-4(wa14)IV ], PD8488 [rrf-1(pk1417)I ], and DH1390 [rme-
2(b1008)IV ]. RNAi was performed at 25°C on either L1 or L4 larva by the feeding
method3. HT115 bacterial feeding strains were obtained from the genome-wide library4.
To confirm that the feeding strains contained the correct genes, we performed PCR on
plasmid preps using gene specific primers internal to the cloned region or sequenced the
cloned region. RNAi phenotypes were compared to those of null mutants to determine
effectiveness. DAPI staining and direct observation using DIC microscopy were used to
determine the locations of self-derived sperm in wild-type and mutant hermaphrodite
gonads. For the RNAi screen, RNAi hermaphrodites were scored for loss of sperm a
day earlier than controls (unc-4 and unc-24 RNAi). Absence of sperm was indicated by
the presence of unfertilized eggs in the uterus and on the plate and accumulation of
oocytes in the proximal gonad.
MitoTracker staining and mating assay
MitoTracker Red CMXRos (Molecular Probes) was used to label male sperm by
modification of a method described in previous studies5,6.
Briefly, MitoTracker was
diluted in DMSO to a 1 mM concentration. Approximately fifty males were placed in 300
µl M9 buffer in a watch glass. MitoTracker was added to a final concentration of 10 µM.
The males were incubated in the dark for 2 hours, then transferred to fresh plates and
allowed to recover overnight. For imaging freshly mated hermaphrodites, approximately
25 males were placed with 6-8 anesthetized hermaphrodites [0.1% tricaine and 0.01%
tetramisole7] on NGM plates containing a 1 cm in diameter drop of bacteria. After mating
for 20-30 minutes, anesthetized hermaphrodites were mounted for microscopy on 2%
agarose pads.
Anesthesia does not affect sperm motility.
MT sperm are
indistinguishable from non-labeled sperm in appearance and motility.
Microscopy and sperm movement analysis
Imaging was performed using a Zeiss Axioskop 2 plus equipped with epi-fluorescence, a
63X objective, MRM Axiocam Hi-Res digital camera, PC computer, and Axiovision
software. To analyze subcellular localization, axial scans were performed, and out-offocus light was removed with deconvolution software (Axiovision). Sperm movement
was analyzed from traces generated from time-lapse videos. Images were taken every
30 seconds. Axiovision software was used to measure distances. Vectorial velocity
toward the spermatheca was measured by creating a straight line through the uterus
from the vulva to the spermatheca. The distance traveled along this line from the
beginning of a sperm trace to the end was divided by time. Negative values indicate
movement away from the spermatheca relative to the starting point. A reversal was
defined as occurring when the angle generated from a sperm trace during three
consecutive time-lapse frames is less than 90 degrees. Sperm traces range from a
minimum of 2.5 minutes to a maximum of 15 minutes. At least 4 videos from different
animals were used for quantitation.
A two sample T-test was used to test for
significance.
Fatty acid supplementation, cholesterol extraction, and starvation
For Bodipy-FA experiments, 200 µM solutions dissolved in DMSO were dropped onto
seeded plates and allowed to dry. L4 stage hermaphrodites were added to the plates
and kept in the dark for 24-48 hours at 20°C. Two BODIPY 500/510 probes (Molecular
Probes) were used, each yielding identical results. The probes differ in the placement of
the fluorescent fluorophore, either in the middle or near the end of the terminal
carboxylate group.
For dietary PUFA supplementation experiments, PUFA stocks were prepared by
diluting fatty acid salts (Nu-Chek Prep, Elysian, MN) to 100 mM in ddH2O as described
elsewhere8,9.
PUFAs were added slowly to cooled NGM media, with stirring, to final
concentrations of 40 and 160 µM. Plates were kept in the dark and seeded with NA22
bacteria after 24 hours. NA22 bacteria and worms were collected and evaluated for lipid
content by Gas Chromatography8. Unlike OP50, the NA22 bacteria does not readily
incorporate PUFAs into its lipids (1-5% accumulation in our experiments).
Sperm
motility was rescued when fat-2(wa17) hermaphrodites were grown on 40 and 160 µM
PUFA-containing plates. Addition of 40 µM linoleic acid and eicosapentaenoic acid to
plates was the minimum amount necessary for complete rescue.
The worms
accumulated the exogenous PUFAs in their membranes to the extent that 2% of their
total fatty acids were derived from dietary supplementation, a two-fold increase relative
to unsupplemented fat-2(wa17) mutants. For arachidonic acid, addition of 160 µM to
plates was the minimum amount necessary for complete rescue. Despite the higher
concentration, the worms still accumulated the exogenous PUFAs to the extent that 2%
of their total fatty acids were derived from supplementation, suggesting that arachidonic
acid may be more unstable or metabolized more rapidly than linoleic acid and
eicosapentaenoic acid.
For cholesterol depletion experiments, worms were grown on plates containing
ether-extracted peptone and agarose without cholesterol addition as previously
described10. Gravid adults were placed on cholesterol-free plates and allowed to lay
eggs.
These F1 hermaphrodites were mated to MT males grown on cholesterol-
containing plates and evaluated for defects in sperm motility. F1 adults were pale, had a
variety of gonadal defects and their F2 progeny did not survive.
For starvation experiments, adult hermaphrodites were washed several times in
M9 buffer and transferred to unseeded plates for 8, 16 or 24 hours. These starved
hermaphrodites were mated to fed MT males. Defects in directional sperm movement
were observed in 16 and 24 hour starved hermaphrodites, but not 8 hour starved
hermaphrodites. Eventually sperm stop moving altogether. These MT sperm will start
moving again upon mating to unlabelled wild-type males, suggesting that males provide
a factor in the seminal fluid that promotes motility.
Yolk complex purification
Vitellogenin-containing or yolk complexes were purified from rme-2(b1008)
hermaphrodites, which accumulate yolk in the pseudocoelom. Worms were collected
from forty 150 cm plates containing approximately 3000 to 8000 adults. The worms
were washed twice in M9 buffer and twice in Buffer A [50 mM Tris-HCl, pH 7.8, 0.2 M
NaCl, 5 mM MgCl2, 0.5 mM CaCl2] (ref. 11).
To release yolk from r m e -2(b1008)
hermaphrodites with as little contaminating material as possible, concentrated worms
were placed in a 6 cm culture plate and macerated using a new razor blade. This
process, which liberates yolk from within the cuticle, was continued until nearly all worms
were cut at least once (15-30 minutes). The yolk and carcass-containing solution was
transferred to a polypropylene tube and vortexed for five minutes. The carcasses were
pelleted in a clinical centrifuge for 8 minutes at 3000 RPM.
The yolk-containing
supernatent was frozen at -80°C and the entire procedure was repeated with 40 fresh
rme-2(b1008) plates four more times. Thawed yolk-containing supernatants (~10ml)
were transferred to a glass vial and sonicated for four 30 second pulses. Next, the yolk
solution was centrifuged at 10,000 RPM for 10 minutes to pellet remaining insoluble
material. In some cases, we concentrated the yolk-containing supernatants by dialysis
using a 30% w/v solution of PEG 8000 in Buffer A. To clean up the samples further, we
set up a discontinuous sucrose gradient containing 2.5 ml 2.0 M sucrose, 2.5 ml 1.35 M
sucrose, 5 ml load (yolk-containing supernatent with 0.25 M sucrose), and 1 ml Buffer A.
The gradient was run for 2 hours at 28,500 RPM (4°C) in an SW41 rotor.
Yolk
complexes stay in the load layer while most contaminating cellular debris accumulates
between the 1.35 M and 2.0 M sucrose layers. Next, yolk from the load layer of gradient
#1 was concentrated using another discontinuous sucrose gradient containing 3 ml 2.0
M sucrose, 5 ml load (from the first gradient), and 3 ml Buffer A. The gradient was run at
28,500 RPM for 8 hr (4° or 20° C) in an SW41 rotor. 500 µl fractions were collected and
evaluated for purity using SDS-PAGE.
Most vitellogenin-containing complexes
concentrate between the load and 2.0 M sucrose layers.
This method yields
vitellogenin-containing complexes of high purity (Fig. 2b). Fractions can be further
concentrated by dialysis in a 30% w/v solution of PEG 8000.
Lipid analysis
Lipids were extracted from frozen yolk pellets with (1:1) chloroform/methanol at –20°
overnight. The extract was washed with 0.2M H3PO4, 1M KCL. Lipids were recovered in
the chloroform phase, dried under N2, and redissolved in chloroform. Yolk lipids and
authentic standards were separated by thin layer chromatography using a two-step
development scheme. The plate was developed with chloroform/methanol/water/acetic
acid (65:43:3:2.5) solvent mixture until the solvent front reached halfway up the plate.
The plate was transferred to the second solvent mixture consisting of
hexane/ether/acetic acid (80:20:1.5) and developed until the solvent reached the top of
the plate. For estimation of cholesterol, lipids were visualized with acidic ferric chloride
solution. For quantification of all other lipid classes, separated lipids were visualized
using I2 vapor, compared to authentic standards and scraped immediately for fatty acid
methyl ester derivitazation. The internal standard (15:0) was added prior to
esterification. Fatty acid methyl esters were prepared with 2.5% methanolic H2SO4 for
analysis by gas chromatography as previously described12.
Microinjection
Arachidonic acid (20:4n6, 500 µM), purified PUFA-containing yolk complexes (250 µg/ml
protein), or PBS buffer was microinjected through the vulva into the fat-2(wa17)
hermaphrodite reproductive tract using a Zeiss Axiovert 200 microscope, hydraulic fine
type micromanipulator, and Narishige IM-30 microinjector. Fluid disperses throughout
the uterus, spermatheca, and proximal gonad. This method causes much less internal
damage than microinjection directly into the proximal gonad. Microinjected animals were
allowed to recover for 2 hours before mating with MT males.
1.
Brenner, S. The genetics of Caenorhabditis elegans. Genetics 77, 71-94 (1974).
2.
Epstein, H. F., Waterston, R. H. & Brenner, S. A mutant affecting the heavy chain
of myosin in Caenorhabditis elegans. J Mol Biol 90, 291-300 (1974).
3.
Timmons, L. & Fire, A. Specific interference by ingested dsRNA. Nature 395, 854
(1998).
4.
Kamath, R. S. et al. Systematic functional analysis of the Caenorhabditis elegans
genome using RNAi. Nature 421, 231-7 (2003).
5.
Hill, K. L. & L'Hernault, S. W. Analyses of reproductive interactions that occur
after heterospecific matings within the genus Caenorhabditis. Dev Biol 232, 10514 (2001).
6.
Kosinski, M., McDonald, K., Schwartz, J., Yamamoto, I. & Greenstein, D. C.
elegans sperm bud vesicles to deliver a meiotic maturation signal to distant
oocytes. Development 132, 3357-69 (2005).
7.
McCarter, J., Bartlett, B., Dang, T. & Schedl, T. On the control of oocyte meiotic
maturation and ovulation in Caenorhabditis elegans. Dev Biol 205, 111-28.
(1999).
8.
Watts, J. L., Phillips, E., Griffing, K. R. & Browse, J. Deficiencies in C20
polyunsaturated fatty acids cause behavioral and developmental defects in
Caenorhabditis elegans fat-3 mutants. Genetics 163, 581-9 (2003).
9.
Kahn-Kirby, A. H. et al. Specific polyunsaturated fatty acids drive TRPVdependent sensory signaling in vivo. Cell 119, 889-900 (2004).
10.
Merris, M. et al. Sterol effects and sites of sterol accumulation in Caenorhabditis
elegans: developmental requirement for 4alpha-methyl sterols. J Lipid Res 44,
172-81 (2003).
11.
Sharrock, W. J., Sutherlin, M. E., Leske, K., Cheng, T. K. & Kim, T. Y. Two
distinct yolk lipoprotein complexes from Caenorhabditis elegans. J Biol Chem
265, 14422-31 (1990).
12.
Watts, J. L. & Browse, J. Genetic dissection of polyunsaturated fatty acid
synthesis in Caenorhabditis elegans . Proc Natl Acad Sci U S A 99, 5854-9
(2002).
13.
Grant, B. & Hirsh, D. Receptor-mediated endocytosis in the Caenorhabditis
elegans oocyte. Mol Biol Cell 10, 4311-26. (1999).
SUPPLEMENTARY INFORMATION, TABLES 1 and 2
S.I. Table S1. Candidates tested in the pilot RNAi screen.
Gene
Description
H02I12.8
Class 4 cytochrome P450 predicted to hydroxylate PUFAs
K08F4.7
Related to glutathione-requiring Prostaglandin D synthases
C01F6.1
Copine family of calcium-dependent phospholipid binding proteins
T28F3.1
Copine family of calcium-dependent phospholipid binding proteins
R107.7
Glutathione S-transferase, pi class
F27C8.6
Related to arylacetamide deacetylase, a putative microsomal lipase
F21G4.2
Member of the ATP-binding Cassette (ABC) transporter family
C10C6.5
Member of the ATP-binding Cassette (ABC) transporter family
F35C11.5
Phospholipase
C45B2.6
Phospholipase
F15B9.5
Phospholipase
W07A8.2
Phospholipase
C42C1.11
Related to Leukotriene A-4 hydrolase
M106.3
Oxidoreductase
C01G8.3
Related to short-chain alcohol dehydrogenases
K06G5.2
Cytochrome P450
ZK177.5
Cytochrome P450
K09D9.2
Cytochrome P450
C06B3.3
Cytochrome P450
C01G6.6
Cytochrome P450 oxidoreductase
F41C3.3
Fatty acid elongase
T06E8.1
Fatty acid elongase
C30G12.2
Short chain-type dehydrogenase
R01H2.3
Member of the low-density lipoprotein receptor family
F53C3.13
Lipid Phosphate Phosphatase
T28D9.3
Lipid Phosphate Phosphatase
T06D8.3
Lipid Phosphate Phosphatase
F11G11.1
Glutathione S-transferase
F11G11.2
Glutathione S-transferase
C07D8.6
Reductase
C01G5.5
Reductase/dehydrogenase
T10B11.2
Sphingosine kinase
F11A6.2
Phospholipid scramblase
C23H3.4
Serine Palmitoyl Transferase
C52E12.3
Sugar and small molecule transporter
These genes have mRNAs expressed in oocytes based on genome-wide DNA microarray
and in situ hybridization studies (see text for references). Descriptions are based on
BlastP searches.
S.I. Table S2. Human gene classes implicated in eicosanoid signaling are present in the C.
elegans genome.
Human Gene Class
Role in eicosanoid signaling
Genome*
Sperm
Recruitment
C. elegans
Apolipoprotein B-100
PUFA transport
Yes
Yes
LDL receptor
PUFA transport
Yes
Yes
Phospholipases A and C
PUFA hydrolysis
Yes
N.D.**
Phospholipase activating
protein (PLAP)
PUFA hydrolysis
Yes
N.D.
Cyclooxygenase
Conversion of AA into prostaglandin H2
No
No
Prostaglandin E synthase
Conversion of prostaglandin H2 into E2
No
No
Prostaglandin D synthase
Conversion of prostaglandin H2 into D2
Yes
Yes
Prostacyclin synthase
Conversion of prostaglandin H2 into I2
No
No
Thromboxane synthase
Conversion of prostaglandin H2 into TXA2
Yes
N.D.
Cytochrome P450
epoxygenase
Conversion of AA into epoxides
Yes
N.D.
Epoxide hydrolase
Conversion of epoxides into diols
Yes
N.D.
Cytochrome P450, class 4
Omega hydroxylation of PUFAs,
prostaglandins, and leukotrienes
Yes
Yes
Lipoxygenase
Conversion of AA into LTA4
No
No
5-lipoxygenase-activating
protein (FLAP)
Conversion of AA into LTA4
No
No
Leukotriene A4 hydrolase
Conversion of LTA4 into LTB4
Yes
N.D.
Leukotriene C4 synthase
Conversion of LTA4 into LTC4
No
No
Gamma-glutamyl
transpeptidase
Conversion of LTC4 into LTD4
Yes
N.D.
ABC transporter, subfamily C
PUFA and eicosanoid transport
Yes
Yes
Prostaglandin transporter
Prostaglandin transport
Yes
N.D.
†
Prostaglandin receptor
Signal transduction
No
No†
Leukotriene receptor
Signal transduction
No†
No†
* Reciprocal BlastP searches and functional data, where applicable, were used to identify putative
homologs. The vitellogenins are thought to be homologous to Apolipoprotein B-100.
** The presence of unesterified PUFAs in mature yolk complexes is suggestive of phospholipase
activity in the intestine or during yolk transport.
†
Although the C. elegans genome does not contain clear homologs of eicosanoid receptors, it
does contain G protein coupled and olfactory receptors generally similar in structure. Several
of these receptors are expressed during spermatogenesis. N.D., not determined.
SUPPLEMENATRY INFORMATION, Video Legends
Video 1. Time-lapse video showing wild-type MT sperm movement within the uterus of a
wild-type hermaphrodite. This video was created shortly after mating. The first 20
minutes were generated from images taken every 30 seconds. The last 40 minutes
were generated from images taken every 2 minutes. Although over 95% of sperm
accumulate at the spermatheca-uterine valve, few sperm enter the spermatheca due to
blockage of the valve during mounting. For tracing sperm paths within the uterus,
hermaphrodites containing fewer sperm are optimal (Quicktime, 3.3MB).
Video 2. Time-lapse video showing wild-type MT sperm movement within the uterus of a
fat-2(wa17) hermaphrodite. This 20 minute video was created shortly after mating.
Images were taken every 30 seconds. Notice how sperm tend to move in circles. Some
sperm reach the spermatheca over time (typically less than 20% of the total
inseminated), likely due to remaining ∆12-desaturase activity and random movement.
Identical results are observed in longer videos. FC indicates a change in the focal plane
(Quicktime, 1.6MB).
Video 3. Time-lapse video showing wild-type MT sperm movement within the uterus of
an rme-2(b1008) null hermaphrodite. This 28 minute video was created shortly after
mating. Images were taken every 30 seconds. Some sperm reach the spermatheca
over time (typically less than 10% of the total inseminated), likely due to RME2–independent PUFA transport to oocytes and random movement (Quicktime, 2.7MB).
Video 4. Time-lapse video showing wild-type MT sperm movement within the uterus of
an rme-2 RNAi hermaphrodite. This 25 minute video was created one hour after mating.
Images were taken every 30 seconds. Even after one hour, the time it takes for nearly
all sperm to reach the spermatheca-uterine valve in wild-type hermaphrodites (Video 1),
a small percentage of sperm are near the spermatheca. Notice how numerous sperm
near the vulva have been ejected into the external environment (Quicktime, 2.1MB).