Supporting Online Material for

Transcription

1 Supporting Online Material for Resolving Mechanisms of Competitive Fertilization Success in Drosophila melanogaster Mollie K. Manier, John M. Belote, Kirstin S. Berben, David Novikov, Will T. Stuart, Scott Pitnick* *To whom correspondence should be addressed. This PDF file includes Materials and Methods SOM Text Figs. S1 to S13 Table S1 References Published 18 March 2010 on Science Express DOI: /science Other Supporting Online Material for this manuscript includes the following: (available at Movies S1 to S6

2 MATERIALS AND METHODS Construction of the fluorescently tagged Protamines A and B Because of the 1.8 mm length of Drosophila sperm tails (S1), quantifying sperm numbers, position and behavior in vivo is not possible with sperm tail labeling as previously done with GFP-tagged don juan transgenes (S2-S4). To make transgenic flies whose sperm heads express a fluorescent tag, we labeled Mst35Ba and Mst35Bb (also known as CG4479 and CG4478) that encode similar sperm-specific chromosomal proteins called ProtamineA and ProtamineB, respectively. These transcription units are adjacent and oriented in the same direction and are located in region 35B6 on chromosome arm 2L (S5). During the late stages of Drosophila spermiogenesis, sperm nuclei elongate and condense into needle-shaped structures, accompanied by a replacement of at least some histones by these protamines (S5). A BAC clone (RPCI C.24) of the genomic region containing the Mst35Ba (ProtamineA) and Mst35Bb (ProtamineB) genes was obtained from Children's Hospital Oakland Research Institute (S6) and DNA prepared using a Qiagen Plasmid Midi-Prep Kit (Qiagen Inc., Valencia, CA, USA). The published sequence of this clone (accession number AC092246, GenBank, predicted a 4.2 kb KpnI-PstI fragment containing the ProtamineB gene region, with 1.3 kb of sequence upstream of the transcription start site and 1.2 kb of sequence downstream of the polyadenylation site. BAC DNA was therefore treated with these enzymes and the 4.2 kb fragment gel-purified and ligated into pbluescriptks+ (Stratagene, La Jolla, CA, USA) to obtain pbs/protb4.2kp. Sitedirected mutagenesis, using primers NdeMut-1 = 5' CGCCACAAGCGCCGACGCATATGCAAGTAATACTG 3', and NdeMut-2 = 5' 2

3 CAGTATTACTTGCATATGCGTCGGCGCTTGTGGCG 3', was carried out to generate an NdeI restriction site at the 3' end of the ProtamineB coding region to yield plasmid pbs/protb4.2ndekp. A DsRed-Monomer NdeI cassette was made by PCR amplifying the coding region of DsRed-Monomer from pdsred-monomer (Clontech Laboratories, Inc., Mountain View, CA, USA) using the primers Nde-DsRedM5 = 5' CCATATGCATGCATGGACAACACCGAGGAGGTCATC 3' and Nde-DsRedM3 = 5' TCATATGTCTACTGGGAGCCGGAGTGGCGGGC 3'. The 692 bp product was cloned into the TOPO cloning plasmid pcr2.1 TOPO using the TOPO TA cloning kit (Invitrogen, Carlsbad, CA, USA) to give ptopo/dsredm-nde-1. This clone was sequenced to confirm that no PCR-induced mutations had occurred, and then the DsRed-Monomer sequence was cut out with NdeI and ligated into the NdeI site of pbs/protb4.2ndekp. Orientation of the inserted DsRed-Monomer sequence was checked by digestion with StuI and EcoRI. The 4.2 kb insert of one correct clone, pbs/protb4.2kp-nde-dsred13, was cut out with KpnI and BamHI and ligated into the transformation vector pw8 to give pw8/protb-dsredm (Fig. S1). To create GFP-tagged ProtamineB, an EGFP NdeI cassette was made by PCR amplifying the coding region of EGFP from pegfp (Clontech Laboratories, Inc.) using the primers NdeGFP5B = 5' TGCATATGCAAGGTGAGCAAGGGCGAGGAGCTGTTCACC 3' and 3'GFP-Ndejb = 5' TCCATATGCAGAGGTTTTCACCGTCATCACCGAAACG 3'. The 875 bp product was cloned into the TOPO cloning plasmid pcr2.1 TOPO using the TOPO TA cloning kit (Invitrogen) to give ptopo/gfpnde3, which was sequenced to confirm that no mutations had occurred. The NdeI EGFP cassette was cut out and cloned into the NdeI site of pbs/protb4.2ndekp to give pbs/protb-gfp4. Orientation of the inserted EGFP cassette 3

4 was checked by digestion with EcoRI, and the 5.1 kb KpnI/BamHI fragment was cut out and ligated into pw8 to give pw8/protb-gfp1 (Fig. S1). In order to create EGFP- and DsRed-tagged ProtamineA constructs, BAC clone RPCI C.24 was digested with NruI and Asp700 (an XmnI isoschizomer) and electrophoresed on a 0.7% agarose gel. A 6.0 kb fragment containing the ProtamineA coding region with 4.8 kb of sequence upstream of the transcription start and 0.5 kb downstream of the polyadelylation site, was gel purified using the QIAQuick gel extraction kit (Qiagen) and made blunt-ended using the Stratagene Polishing Kit (Stratagene). This DNA was then ligated into the HinCII site of pbluescript KS+ to give pbs/prota6.0na13. Site-directed mutagenesis, using primers ProtAMut-1 = 5' CGCCACAAGCGCCGACGCATATGCCAGCAATACTGAG 3', and ProtAMut-2 = 5' CTCAGTATTGCTGGCATATGCGTCGGCGCTTGTGGCG 3' was carried out to generate an NdeI restriction site at the 3' end of the ProtamineA coding region to yield plasmid pbs/prota6.0na-nde2. The NdeI DsRed-Monomer cassette from ptopo/dsredm- Nde-1 was inserted into the NdeI site of pbs/prota6.0na-nde2 to generate pbs/protana/ DsRedM20. The orientation of the inserted cassette was checked by cutting with StuI and EcoRV. The 6.7 kb KpnI-BamHI fragment of this clone was subcloned into pw8 to give pw8/prota-dsredm3. To create an EGFP tagged ProtamineA construct, the NdeI EGFP cassette of ptopo/gfpnde3 was ligated into the NdeI site of pbs/prota6.0na-nde2 to give pbs/prota-egfp6, and the orientation checked by cutting with EcoRI. The 6.9 kb KpnI/BamHI insert was cut out and subcloned into pw8 to give pw8/prota-gfp8 (Fig. S1). Both the GFP- and RFP-labeled constructs give strong signal in the heads of mature sperm and are easily visualized in the female sperm-storage organs following mating (Fig. S2). We examined sperm cysts in testes and compared slide preparations of fixed and 4,6-4

5 diamidino-2-phenylindole (DAPI)-stained sperm to confirm that every sperm manufactured by males is fluorescently labeled. Transformed lines were backcrossed six generations to the LHm wild type stock and fixed. LHm is a genetically variable population (S7) that was obtained from A. Chippindale (Biology, Queens University) and has been maintained since its inception as an outbred population in cages (N > 1000 individuals) with overlapping generations. Maintenance of stocks All stocks were maintained at ambient room temperature and natural photoperiod on standard corn meal-molasses-yeast-agar medium in half pint milk bottles with a starting density of 30 flies. Adults were transferred to new bottles every three days. Fitness assays of transgenic lines To ensure that ejaculates of transgenic males are not dysfunctional, we conducted separate experiments to quantify (i) productivity of females following a single insemination (i.e., number of adult progeny produced over 20 days in five successive vials), (ii) first-male sperm precedence (ejaculate defense; P 1 ) and (iii) second-male sperm precedence (ejaculate offense; P 2 ). There were three treatments per experiment, with LHm, GFP or RFP test males. All females used were virgin LHm, and all test males in experiments (ii) and (iii) were competed against a standard LHm-B male (bearing a dominant brown-eye mutation, allowing paternity assignment). All three experiments were conducted simultaneously, with all females and competitor males randomly assigned to both experiments and treatments within experiments. This experimental design allowed us to forego measuring body size of all but test males, for 5

6 which dry mass was determined to the nearest 1.0 µg and entered as a covariate in all statistics. For (i), N = 33 per treatment, and for (ii) and (iii), N = 40 per treatment. Over all experiments, data for 24/219 females were excluded from statistical analyses, because females were killed, lost, did not remate before 10 days, or produced few to no progeny; there were no treatment biases with exclusion. Proportions of progeny from the first male (P 1 ) and second male (P 2 ) were arcsine squareroot transformed and analyzed with the number of progeny produced prior to remating as a covariate. Single-insemination productivity ANCOVAs comparing cumulative progeny produced reveal no significant difference between LHm, GFP and RFP males throughout the first 9 days. Mates of LHm males produced significantly more progeny than mates of GFP or RFP males only after 14 days (well past the time of remating by most females), and continuing through day 20 (Fig. S3). Differences observed could be due to greater sperm numbers per ejaculate, greater sperm use efficiency by females and/or greater sperm longevity with inseminations by LHm versus GFP and RFP males. Sperm defense GFP and RFP males performed equally well to one another at first male sperm precedence, but both sired a significantly lower proportion of progeny than did LHm males (Fig. S4). It is important to note that P 1 (and P 2 ) is not interpretable without knowledge of the number of sperm competing. In fact, females waited significantly longer to remate following insemination by an RFP male relative to either LHm or GFP males (Fig. S5; Fisher s PLSD, P 6

7 < 0.05). As a consequence, females produced significantly more progeny prior to remating following insemination by RFP males (with no significant difference between LHm and GFP males; Fig. S6). Summing progeny produced prior to remating with first-male progeny sired after remating revealed RFP males have the highest total competitive reproductive success (sperm defense) and GFP males the lowest (Fig. S6). Sperm offense Both LHm and GFP males exhibited extremely (and equally) high second male competitive fertilization success when competing against LHm-B males. Sperm of RFP males competed significantly less well (Fig. S7), although a mean P 2 for RFP males of 76% is not much below the typically reported average for studies with D. melanogaster. In conclusion, a diversity of indices of male fitness relevant to sperm/ejaculate function reveal that the GFP and RFP populations perform equally well to one another and to the LHm population in some assays, and better than or worse than in others. All lines fall within normal variation observed in the literature for all assays. In addition, some of the same fitness parameters were measured in the time series experiment and failed to show any differences between the GFP and RFP lines (see below). Time series experiment Virgin LHm females were mated to one virgin male and provided a 6 hr long opportunity to remate to a second virgin male. Females failing to remate were provided additional opportunities on days four and five. Females and males were randomly assigned among thirteen experimental treatments that were flash frozen in liquid nitrogen at different time 7

8 points after the start of the second mating (ASM): 0 min, 5 min, 10 min, 15 min, 30 min, 45 min, 60 min, 90 min, 2 hr, 8 hr, 24 hr, 48 hr, and 72 hr. In each group, half of the females were mated first to a GFP male, and then remated to an RFP male, and in the other half, the mating order was reversed. The entire experiment was replicated twice, though the first replicate lacked the 90 min, 48 hr and 72 hr time treatments (sample sizes shown in Table S1). All flies were 3-4 days old at time of their first mating. For all matings, pairs were aspirated into plastic vials with 1.5 cm of medium and observed continuously. Females were transferred to fresh food vials the afternoon before the remating. For time treatments below 30 min ASM, pairs in copula were flash frozen, and for all other treatments, copulation duration was recorded. Females or pairs were frozen by holding the bottom half of the vial in liquid nitrogen for seconds. Vials were then stored at -20 C until dissection. For time treatments at or above 90 min ASM, females were transferred by aspiration within 30 minutes after the end of copulation to a fresh food vial in which she was frozen at the appropriate time. Eggs in the vial were counted when the female was dissected. No females laid eggs prior to 2 hours ASM. Females with no prior progeny from the first mating were not given the opportunity to remate. Females were excluded from analyses if copulations were of normal duration yet no sperm was transferred. Thawed females were dissected using fine forceps by removing reproductive tracts into a drop of phosphate-buffered saline (PBS) on a glass slide, being careful not to compress or distort the tissue. The lower reproductive tract was removed without disturbing the sperm within by gently severing the ovipositor from surrounding cuticle and longitudinally opening the mid-ventrum, then severing digestive tract close to the anus and severing the ovaries at the base of the lateral oviducts. To facilitate clearer visualization of sperm within the seminal 8

9 receptacle (SR), the fine tracheoles binding loops of the long tubule were gently disrupted using insect pin (size 000) in a vise handle. Tracts were individually mounted under a glass coverslip and sealed twice with paper cement, which does not affect fluorescent signal. Slides were stored at 4 C until counting, and all counts were made within 3-4 days of dissection. Mounted preps were viewed at 400X on an Olympus BX60 compound microscope (Olympus America Inc., Center Valley, PA, USA) with an X-Cite Series 120 fluorescent lamp (EXFO Life Sciences, Mississauga, Ontario, Canada) and multiband GFP-DsRed-A filter set (Semrock, Rochester, NY, USA). Individual GFP and RFP fluorescent sperm were counted in the bursa, proximal and distal portions of the SR, spermathecal ducts and spermathecal capsules. The proximal and distal portions of the SR were discriminated by the point at which lumen diameter increases from narrow to wide (S8). Sperm were counted using a multiple tally counter, and at least one voucher digital photo was taken of each slide using a DP71 digital camera (Olympus America Inc.). Sperm counts were square root transformed. In order to test for statistical differences attributable to mating order (GFP first or RFP first), the data for each region were individually fit to a quadratic regression model, with transformed counts as the dependent variable and logtransformed minutes ASM (time) and time 2 as the independent variables. We then compared the model coefficients between both mating orders using a two-sample t-test. The same approach was used to evaluate differences between replicates for each region of the tract. For the first replicate, sperm numbers did not differ between GFP-RFP and RFP-GFP male mating orders for all time treatments and regions of the reproductive tract. For the second replicate, ejaculates at 15 min ASM from RFP males were significantly larger (mean ± se = 1682 ± 94 sperm) than those from GFP males (1060 ± 59 sperm; t = -5.52, P < ). 9

10 Nevertheless, male order did not influence the timing of key events across replicates (Figs S8- S10). We also examined differences in performance of GFP and RFP males in several aspects of the dataset to supplement results obtained from the fitness assays reported above. Although females waited significantly longer to remate with RFP males in the fitness assay (Fig. S5), we found no difference in the time series experiment in average remate day for females remated to GFP versus RFP males (t = 0.43, P = 0.66). GFP males in the fitness assays also had a higher proportion of second male paternity (Fig. S7), but we again found no difference in P 2 for GFP and RFP males (t = 0.22; P = 0.83) in the time series experiment, and they produced equal numbers of progeny (t = 0.03; P = 0.98). When we compared second male sperm counts in the reproductive tract pooled across the time points between the peak of transfer and ejection (15 to 45 min ASM; i.e., sperm transferred by second male), we found that RFP males transferred more sperm in Replicate A (though to a lesser degree: t = 2.58, P = 0.015, N = 33), as well as in Replicate B (t = 4.60, P < , N = 99). When we compared second-male sperm counts in storage pooled across the time points between the peak of storage and when egg laying begins (45 min to 8 hr ASM; i.e., sperm stored by second male), we also found that females stored significantly more RFP sperm in Replicate B (t = 6.38, P < ) but not Replicate A (t = 1.93, P = 0.06). Consistent with this result, more resident sperm were displaced from storage into the bursa at 45 and 60 min ASM when the second male was RFP than when GFP in Replicate B (t = 3.81, P = ) but not in Replicate A (t = 0.82, P = 0.43). However, as mentioned above, there was no difference in P 2 for GFP and RFP males (as recorded from Replicate B only), even though RFP males were transferring, storing and 10

11 displacing more sperm. When we examined line differences in the proportion of second male sperm (S 2 ) in storage between 45 and 8 hr ASM, we found a difference in the spermathecae (t = 3.00, P = ) but not in the SR (t = 1.09, P = 0.28), suggesting that sperm are displaced disproportionately more from the spermathecae than from the SR. P 2 is highly associated with S 2 in the SR but not the spermathecae, explaining the observed differences in numbers of GFP and RFP sperm stored but not used for fertilization. These results suggest that, although larger ejaculates displace more sperm, they do so differentially in the two types of sperm-storage organs and in a way that does not necessarily translate to an advantage in paternity. Sperm ejection To establish whether females eject excess and displaced sperm from their reproductive tracts, up to five females were transferred without anesthetization to each glass cube - constructed from 22 mm coverslips - after an initial mating to a GFP or RFP male, and then again after a second mating three days later to male of the opposite genotype. Females were aspirated into the cubes within 30 minutes after the end of copulation and remained in the cubes for three hours, after which they were saved in a fresh media vial for the second mating. All matings occurred in individual vials. We dismantled cubes and examined coverslips at X under epifluorescence as described above for sperm counts. To more carefully quantify ejection behavior, a subsequent experiment was conducted in which virgin LHm females were mated to GFP males, noting the time and duration of copulation. After copulation, each female was transferred to a well of a glass 3-well spot plate, which was covered with a glass coverslip secured with clay. Females were monitored for ejection every ten minutes for up to four hours, and timing of ejection was recorded. Ejected 11

12 masses were slide-mounted in PBS, and presence of sperm was verified under fluorescence at 400x. Motility To determine whether motility of first male sperm stored in the SR changes upon remating with a second male, an experiment was conducted with females randomly assigned to one of four experimental treatments: (i) dissection 60 min after start of first mating (N = 24); (ii) dissection 5 min after start of second mating (N = 23); (iii) dissection 60 minutes after the start of second mating (N = 20); and (iv) females not given the opportunity to remate but dissected on same timescale as those that did remate (Control; N = 18). GFP males were used for the first mating, RFP for the second mating. Females remated 3-6 days after first mating, and copulation duration was recorded for first and second matings. Females were anesthetized with CO 2 and their reproductive tracts were removed as described above and mounted under a coverslip in 20 µl (to standardize tract compression) of Grace s Supplemented Insect Medium (Invitrogen) at room temperature. Ten second long movies (MPEG-1) were recorded within 3-7 minutes after anesthetization on an Olympus DP71 digital camera using DPController Software version (Olympus America Inc.). To establish whether removal of the female reproductive tract into saline generated any physiological effects that alter the normal distribution of sperm, we compared the numbers and distributions of sperm in all areas of the reproductive tract in females that were flash frozen (time series experiment) with those anesthetized with CO 2 before live dissection (motility experiment). We found only one significant difference (after statistical correction for multiple comparisons) in 30 comparisons of counts of first or second male sperm in five different 12

13 regions of the reproductive tract (bursa, proximal SR, distal SR, spermathecal ducts and capsules) across three different time points (0 min ASM, 5 min ASM and 60 min ASM). There were significantly fewer second male sperm in the bursa at 5 min ASM in females that were dissected after CO 2 anesthetization than in females that had been flash frozen (t = 3.32; P = ; Bonferroni corrected alpha = ). These results suggest that males in the times series experiment were transferring more sperm at 5 min ASM than in the motility experiment. Alternatively, since there were fewer flies to observe, we may have been more accurate in catching the start times of copulations. In either case, we do not find support for the concern that distributions of sperm may be shifting in response to dissection after CO 2 anesthetization. In addition, the length of the sperm flagellum (~1.8 mm) makes it difficult to dislodge sperm completely from its original position. In the occasions in which a storage organ was ruptured and some sperm spilled out, we could always unambiguously determine the origin of the displaced sperm based on where their flagella remained under DIC light microscopy. It is, of course, impossible to compare the motility of sperm in a dissected tract with those residing in an intact female. We captured movies of the distal SR for all treatments, as well as the bursa for treatments involving females dissected after the second mating. Each movie was converted to AVI using VirtualDub (v1.9.8, VirtualDub.org) and imported into NIH ImageJ (v. 1.42q, National Institutes of Health, Bethesda, MD, USA, as a grayscale stack, which was then inverted from a dark background to a light background. We measured slice-by-slice instantaneous linear velocities (µm/s) for 10 sperm per stack using the Manual Tracking plugin for ImageJ (available at and logtransformed the velocities. Sperm counts of the entire tract were obtained for each female as 13

14 well, as described above. Average instantaneous velocities were calculated per tracked sperm, which were averaged per female. We used analysis of variance with number of sperm in the seminal receptacle as a covariate to test for significant differences between select treatment groups. Comparing sperm velocities in remated females at 5 min ASM with those that did not remate (treatment ii with iv) tests for an effect of early events in remating (but with few-to-no second male sperm present). Comparison between remated females at 60 min ASM and those that did not remate (treatment iii with iv) also tests for an effect of remating, but in this case a full second copulation and the full complement of second male sperm present. A difference between 5 min ASM and 60 min ASM (treatment ii with iii) would indicate temporal effects post-mating, and comparing females dissected 60 min after the first mating and control females (treatment i with iv) would reveal whether motility changes after several days of storage (and with concomitant reduced sperm density). Finally, comparison females dissected 60 min after the first mating with 60 min after the second mating (treatment i with iii) examines the effect storage and remating on first male sperm motility while holding the sperm density in the SR approximately constant. We also quantified motility as the proportion of motile sperm in storage, by examining each movie visually, and assigning a score of 1 through 5: motility scored as 1 if <10% of sperm were motile, 2 for 10-30% motile, 3 for 30-60% motile, 4 for 60-90% motile, and 5 for >90% motile. Differences among all treatments in motility scores were statistically analyzed using Kruskall-Wallis tests. 14

15 Sperm proportions and offspring sired The relationship between the proportion of competing sperm in storage and competitive fertilization success was examined using females in the 24, 48 and 72 hr ASM treatments of the time series experiment (above). After remating, females were transferred to fresh, yeasted food vials daily until freezing. All adult sons rearing from these vials were dissected 3 days post-eclosion, with a single testis and seminal vesicle from each examined under epifluorescence to determine sperm color (GFP or RFP). These data were used to calculate the proportion of offspring sired by the second male (P 2 ). We also counted numbers of green and red sperm in the female reproductive tract at the time of freezing (i.e., immediately following the production of the progeny assayed for paternity) in order to calculate the proportional representation of second male sperm (S 2 ) in each region of the reproductive tract, as described above. We discarded paternity data from females with less than five sons. We analyzed the relationship between P 2 and S 2 separately for the seminal receptacle and spermathecae to examine any differences in sperm use between the two types of storage organ. Simple linear regressions with both variables arcsine square root transformed were used to test for a significant (non-zero) relationships. If found, we proceeded to quantify the slope of the relationship using reduced major axis regression analysis in Matlab (v.r2007a, The MathWorks, Inc., Natick, MA, USA). Several alternative hypotheses of sperm use were tested by statistically comparing the slope of the reduced major axis regression of P 2 on S 2. If sperm from both males are used for fertilizations in direct proportions to their prevalence in storage, indicating a fair raffle process, the slope of the regression between S 2 and P 2 will be 1. However, if sperm use is disproportionately biased toward one male, indicating a loaded raffle, the slope will be 15

16 significantly less than 1 (favoring the second male) or greater than 1 (favoring the first male). Deviation from a fair raffle (b = 1) was examined using the test statistic provided by Clarke (S9) and his equation 5.1 to calculate degrees of freedom. 16

17 FIGURE LEGENDS Fig. S1. Fluorescently-tagged Protamine A and Protamine B gene constructs. The gray boxes are the coding regions. The green and red rectangles represent the EGFP and DsRed-Monomer coding regions, respectively, that are inserted into the NdeI site. Fig. S2. (A, C) GFP- and (B, D) RFP-labeled sperm heads in the female s (A, B) seminal receptacle and (C, D) spermathecae; (E) seminal receptacle of a female initially mated to a GFP male, then remated to an RFP male. Fig. S3. Cumulative number of progeny produced over time by LHm females following a single insemination by a GFP, RFP or LHm male. F and P statistics are from ANCOVAs of cumulative progeny at respective time points. Fig. S4. Proportion 1st male paternity. Fig. S5. Number of days before female remating. Fig. S6. Total first male competitive reproductive success. Fig. S7. Proportion of second male paternity. Fig. S8. Average number of sperm in the bursa at different time points after the start of remating (ASM) for two experimental replicates (A and B), for GFP first male (dark green), RFP first male (orange), GFP second male (green) and RFP second male (red). Error bars represent standard error. Fig. S9. Average number of sperm in the seminal receptacle at different timepoints after the start of remating (ASM) for two experimental replicates (A and B), for GFP first male (dark green squares), RFP first male (orange diamonds), GFP second male (green triangles) and RFP second male (red circles). Error bars represent standard error. 17

18 Fig. S10. Average number of sperm in the spermathecae at different timepoints after the start of remating (ASM) for two experimental replicates (A and B), for GFP first male (dark green squares), RFP first male (orange diamonds), GFP second male (green triangles) and RFP second male (red circles). Error bars represent standard error. Fig. S11. Average arcsine square root proportions of first-male sperm in the proximal (blue) and distal (yellow) at various timepoints after the start of remating (ASM). Error bars represent standard error. Fig. S12. Relationship between the proportion of resident sperm displaced from storage and size of the second male s ejaculate at 45 min and 60 min ASM. 45 min ASM: y = x , P = 0.012; 60 min ASM: y = x , P < Fig. S13. Mean motility score and velocity (µm/s) of first-male sperm 60 min after first mating, 5 min after second mating, and 60 min after second mating. Control females did not remate. Error bars represent standard error. 18

19 Fig. S1. Fig. S2. 19

20 Fig. S3. Fig. S4. Proportion 1st male paternity Fig. S5. Number of days before female remating Fig. S6. Total first male competitive reproductive success Fig. S7. Proportion of second male paternity 20

21 2000 Replicate A m 5m 10m 15m 30m 45m 60m 90m 2h 8h 24h 48h 72h 0m 5m 10m 15m 30m 45m 60m 2h 8h 24h Number of sperm Fig. S8. Replicate B Time after start of remating (ASM) 21

22 m 5m 10m 15m 30m 45m 60m 90m 2h 8h 24h 48h 72h 0m 5m 10m 15m 30m 45m 60m 2h 8h 24h Number of sperm Fig. S9. Replicate A 400 Replicate B Time after start of remating (ASM) 22

23 m 5m 10m 15m 30m 45m 60m 90m 2h 8h 24h 48h 72h 0m 5m 10m 15m 30m 45m 60m 2h 8h 24h Number of sperm Fig. S10. Replicate A 200 Replicate B Time after start of remating (ASM) 23

24 Fig. S11. 0m 5m 10m 15m 30m 45m 60m 90m 2h 8h 24h 48h 72h Arcsine sqrt proportion first-male sperm Time after start of remating (ASM) 24

25 Arcsine sqrt proportion resident sperm displaced Fig. S min ASM R² = min ASM R² = Number of sperm transferred by 2nd male 25

26 Fig. S13. Sperm velocity (μm/s) Motility score 60 min Mate 1 5 min Mate 2 60 min Mate 2 Control 26

27 Table S1. Sample sizes by time treatment for both experimental replicates and mating orders. appropriate time. Eggs in the vial were counted when the female was dissected. No females laid eggs before 2 hours. Females with no prior progeny from the first mating were not remated. Copulations were excluded, if they completed with few to no sperm transferred. Time ASM Rep 1 Male 1=GFP Rep 1 Male 1=RFP Rep 2 Male 1=GFP Rep 2 Male 1=RFP 0 min min min min min min min min hr hr hr hr hr

28 MOVIE LEGENDS Movie S1. GFP-labeled sperm in the seminal receptacle in a female dissected 2 hr after copulation. Movie S2. RFP-labeled sperm in the seminal receptacle in a female dissected an unknown time after copulation. Movie S3. GFP-labeled sperm in a spermatheca in a female dissected an unknown time after copulation. Movie S4. GFP- and RFP-labeled sperm in a twice-mated female (first male was GFP) in a female dissected 60 min after copulation. Movie S5. GFP- and RFP-labeled sperm in a twice-mated female (first male was GFP) in a female dissected 60 min after copulation. Movie S6. GFP-labeled sperm in bursa in a female dissected immediately after copulation. 28

29 SUPPLEMENTAL REFERENCES 1. Pitnick, S., Markow, T. A. & Spicer, G. S. Delayed male maturity is a cost of producing large sperm in Drosophila. Proc. Natl. Acad. Sci. USA 92, (1995). 2. Santel, A., Winhauer, T., Blumer, N. & Renkawitz-Pohl, R. The Drosophila don juan (dj) gene encodes a novel sperm specific protein component characterized by an unusual domain of a repetitive amino acid motif. Mech. Dev. 64, (1997). 3. Price, C. S. C., Dyer, K. A. & Coyne, J. A. Sperm competition between Drosophila males involves both displacement and incapacitation. Nature 400, (1999). 4. Civetta, A. Direct visualization of sperm competition and sperm storage in Drosophila. Curr. Biol. 9, (1999). 5. Jayaramaiah Raja, S. J. & Renkawitz-Pohl, R Replacement by Drosophila melanogaster protamines and Mst77F of histones during chromatin condensation in late spermatids and role of sesame in the removal of these proteins from the male pronucleus. Mol. Cell Biol. 25, Hoskins, R. A. et al. A BAC-based physical map of the major autosomes of Drosophila melanogaster. Science 287, (2000). 7. Chippindale, A. K., Gibson, J. R. & Rice, W. R. Negative genetic correlation for adult fitness between sexes reveals ontogenetic conflict in Drosophila. Proc. Natl. Acad. Sci. USA 98, (2001). 8. Pattarini, J. M., Starmer, W. T., Bjork, A. C. & Pitnick, S. Mechanisms underlying the sperm quality advantage in Drosophila melanogaster. Evolution 60, (2006). 9. Clarke, M. R. B. The reduced major axis of a bivariate sample. Biometrika 67, (1980). 29