Investigating the Female's Role in Sperm Competition In

INVESTIGATING THE FEMALE’S ROLE IN SPERM COMPETITION

IN DROSOPHILA MELANOGASTER

A Thesis

Presented to the Faculty of the Graduate School

of Cornell University

In Partial Fulfillment of the Requirements for the Degree of

Master of Science

Simone L. White

May 2017

ABSTRACT

The process of fertilization involves many interactions between males, females, and their gametes. This is even more complex in cases of multiple mating, such as in Drosophila melanogaster, as the presence of ejaculates from multiple males and the female’s ability to store sperm presents the opportunity for sperm competition to occur. In D. melanogaster, male- derived seminal fluid proteins are known to influence various post-mating responses in the female including sperm competition outcomes. While studies have shown that female genotype is also important for sperm competition outcome, the mechanisms underlying the female’s contribution to the success of a particular male’s sperm are less understood.

To begin to examine the female’s role in sperm competition, we took two approaches in

D. melanogaster: First, we used RNAi knockdown of candidate genes to assess the impact of decreased expression of these genes on sperm competition outcomes. We found that of 29 candidate genes tested, 10 affect sperm competition outcomes when knocked down in females.

Second, we used fly lines mutant for neuromodulators to assess sperm storage outcomes from two different types of males mated sequentially. We found that the neuromodulators octopamine and tyramine may influence relative sperm storage of sperm from competing males. Collectively, results from these experiments provide a clearer picture of the genes and mechanisms involved in the female control of sperm competition outcomes and sperm dynamics.

BIOGRAPHICAL SKETCH

Simone White received a B.A. in Biological Sciences from Binghamton University,

Binghamton, NY in 2011.

iii

This thesis is dedicated to my parents and sister who always unconditionally support and love

me. I would not be where I am today without the three of you. You will always be the most

important people in my life, and for that I am extremely privileged and grateful.

iv ACKNOWLEDGEMENTS

First, I would like to thank my advisors, Mariana Wolfner and Andrew Clark, for their constant support, guidance, and kindness during my time in their labs. You both have provided a great working environment that I have truly felt welcome in and an important part of the group.

Secondly, I would like to thank my committee members, Brian Lazzaro and Paula Cohen, for being a part of my committee and offering helpful suggestions and insights on my research projects. In addition, I would like to thank the present and past members of the Wolfner and

Clark labs for their support and helpful discussions.

I would also like to thank the following sources of funding: Howard Hughes Medical

Institute Gilliam Fellowship for Advanced Study, an NIH grant awarded to Mariana Wolfner and

Andrew Clark (R01-HD059060), as well as a Diversity Supplement under this grant awarded to me for my support, and a SUNY Graduate Diversity Fellowship from the Graduate School at

Cornell University.

Lastly, I would like to thank my family for their love and encouragement, and for always being there when I need you. I cannot thank you enough for the values you have instilled within me, and the lengths you go to help and support me.

v TABLE OF CONTENTS

BIOGRAPHICAL SKETCH……………………………………………………………………..iii

DEDICATION……………………………………………………………………………………iv

ACKNOWLEDGEMENTS……………………………………………………………………….v

TABLE OF CONTENTS…………………………………………………………………………vi

LIST OF FIGURES……………………………………………………………………………...vii

LIST OF TABLES……………………………………………………………………………...viii

INTRODUCTION………………………………………………………………………………...1

MATERIALS AND METHODS………………………………………………………………….5

RESULTS………………………………………………………………………………………..12

DISCUSSION……………………………………………………………………………………24

REFERENCES…………………………………………………………………………………..26

vi LIST OF FIGURES

Figure 1. PCR showing knockdown efficiency of Zasp66 RNAi line…………………………….8

Figure 2. Control and tβh mutant females’ sperm storage from 2 different males………………22

Figure 3. Control and tdc2 mutant females’ sperm storage from 2 different males……………..23

vii LIST OF TABLES

Table 1. VDRC RNAi lines used for sperm competition assays………………………….7

Table 2. Measures of first male precedence (P1 score) in RNAi knockdowns of candidate genes………………………………………………………………………..……………13

Table 3. Control and tβh mutant females’ sperm storage from 2 different males……….22

Table 4. Control and tdc2 mutant females’ sperm storage from 2 different males……...23

viii INTRODUCTION

In Drosophila melanogaster, females often mate with and store sperm from multiple males. Female D. melanogaster can store sperm in two types of sperm storage organs, a single seminal receptacle and a pair of spermathecae. Storage of sperm from multiple males allows sperm competition to occur inside the female’s reproductive tract (Milkman & Zeitler 1974;

Prout & Bundgaard 1977; Imhof et al. 1998). The ability of Drosophila females to store sperm, along with the numerous tools available for study of Drosophila genes make them a particularly good model to study sperm competition.

The outcome of sperm competition depends on a network of complex interactions between male and female molecules. However, males and females often have different reproductive interests, leading to sexual conflict. For example, in systems where females mate multiply, males are often in direct competition with other males (Parker 1970). This may result in males having a damaging effect on females through seminal protein toxicity (Chapman et al.

1995; Wigby & Chapman 2004; Mueller et al. 2007) and through inhibiting a female’s likelihood of remating. However, the reproductive interests of females are to produce the most highly fit offspring. So, for example, it is often in the female’s interest to mate multiply, so as to increase the fitness of progeny, while at odds with the male’s strategy to inhibit remating by females. This conflict from different evolutionary goals of males and females suggests that the mechanisms behind sperm competition may be distinct in both sexes.

D. melanogaster males have been shown to influence a variety of post-mating characteristics through transfer of seminal fluid proteins. These characteristics include sperm competition (Clark et al. 1995; Fiumera et al. 2005, 2006, 2007; Chow et al. 2010; Greenspan &

Clark 2011). Moreover, male genotype and seminal fluid proteins have been shown to be

1 important for determining the outcome of sperm competition (Clark et al. 1995, 1999; Fiumera et al. 2005, 2007; Chow et al. 2010; Civetta et al. 2008).

The female’s role sperm competition is much less understood. Female genotype has been shown to affect sperm competition outcome (Clark & Begun 1998; Clark et al. 1995, 1999;

Chow et al. 2010), as do male X female genotype interactions (Clark & Begun 1998; Clark et al.

1999; Chow et al. 2010). However, there is limited knowledge about the individual genes and processes involved in the female contribution to sperm competition and dynamics. Females may influence sperm competition in several different ways. For example, females could modulate muscular contractions in the reproductive tract that move sperm into and out of storage through nervous system inputs, or spermathecal secretions promoting sperm viability while sperm are in storage could influence sperm competition outcomes. Genes expressed in the female reproductive tract might produce proteins that influence sperm competition outcomes (Swanson et al. 2004; Mack et al. 2006; Allen & Spradling 2008; Kapelnikov et al. 2008; Prokupek et al.

2008, 2009, 2010). Only two genes when altered in female D. melanogaster, Sex Peptide

Receptor (SPR; Chow et al. 2010) and Neprilysin 2 (Nep2; Sitnik et al. 2014) have been shown to influence sperm competition.

Although contributions from the female reproductive tract are likely important in sperm competition, there are probably other contributions from the female outside of her reproductive tract. To examine the genes that contribute to sperm competition, without bias as to site of expression, an unbiased GWAS for female genes that influence variation in sperm competition was conducted (Chow et al. 2013). Briefly, two males from standard laboratory lines (cn bw

[first male] and bwD [second male]) were mated in sequence to females from 39 lines from the

Drosophila Genetic Reference Panel (DGRP), which is a set of wild- derived, inbred lines whose

2 genome sequences are available (Mackay et al. 2012). The double mating experiments showed that sperm competition parameter P1 (proportion of first male progeny) was highly variable between DGRP lines. Statistical tests revealed single nucleotide polymorphisms (SNPs) that were associated with variation in the female’s effect on sperm competition across DGRP lines.

All SNPs are synonymous substitutions or are located in non-coding regions, suggesting that these SNPs somehow affect expression of candidate genes. Interestingly, 15 of the 33 top candidate genes identified by the GWAS are expressed in the nervous system or have specific neuronal functions. The functions of these neurological genes include encoding structural components of ion channels, and other aspects of nervous system function or development

(Chow et al. 2013). The potential importance of the nervous system suggests that the female may be playing a more active role than previously thought in sperm competition outcome.

The role of neural genes in female reproductive success has been suggested in previous studies as well. D. melanogaster females are able to actively regulate the release of stored sperm through the neuromodulators octopamine (OA) and tyramine (TA; Avila et al. 2012). Both of these neuromodulators are synthesized in neurons that innervate the female reproductive tract

(Middleton et al. 2006). While sperm entry and accumulation into storage were unaffected by the absence of OA and TA, sperm release from storage requires OA and TA (Avila et al. 2012).

Furthermore, the uptake and release of several neuromodulators changes throughout the female reproductive tract after mating (Heifetz et al. 2014). This creates unique combinations of neuromodulator levels in different regions of the reproductive tract at different times, which may coordinate female post-mating responses such as sperm and egg movement and release, potentially influencing sperm competition outcomes.

3 The nervous system input has been shown to be important for sperm storage in other organisms as well. Neural input and hormones are important for muscular contraction in the spermatheca in locusts (for review, see Lange & da Silva 2007). More specifically, OA has been shown to modulate muscle contractions in the spermatheca in locusts (Clark & Lange 2003).

Taken together, it appears that input from the female nervous system is important for proper sperm storage and may play a role in sperm competition. However how the nervous system works to control sperm use is unclear.

To examine the female’s role in sperm competition, we first performed ubiquitous RNAi knockdowns of candidate genes from Chow et al. 2013 to assess the impact of decreased expression of these female-expressed candidate genes on sperm competition. Knockdown and control females were scored for sperm competition effects using progeny-phenotype assays. Of

29 genes tested, knockdown of 10 affected sperm competition outcomes. Effects on sperm competition outcome of knockdown of these genes in specific neurons were tested to identify those neurons that were essential for sperm competition. Separately, we examined the storage of sperm from multiple males after mating to female flies lacking octopamine or tyramine.

Preliminary results suggest that octopamine and tyramine also play a role in how sperm is stored from multiple males over time. Overall, our results provide clues as to how females may actively influence sperm competition.

4 MATERIALS AND METHODS

Generating control and experimental females

To knock down female-expressed candidate genes, we generated RNAi females by crossing the ubiquitous driver Tubulin-GAL4/TM3, Sb virgin females to UAS-RNAi generating males (Vienna Drosophila RNAi Center [VDRC]). VDRC lines used are included in Table 1.

Control females were generated by crossing virgin females from the same Tubulin-GAL4/TM3,

Sb driver line to males from the AttP or w1118 RNAi background line (VDRC). Control and knockdown females were therefore identical except for the insertion of the RNAi transgene in experimental females. Only Sb+ progeny were used from each cross. Eighty to 100 virgin control and knockdown females were collected for each treatment. Typical sample size at the end was

30-60 doubly mated females, as females that did not mate the first or second time were not included in the final analysis. Knockdown efficiency (Table 1) was examined by PCR of cDNA synthesized from RNA isolated from whole control and knockdown female flies. A representative figure showing efficient knockdown is shown in Figure 1. In cases where knockdown was not determined, PCR was unsuccessful or knockdown has not yet been assessed.

In cases where ubiquitous knockdown of a candidate gene was lethal, a Tubulin-

GAL80ts/Tubulin-GAL80ts; Tubulin-GAL4/TM3, Sb was used to allow for temporal control of

GAL4 expression during adulthood only, by shifting crosses from room temperature to 29°C after flies had eclosed. Females were aged and maintained at 29°C throughout the assay.

We also examined candidates for a neural effect on sperm competition by driving knockdown using a pan-neuronal driver, nSyb-GAL4 (Hindle et al. 2013). If an effect was observed from pan-neuronal knockdown, we then used ppk-GAL4 (Matthews et al. 2007) to drive knockdown in the ppk neurons that innervate the female reproductive tract, and in one

5 case, tdc2-GAL4 (Cole et al. 2005), which targets octopaminergic neurons, which are important for sperm release from storage (Avila et al. 2012). All flies were collected as virgins under CO2 anesthesia and aged 3-7 days in single-sex vials of 30-40 flies, and maintained on a yeast-glucose agar medium at room temperature on a 12-hour light/dark cycle.

Table 1. VDRC RNAi lines used for sperm competition assays. N.D., Not Determined % Knockdown, approximate percentage of gene knocked down Gene VDRC ID % Knockdown Cyp313a2 110504 N.D. Shab 102218 90% sima 106504 99% spz5 102389 N.D. CG42796 102356 N.D. btsz 102608 90% Msp300 107183 N.D. CG10962 106396 90% CG32532 106534 90% para 104775 N.D. SK 103985 N.D. Rab2 105358 N.D. Rim 39384 N.D. hid 8269 N.D. uif 101153 99% CG9850 107199 99% CG15800 110049 50% CG32834 100964 80% Rbp6 29799 99% caup 105705 N.D. CG32264 101503 80% CG15765 101194 40% CG33095 38604 90% CG33298 42776 0% CG31872 102669 99% CG6163 22267 95% Ddr 101831 99% Zasp66 102980 97% CG13594 36590 N.D. Dh44 108473 N.D.

1 1:2 1:4 1:10 1:40 1:100 1 1:2 Control Knockdown

Figure 1. PCR showing knockdown efficiency of Zasp66 RNAi. The band is a 180 bp fragment of the cDNA of Zasp66. Control and knockdown samples were diluted as shown to gauge knockdown efficiency. Figure shows Zasp66 is knocked down about 97% (undiluted knockdown is comparable to control 1:40 dilution).

Sperm competition assays

Control and experimental females were given the opportunity to mate in sequence to males from the standard laboratory lines cn bw and bwD. Sperm competition assays were performed by methods similar to those in Chow et al. (2013). Control and knockdown females have wild-type red eyes, while cn bw males have recessive mutations resulting in white eyes. The F1 progeny sired by cn bw males have red eyes. The bwD males are homozygous for a dominant mutation

D D that causes brown eyes. F1 progeny sired by bw males inherit one copy of the dominant bw allele, resulting in brown eyes. Young males were also collected at the same time as control and experimental females, and were handled and housed in a similar manner to virgin females.

For the first mating, virgin control and knockdown females were singly mated to cn bw males in glass vials (day 0; vial 1). Vials were observed for copulation every 15 minutes, and males were removed from vials after mating concluded. The next night (day 1), two bwD males were placed with each female for the second mating for 12 hours overnight. In the morning on day 2, bwD males were discarded and females were placed into a new vial (vial 2) and allowed to lay eggs for 48 hours. Again on days 4, 6, and 8, individual females were placed into a new vial

(vials 3, 4, and 5) and allowed to lay eggs for 48 hours in each vial. On day 10, females were discarded. Adult female progeny were then scored for eye color (red vs. brown) to determine paternity. To be scored accurately, the bwD eye color phenotype requires a w+ background. Since

VDRC lines as well as driver lines are in a w- background, only female progeny are scored as they receive a w+ chromosome from their bwD fathers.

Statistical analysis

9 A P1 score was calculated for control and experimental females. P1 is the proportion of progeny sired by the first male after the second mating. In this case, this is represented by the number of red-eyed progeny divided by the total number of progeny (red/[red+brown]) from vials 2-5 of the individual females that remated. Since the second mating occurs in vial 1, vial 1 is not included in calculating the P1 score because it is not possible to determine which red-eyed progeny were deposited as embryos before the second mating occurred. If P1=0 or 1, we confirmed that there were both types of progeny in vial 1 (verifying that both matings occurred). P1 was then compared between control and experimental females using a one-way ANOVA. All statistical analyses were performed in R.

Examining sperm storage and dynamics from two different males in OA and TA mutant females

Experiments were carried out by doubly mating 3-5 day old virgin control or mutant females to

3-5 day old males that express a protamine labeled with either GFP (first male) or RFP (second male), which show green or red fluorescence, respectively, in sperm heads (Manier et al. 2010).

Fly lines from which virgin control and mutant females were derived from were available in our lab (tβhM18/FM7 and tdc2RO54/Gla, Rubinstein and Wolfner 2013) and from the Bloomington

Stock Center (Df(2R)42/SM5). All flies were maintained on a standard yeast-glucose medium at room temperature on a 12-hour light/dark cycle.

Double-mating experiments were done in a similar manner to the sperm competition assays described above. Virgin control (tβhM18/FM7, tdc2RO54/SM5, Df(2R)42/Gla) and virgin null-mutant (tβhM18/tβhM18, tdc2RO54/Df(2R)42) females were first singly mated to GFP-sperm labeled males. GFP-sperm males were removed after mating was complete. The next day, all females who successfully mated with GFP-sperm labeled males were given the opportunity to

10 mate with two RFP-sperm labeled males overnight. The next morning the RFP-sperm male flies were removed, and females were maintained in vials for 1, 4, or 10 days, then flash frozen, and stored at -80°F to later be dissected. To visualize sperm, female reproductive tracts were dissected in PBS on a microscopic slide, covered with a class coverslip sealed with rubber cement, and examined at 400x under a florescence microscope.

The number of sperm stored within the individual sperm storage organs was counted for each time point at 400X on a fluorescence microscope. T-tests were used to determine if there was a difference in sperm stored from each type of male in each sperm storage organ over time.

11 RESULTS

Functional testing of candidate genes

To examine the possible role of candidate genes in sperm competition in females, we used RNAi to ask if reduced expression changes sperm competition outcomes. Of 29 genes tested, knockdown of 10 significantly affected P1 score (Table 2). Four candidates from Chow et al. (CG34027, CG10858, RFeSP and sti) do not have RNAi lines available, and one (CG13594) was not tested due to its 94 predicted off-targets. One additional candidate (Dh44) was tested due to its previously known involvement in sperm ejection (Kim et al. 2015). In cases where there is no ubiquitous knockdown shown, ubiquitous knockdown was too detrimental to females for a sperm competition assay to be performed (females did not survive through the assay). Additional tests were performed on some candidates to determine if using different types of males

(Protamine-GFP and Protamine-RFP) and/or reversing the mating order of males would still show an effect on sperm competition outcomes.

Many candidates remained significant in their effect on P1 score when knocked down specifically in the nervous system as well. The neuronal candidates may be involved in the development of neural networks that are important for sperm competition, and/or could function at the time sperm competition occurs. Additional tests will need to be done to determine how candidates are acting to affect sperm competition outcomes.

Table 2. Measures of first male precedence (P1 score) in RNAi knockdowns of candidate genes. Ten of 29 genes tested affect P1 score when knocked down in females. P-values reflect the result of an ANOVA statistical test. Note: Knockdown of CG10962 results in a close-to-significant p- value. However, this RNAi line has 2 off targets and knockdown flies were not healthy. KD, knockdown N, number of doubly mated females Median, median P1 score Progeny, total progeny counted (missing in cases where experiment is from Chow et al. 2013) *Significant result † Data in row from Chow et al. 2013 ‡ GFP- (first male) and RFP- (second male) sperm labeled males used for assay § Reversed mating order: bwD first male, cn bw second male ǁ Candidate gene not from Chow et al. 2013 DGRP screen Δ Data collected with Martik Chatterjee

13 Gene Driver Control N, KD N, p-value Box plot knocked Median Median down Progeny Progeny counted counted Rim † ppk-GAL4 44 41 0.017* 0.560 0.519

Figure 3 Level of ﬁrst-male sperm precedence (P1 score) in RNAi knock- ‡ ppk-GAL4 64 51 0.2285 Figure 3 Level of ﬁrst-male sperm precedence (P1 score) in RNAi knockdown of select neurological candidates. Three of the0.8 four candidates 0.597 0.671 down of select neurological candidates. Three of the four candidates chosen for functional analysis affect P1 score when knocked down in Figure 2 Average expression of candidate genes associated with sperm chosenP1 for functional analysis affect P1 score when knocked down in

Figure 2 Average expression of candidate genes associated with sperm 0.4 competition in different tissues. The mean expression levels of the 33 16,304sensory neurons innervating13,762 the female reproductive tract. KD, knock- competition in different tissues. The mean expression levels of the 33 sensory neurons innervating the female reproductive tract. KD, knock- candidate genes are higher in neuronal tissues (dark shading) than in down; *P , 0.05. candidate genes are higher in neuronal tissues (dark shading) than in down; *P , 0.05. nonneuronal tissues (light shading) (P = 0.0019). Tissue-specific expres- 0.0 nonneuronal tissues (light shading) (P = 0.0019). Tissue-specific expression data were taken from FlyAtlas. TaG, thoracicoabdominal ganglion; fi Control Rim sion data were taken from FlyAtlas.Knockdown TaG, thoracicoabdominal of three of these ganglion; genes signi cantly affected SG, salivary gland; Spt, spermatheca; L_CNS, larval central nervous sys- Knockdown+ of three of these genes significantly affected SG,ppk salivary-GAL4 gland; Spt, spermatheca;32P1 score. When L_CNS,34Rab2 larval centralwas knocked nervous0.0084* sys- down in ppk neurons, tem. Mean 6 SD is shown. P1 score.0.8 When Rab2 was knocked down in ppk+ neurons, tem. Mean 6 SD is shown.0.41knockdown females0.263 had a significant increase in P1 score knockdown females had a significant increase in P1 score compared to control females [control medianP1 P1 = 0.685, 3,881 4,105 0.4 Future experimentation is needed to establish a functional N = 38; knockdown (KD) median P1 = 0.836,comparedN = 40;to controlP = females [control median P1 = 0.685, Future experimentation is needed to establish a functional N = 38;+ knockdown (KD) median P1 = 0.836, N = 40; P = role of these genetic variants in sperm competition. 0.04] (Figure 3). Thus loss of Rab2 function in 0.0 ppk neurons + role of these genetic variants in sperm competition. 0.04] (FigureControl 3). ThusRim loss of Rab2 function in ppk neurons Functional testing results in an increase in first-male progeny compared to § Functionalppk-GAL4 testing 42controls. Knockdown57 of para0.0321*in sensory results neurons in resulted an increase in first-male progeny compared to To validate potential roles of neurological genes in sperm in reduction in P1 score compared to thatcontrols. in control0.8 Knockdown females of para in sensory neurons resulted To validate potential0 roles of neurological0.015 genes in sperm competition in females, we used RNAi to ask whether re- (control median P1 = 0.580, N = 40;in KD reduction median inP1 P1 = score compared to that in control females P1

competition in females,5,053 we used RNAi8,060 to ask whether re- 0.4 duction in the expression of these genes changes sperm 0.506, N = 42; P = 0.008) (Figure 3). Similarly,(control knockdown median P1 = 0.580, N = 40; KD median P1 = duction in the expression of these genes changes sperm competition outcomes. These candidates could be poten- of Rim in sensory neurons also resulted in0.506, a signiN =ficant 42; P re-= 0.008) (Figure 3). Similarly, knockdown tially involved in the development of neuralcompetition networks outcomes.duction These in candidates P1 score (control could median be poten- P1 = 0.560,of Rim0.0 Nin= sensory 44; KD neurons also resulted in a significant re- specificforspermcompetitionand/ortheycouldfunctiontially involved in themedian development P1 = 0.519, of neuralN = 41; networksP = 0.017) (Figureduction 3).in P1Control Knock- score (controlRim median P1 = 0.560, N = 44; KD at the time of sperm competition.Rab2 Given † specippk that-fiGAL4 knockdowncforspermcompetitionand/ortheycouldfunction 38down of para or40Rim in ppk+0.04*neurons resultsmedian in production P1 = 0.519, N = 41; P = 0.017) (Figure 3). Knock- of at least some of these neural genes isat lethal the time or affects of sperm0.685of competition. fewer first-male Given0.836 progeny that knockdown after a seconddown mating. of para or Rim in ppk+ neurons results in production behaviors unrelated to sperm competitionof (e.g. at, least locomotion some of theseWe neural observed genes no is difference lethal or affects in P1 scoreof fewer whenfirst-maleSK was progeny after a second mating. behaviors that could affect courtship or mating)behaviors (Loughney unrelated toknocked sperm competition down in ppk (e.g.+ ,neurons locomotion (controlWe median observed P1 = no difference in P1 score when SK was et al. 1989; Lloyd et al. 2000; Schulte et al.behaviors2010), we that chose could0.578, affect courtshipN = 40; or KD mating) median (Loughney P1 = 0.578,knockedN = 40;downP = in ppk+ neurons (control median P1 = fi to test a speci chypothesis:thatneuralgenesmightin-et al. 1989; Lloyd et al.0.344)2000; (Figure Schulte 3).et al. The2010), previous we chose success0.578, of thisNppk-GAL4= 40; KD median P1 = 0.578, N = 40; P = fluence sperm competition through neurons that inner- to test a specifichypothesis:thatneuralgenesmightin-driver and the lethality of a ubiquitous0.344) knockdown (Figure of SK 3). The previous success of this ppk-GAL4 vate the female reproductive tract. This approach likely fluence sperm competition(data not through shown) neurons suggest that knockdowninner- driver is occurring, and the lethalityal- of a ubiquitous knockdown of SK underestimates the extent of contribution of these genes vate the female reproductivethough we tract. cannot This gauge approach its extent. likely It is unclear(data not whether shown)SK suggest that knockdown is occurring, al- that we found, but is a simple, initial, andunderestimates direct test of the extent of contribution of thesefi genes ‡ ppk-GAL4 57knockdown level51 was insuf 0.9605cient to givethough a sperm we competi- cannot gauge its extent. It is unclear whether SK

function. 0.8 that we found, but istion a simple, phenotype, initial, whether and direct SK’s role test in of spermFigureknockdown competition 3 Level level of first-male is was insuf spermficient precedence to give (P1 score)a sperm in RNAi competi- knock- Candidate gene function was tested by knocking down 0.467 0.482+ function. not through ppk neurons, whether compensatorydowntion phenotype, of select mecha- neurological whether candidates. SK’s role Three in sperm of the competition four candidates is + P1 expression in the ppk sensory neurons that innervate the chosen0.4 for functional analysis affect P1 score when knocked down in Figure Candidate 2 Average gene expression function13,434nisms of candidate arewas at tested play, genes11,797 or associatedby whether knocking with the downsperm SK association was a false+ sensorynot through neuronsppk innervatingneurons, the female whether reproductive compensatory tract. KD, mecha- knock- female reproductive tract. We tested candidatescompetitionexpression in in in differentthese the ppk tissues.positive.+ sensory The mean neurons expression that levels innervate of the the 33 down;nisms* areP , at0.05. play, or whether the SK association was a false neurons because SPR,theonlyfemalegeneknowntobecandidate genes are higher in neuronal tissues (dark shading) than in 0.0 female reproductive tract. We tested candidatesfi in these positive. important for sperm competition, signalsnonneuronal through tissues them (light shading) (P = 0.0019). Tissue-speci c expres- Control Rab2 neurons because SPRDiscussion,theonlyfemalegeneknowntobe (Hasemeyer et al. 2009; Yang et al. 2009).sion data Knockdown were taken from FlyAtlas. TaG, thoracicoabdominal ganglion; Knockdown of three of these genes significantly affected SG,important salivary gland; forSpt, sperm spermatheca; competition, L_CNS, signals larval central through nervous them sys- was achieved with the ppk-GAL4 driver that is specifically Despite overwhelming evidence that theP1Discussion female score. genotype When Rab2 was knocked down in ppk+ neurons, tem.(Hasemeyer Mean 6 SDet is shown. al. 2009; Yang et al. 2009). Knockdown expressed in this subset of neurons (Grueber et al. 2007). affects sperm competition (Clark and Begunknockdown 1998; females Clark had a significant increase in P1 score was achieved with the ppk-GAL4 driver that is specifically Despite overwhelming evidence that the female genotype Knockdown females were compared to controls in a sperm et al. 1999; Fiumera et al. 2005, 2007; Civettacomparedet al. to2008; control females [control median P1 = 0.685, expressed in this subset of neurons (Grueber et al. 2007). affects sperm competition (Clark and Begun 1998; Clark competition experiment identical to theFuture DGRP experiment experimentationChow is neededet al. 2010), to establish little is a known functional about theN molecular= 38; knockdown mech- (KD) median P1 = 0.836, N = 40; P = Knockdown females were compared to controls in a sperm et al. 1999; Fiumera et al. 2005, 2007; Civetta et al. 2008; described above. Because we sought torole test ofcandidates these genetic in variantsanisms that in sperm underlie competition. these female contributions.0.04] (Figure We sought 3). Thus loss of Rab2 function in ppk+ neurons competition experiment identical to the DGRP experiment Chow et al. 2010), little is known about the molecular mech- neurons (rather than muscle), we chose candidate genes to identify potential candidate pathwaysresults and molecules in an increase in in first-male progeny compared to Functionaldescribed testingabove. Because we sought to test candidates in anisms that underlie these female contributions. We sought with clear functions in neurons. Four of the neurological the female that play a role in sperm competition.controls. To Knockdown do this, of para in sensory neurons resulted neurons (rather than muscle), we chose candidate genes to identify potential candidate pathways and molecules in candidates, with various levels of significanceTo validate in the associ- potentialwe roles took of neurological advantage14 of genes the variation in sperm capturedin reduction in the inDGRP P1 score compared to that in control females with clear functions in neurons. Four of the neurological the female that play a role in sperm competition. To do this, ation study (Figure S4), were chosen forcompetition functional testing: in females,resource we used (Mackay RNAi toet al. ask2012). whether The re- DGRP(control is a collection median of P1 = 0.580, N = 40; KD median P1 = candidates, with various levels of significance in the associ- we took advantage of the variation captured in the DGRP Rab2, SK, para, and Rim. duction in the expression192 wild-derived of these genes inbred changes lines sperm from a0.506, singleN population.= 42; P = 0.008) (Figure 3). Similarly, knockdown ation study (Figure S4), were chosen for functional testing: resource (Mackay et al. 2012). The DGRP is a collection of competition outcomes. These candidates could be poten- of Rim in sensory neurons also resulted in a significant re- Rab2, SK, para, and Rim. 192 wild-derived inbred lines from a single population. tially involved in the development of neural networks duction in P1 score (control median P1 = 0.560, N = 44; KD 182 C. Y. Chow, M. F. Wolfner, and A. G. Clarkspecificforspermcompetitionand/ortheycouldfunction median P1 = 0.519, N = 41; P = 0.017) (Figure 3). Knock- at the time of sperm competition. Given that knockdown down of para or Rim in ppk+ neurons results in production of182 at leastC. Y. some Chow, of M. these F. Wolfner, neural and genes A. G. Clark is lethal or affects of fewer first-male progeny after a second mating. behaviors unrelated to sperm competition (e.g., locomotion We observed no difference in P1 score when SK was behaviors that could affect courtship or mating) (Loughney knocked down in ppk+ neurons (control median P1 = et al. 1989; Lloyd et al. 2000; Schulte et al. 2010), we chose 0.578, N = 40; KD median P1 = 0.578, N = 40; P = to test a specifichypothesis:thatneuralgenesmightin- 0.344) (Figure 3). The previous success of this ppk-GAL4 fluence sperm competition through neurons that inner- driver and the lethality of a ubiquitous knockdown of SK vate the female reproductive tract. This approach likely (data not shown) suggest that knockdown is occurring, al- underestimates the extent of contribution of these genes though we cannot gauge its extent. It is unclear whether SK that we found, but is a simple, initial, and direct test of knockdown level was insufficient to give a sperm competi- function. tion phenotype, whether SK’s role in sperm competition is Candidate gene function was tested by knocking down not through ppk+ neurons, whether compensatory mecha- + expression in the ppk sensory neurons that innervate the nisms are at play, or whether the SK association was a false female reproductive tract. We tested candidates in these positive. neurons because SPR,theonlyfemalegeneknowntobe important for sperm competition, signals through them Discussion (Hasemeyer et al. 2009; Yang et al. 2009). Knockdown was achieved with the ppk-GAL4 driver that is specifically Despite overwhelming evidence that the female genotype expressed in this subset of neurons (Grueber et al. 2007). affects sperm competition (Clark and Begun 1998; Clark Knockdown females were compared to controls in a sperm et al. 1999; Fiumera et al. 2005, 2007; Civetta et al. 2008; competition experiment identical to the DGRP experiment Chow et al. 2010), little is known about the molecular mech- described above. Because we sought to test candidates in anisms that underlie these female contributions. We sought neurons (rather than muscle), we chose candidate genes to identify potential candidate pathways and molecules in with clear functions in neurons. Four of the neurological the female that play a role in sperm competition. To do this, candidates, with various levels of significance in the associ- we took advantage of the variation captured in the DGRP ation study (Figure S4), were chosen for functional testing: resource (Mackay et al. 2012). The DGRP is a collection of Rab2, SK, para, and Rim. 192 wild-derived inbred lines from a single population.

182 C. Y. Chow, M. F. Wolfner, and A. G. Clark Table 2 (continued) para † ppk-GAL4 40 42 0.008* 0.58 0.506

‡ ppk-GAL4 57 Figure35 3 Level of0.5885ﬁrst-male spermFigure precedence 3 Level (P1 of score)ﬁrst-male in RNAi sperm knock- precedence (P1 score) in RNAi knock- 0.8 0.467 down0.427 of select neurological candidates.down of Three select of neurological the four candidates candidates. Three of the four candidates chosen for functional analysis affect P1 score when knocked down in Figure 2 Average expression of candidateFigure 2 genesAverage associated expression with of sperm candidate genes associated with sperm chosenP1 for functional analysis affect P1 score when knocked down in 13,434 sensory8,335 neurons innervating the female0.4 reproductive tract. KD, knock- competition in different tissues. Thecompetition mean expression in different levels tissues. of the The 33 mean expression levels of the 33 sensory neurons innervating the female reproductive tract. KD, knockdown; *P , 0.05. candidate genes are higher in neuronalcandidate tissues genes (dark are highershading) in than neuronal in tissues (dark shading) than in down; *P , 0.05.

fi 0.0 nonneuronal tissues (light shading)nonneuronal (P = 0.0019). tissues Tissue-speci (light shading)c expres- (P = 0.0019). Tissue-specific expression data were taken from FlyAtlas. TaG, thoracicoabdominal ganglion; Control para sion data were taken from FlyAtlas. TaG, thoracicoabdominalKnockdown of ganglion; three of theseKnockdown genes signi officantly three of affected these genes significantly affected SG, salivary gland; Spt, spermatheca; L_CNS, larval central nervous sys- + btsz SG,Tubulin salivary- gland; Spt, spermatheca;37 L_CNS,P166 larval score. central When nervous0.0281*Rab2 sys-was knockedP1 score. down When inRab2ppk wasneurons, knocked down in ppk+ neurons, tem. Mean 6 SD is shown. tem. Meants6 SD is shown. 0.8 GAL80 ; 0.264 knockdown0.16 females had aknockdown significant increase females in had P1 a score significant increase in P1 score

compared to control femalesP1 [control median P1 = 0.685, Tubulin- 4,702 6,084 compared0.4 to control females [control median P1 = 0.685, Future experimentation is needed to establish a functional FutureGAL4/TM3, experimentationSb is needed toN = establish 38; knockdown a functional (KD) medianN = 38; P1 knockdown = 0.836, N (KD)= 40; medianP = P1 = 0.836, N = 40; P = role of these genetic variants in sperm competition. + role of these genetic variants in sperm0.04] competition. (Figure 3). Thus loss of0.04]Rab20.0 (Figurefunction 3). in Thusppk lossneurons of Rab2 function in ppk+ neurons results in an increase in first-male progenybtsz comparedControl to Functional testing Functional testing results in an increase in first-male progeny compared to Δ nSyb-GAL4 46 controls.52 Knockdown0.003* of paracontrols.in sensory Knockdown neurons of resultedpara in sensory neurons resulted To validate potential rolesTo of neurological validate potential genes0.402 roles in sperm of neurologicalin0.123 reduction genes inin P1 sperm score compared0.8 to that in control females competition in females, we used RNAi to ask whether re- in reduction in P1 score compared to that in control females competition in females, we used RNAi(control to ask median whether P1 re-= 0.580,P1 N = 40; KD median P1 = 6,538 6,284 (control0.4 median P1 = 0.580, N = 40; KD median P1 = duction in the expression of these genes changes sperm 0.506, N = 42; P = 0.008) (Figure 3). Similarly, knockdown duction in the expression of these genes changes sperm 0.506, N = 42; P = 0.008) (Figure 3). Similarly, knockdown competition outcomes. These candidates could be poten- of Rim in sensory neurons also resulted in a significant re- competition outcomes. These candidates could be poten- of Rim0.0 in sensory neurons also resulted in a significant re- tially involved in the development of neural networks duction in P1 score (control median P1btsz = 0.560, ControlN = 44; KD tially involved in the development of neural networks duction in P1 score (control median P1 = 0.560, N = 44; KD specificforspermcompetitionand/ortheycouldfunction median P1 = 0.519, N = 41; P = 0.017) (Figure 3). Knock- specitdc2-fiGAL4cforspermcompetitionand/ortheycouldfunction 25 14 0.1386 median P1 = 0.519, N = 41; P = 0.017) (Figure 3). Knock- at the time of sperm competition. Given that knockdown down of para or Rim in ppk+ neurons0.8 results in production at the time of sperm0.381 competition. Given0.292 that knockdown down of para or Rim in ppk+ neurons results in production of at least some of these neural genes is lethal or affects fi of at least some of these neural genesof fewer is lethalrst-male or affects progeny afterP1 a second mating. of fewer0.4 first-male progeny after a second mating. behaviors unrelated to sperm competition (e.g., locomotion3,404 1,738We observed no difference in P1 score when SK was behaviors unrelated to sperm competition (e.g., locomotion We observed no difference in P1 score when SK was behaviors that could affect courtship or mating) (Loughney knocked down in ppk+ neurons (control median P1 = behaviors that could affect courtship or mating) (Loughney knocked0.0 down in ppk+ neurons (control median P1 = et al. 1989; Lloyd et al. 2000; Schulte et al. 2010), we chose 0.578, N = 40; KD median P1 = 0.578, N = 40; P = et al. 1989; Lloyd et al. 2000; Schulte et al. 2010), we chose 0.578, N =btsz 40; KDControl median P1 = 0.578, N = 40; P = to test a specifichypothesis:thatneuralgenesmightin- 0.344) (Figure 3). The previous success of this ppk-GAL4 Δ toppk test-GAL4 a speci fichypothesis:thatneuralgenesmightin-34 40 0.2637 0.344) (Figure 3). The previous success of this ppk-GAL4 fluence sperm competition through neurons that inner- 0.8 fluence sperm competition throughdriver neurons and the that lethality inner- of a ubiquitous knockdown of SK vate the female reproductive tract. This approach0.185 likely 0.182 driver and the lethality of a ubiquitous knockdown of SK vate the female reproductive tract.(data This not approach shown) likely suggest that knockdown is occurring, al- underestimates the extent of contribution of these4,927 genes 5,603 (dataP1 not shown) suggest that knockdown is occurring, al- underestimates the extent of contributionthough we of cannot these genes gauge its extent.0.4 It is unclear whether SK that we found, but is a simple, initial, and direct test of knockdown level was insuffithoughcient to we give cannot a sperm gauge competi- its extent. It is unclear whether SK that we found, but is a simple, initial, and direct test of fi function. tion phenotype, whether SKknockdown’s role in sperm levelwas competition insuf cient is to give a sperm competi- function. 0.0 Candidate gene function was tested by knocking down + tion phenotype, whether SK’s role in sperm competition is not through ppk neurons, whether compensatorybtsz Control mecha- expression in the ppk+ sensoryCandidate neurons that gene innervate function thewas tested by knocking down not through ppk+ neurons, whether compensatory mecha- + nisms are at play, or whether the SK association was a false female reproductiveMsp300 tract. expression WeTubulin tested- in candidates the ppk85 in sensory these neurons48 that innervate0.0002* the nisms are at play, or whether the SK association was a false ts positive. 0.8 neurons because SPR,theonlyfemalegeneknowntobefemaleGAL80 reproductive; 0.3 tract. We tested0.067 candidates in these positive. P1 important for sperm competition,neuronsTubulin signalsbecause- throughSPR7,812,theonlyfemalegeneknowntobe them 2,735 0.4 Discussion (Hasemeyer et al. 2009; Yangimportantet al. for2009). sperm Knockdown competition, signals through them

GAL4/TM3,Sb Discussion0.0 (Hasemeyer et al. 2009; Yang et al. 2009). Knockdown was achieved with the ppk-GAL4 driver that is specifically Despite overwhelming evidence thatControl the femaleMsp300 genotype was achieved with the ppk-GAL4 driver that is specifically Despite overwhelming evidence that the female genotype expressed in this subset of neurons (Grueber et al. 2007). affects sperm competition (Clark and Begun 1998; Clark Knockdown females were comparedexpressed to in controls this subset in a of sperm neuronset (Grueber al. 1999;et Fiumera al. 2007).et al. 2005,affects 2007; sperm Civetta competitionet al. 2008; (Clark and Begun 1998; Clark competition experiment identicalKnockdown to the females DGRP experiment were comparedChow to controlset al. 2010), in a sperm little is knownet al. 1999; about the Fiumera molecularet al. mech-2005, 2007; Civetta et al. 2008; described above. Because wecompetition sought to experiment test candidates identical in toanisms the DGRP that underlie experiment these femaleChow et contributions. al. 2010), little We is sought known about the molecular mech- neurons (rather than muscle),described we chose above. candidate Because genes we soughtto toidentify test candidates potential candidate in anisms pathways that underlie and molecules these female in contributions. We sought with clear functions in neurons.neurons Four (rather of the than neurological muscle), wethe chose female candidate that play genes a role into sperm identify competition. potential candidate To do this, pathways and molecules in candidates, with various levelswith of clear signifi functionscance in the in neurons. associ- Fourwe took of the advantage neurological of the variationthe female captured that play in a therole DGRP in sperm competition. To do this, ation study (Figure S4), werecandidates, chosen for with functional various testing: levels of signiresourceficance (Mackay in the associ-et al. 2012).we took The DGRPadvantage is a collection of the variation of captured in the DGRP Rab2, SK, para, and Rim. ation study (Figure S4), were chosen192 for15 wild-derived functional testing: inbred linesresource from (Mackay a singleet population.al. 2012). The DGRP is a collection of Rab2, SK, para, and Rim. 192 wild-derived inbred lines from a single population.

182 C. Y. Chow, M. F. Wolfner, and A. G. Clark 182 C. Y. Chow, M. F. Wolfner, and A. G. Clark Table 2 (continued)

nSyb-GAL4 66 58 0.04* 0.8 0.095 0.173 P1 8,010 7,538 0.4 0.0 Control Msp300

Ddr Tubulin- 38 25 0.015* 0.8 GAL4/TM3,Sb 0.462 0.268 P1

5,404 2,479 0.4 0.0 Control Ddr nSyb-GAL4 34 36 0.0311*

0.257 0.127 0.8

4,914 3,969 P1 0.4 0.0 Control Ddr CG31872 Tubulin- 38 33 0.0001* GAL4/TM3,Sb 0.462 0.154 0.8

5,404 3,872 P1 0.4 0.0 CG31872 Control Tubulin- 61 22 0.0073* GAL4/TM3,Sb 0.384 0.171 0.8 P1

7,828 1,777 0.4 0.0 CG31872 Control CG32834 Tubulin- 36 43 0.0038* GAL4/TM3,Sb 0.392 0.226 0.8

4,101 4,316 P1 0.4 0.0 CG32834 Control caup Δ nSyb-GAL4 24 33 0.0015* 0.652 0.204 0.8

2,447 3,726 P1 0.4

0.0 caup Control

16 Table 2 (continued) ppk-GAL4 55 55 0.5753 0.272 0.167 0.8

5,748 5,722 P1 0.4 0.0 caup Control hid nSyb-GAL4 44 57 0.0170* 0.273 0.474 0.8 P1

10,096 12,257 0.4 0.0 Control hid SK † ppk-GAL4 40 40 0.344 0.578 0.578

Δ nSyb-GAL4 49 52 Figure0.9797 3 Level ofFigureﬁrst-male 3 Level sperm of ﬁ precedencerst-male sperm (P1 score) precedence in RNAi (P1 knock- score) in RNAi knock- 0.8 0.47 0.5 down of select neurologicaldown of select candidates. neurological Three candidates. of the four Three candidates of the four candidates

Figure 2 AverageFigure expression 2 Average of candidate expression genes of associatedcandidate genes with sperm associatedchosen with sperm for functionalchosenP1 analysis for functional affect P1 analysis score whenaffect P1knocked score downwhen knockedin down in 6,737 7,216 0.4 competition in differentcompetition tissues. in different The mean tissues. expression The mean levels expression of the 33 levelssensory of the neurons 33 sensory innervating neurons the femaleinnervating reproductive the female tract. reproductive KD, knock- tract. KD, knock- candidate genescandidate are higher genes in neuronal are higher tissues in neuronal(dark shading) tissues than (dark in shading)down; than*P , in 0.05.down; *P , 0.05. nonneuronal tissuesnonneuronal (light shading) tissues (P (light= 0.0019). shading) Tissue-speci (P = 0.0019).fic expres- Tissue-specific expres- 0.0 Control SK sion data were takension datafrom were FlyAtlas. taken TaG, from thoracicoabdominal FlyAtlas. TaG, thoracicoabdominal ganglion; ganglion; Cyp313a2 Tubulin- 62 58 0.1916Knockdown ofKnockdown three of these of three genes of signi theseficantly genes affected significantly affected SG, salivary gland;SG, Spt, salivary spermatheca; gland; Spt, L_CNS, spermatheca; larval central L_CNS, nervous larval sys- centralP1 nervous score. sys- When Rab2 was knocked down in ppk+ neurons, + ts P1 score.0.8 When Rab2 was knocked down in ppk neurons, tem. Mean 6 SDtem. is shown. Mean 6 SD is shown. GAL80 ; 0.084 0.179 knockdown femalesknockdown had a females significant had increasea significant in P1 increase score in P1 score P1

Tubulin- 8,569 6,904 compared to controlcompared0.4 females to control [control females median [control P1 = median 0.685, P1 = 0.685, Future experimentationFutureGAL4/TM3, experimentation is neededSb to establish is needed a to functional establish aN functional= 38; knockdownN = 38; (KD) knockdown median (KD)P1 = median 0.836, N P1= = 40; 0.836,P = N = 40; P = role of these geneticrole of variants these genetic in sperm variants competition. in sperm competition.0.04] (Figure0.04] 3). Thus0.0 (Figure loss of 3).Rab2 Thusfunction loss of Rab2 in ppkfunction+ neurons in ppk+ neurons Control Cyp313a2 Functional testingFunctional testing results in anresults increase in in anﬁ increaserst-male in progenyﬁrst-male compared progeny to compared to Tubulin- 46 41 controls.0.8458 Knockdown of para in sensory neurons resulted controls.0.8 Knockdown of para in sensory neurons resulted ts To validate potentialToGAL80 validate roles; potential of neurological0.113 roles of genes neurological in0.119 sperm genesin in reduction sperm inin P1 reduction score compared in P1 score to that compared in control to that females in control females

competition incompetition females, we in used females, RNAi we to used ask whether RNAi to re- ask whether re- P1 Tubulin- 6,350 5,742 (control median(control P10.4 = median 0.580, P1N = = 40; 0.580, KDN median= 40; P1 KD = median P1 = duction in theductionGAL4/TM3, expression in the ofSb expression these genes of changes these genes sperm changes0.506, spermN = 42;0.506,P = 0.008)N = 42; (FigureP = 0.008) 3). Similarly, (Figure knockdown3). Similarly, knockdown competition outcomes.competition These outcomes. candidates These could candidates be poten- could be poten- fi of Rim in sensoryof Rim0.0 neuronsin sensory also neuronsresulted also in a resulted signi cant in re- a significant re- tially involvedtially in the involved development in the development of neural networks of neural networks duction in P1 scoreduction (control in P1Control score median (controlCyp313a2 P1 = median0.560, N P1= = 44; 0.560, KD N = 44; KD fi speci cforspermcompetitionand/ortheycouldfunctionspeciTubulinficforspermcompetitionand/ortheycouldfunction- 37 48 median0.9591 P1 =median 0.519, N P1= = 41; 0.519,P = 0.017)N = 41; (FigureP = 0.017) 3). Knock- (Figure 3). Knock- at the time ofat sperm the time competition. of sperm Given competition. that knockdown Given that knockdown + + GAL80ts; 0.264 0.356 down of paradownor Rim0.8 ofinparappkorneuronsRim in ppk resultsneurons in production results in production of at least someof atof least these some neural of genes these neuralis lethal genes or affects is lethalof or fewer affectsfirst-maleof fewer progenyfirst-male after progeny a second after mating. a second mating. Tubulin- 4,702 5,824 P1 behaviors unrelatedbehaviors to sperm unrelated competition to sperm (e.g. competition, locomotion (e.g., locomotionWe observedWe no0.4 observeddifference no in difference P1 score when in P1SK scorewas when SK was GAL4/TM3,Sb + behaviors thatbehaviors could affect that courtship could affect or mating) courtship (Loughney or mating) (Loughneyknocked downknocked in ppk downneurons in ppk (control+ neurons median (control P1 = median P1 = 0.0 et al. 1989; Lloydet al.et1989; al. 2000; Lloyd Schulteet al. 2000;et al. 2010), Schulte weet choseal. 2010),0.578, we choseN = 40;0.578, KDN median= 40; P1 KD = median 0.578, P1N = 0.578,40; P =N = 40; P = fi Control Cyp313a2 to test a specito chypothesis:thatneuralgenesmightin- test a specifichypothesis:thatneuralgenesmightin-0.344) (Figure0.344) 3). The (Figure previous 3). The success previous of this successppk-GAL4 of this ppk-GAL4 fl uence spermfl uence competition sperm through competition neurons through that neurons inner- thatdriver inner- and thedriver lethality and of the a lethality ubiquitous of aknockdown ubiquitous of knockdownSK of SK vate the femalevate reproductive the female tract.reproductive This approach tract. This likely approach(data likely not shown)(data suggest not shown) that knockdown suggest that is knockdown occurring, al-is occurring, al- underestimates the extent of contribution of these genes underestimates the extent of contribution of thesethough genes we cannotthough gauge we cannot its extent. gauge It is its unclear extent. whether It is unclearSK whether SK that we found, but is a simple, initial, and direct test of that we found, but is a simple, initial, and directknockdown test of levelknockdown was insuf levelficient was to insuf givefi acient sperm to give competi- a sperm competi- function. function. tion phenotype,tion whether phenotype, SK’s whether role in sperm SK’s role competition in sperm is competition is Candidate gene function was tested by knocking down + Candidate gene function was tested by knockingnot through down ppknot throughneurons,ppk whether+ neurons, compensatory whether compensatory mecha- mecha- expression in the ppk+ sensory neurons+ that innervate the expression in the ppk sensory neurons that innervatenisms are the at play,nisms or are whether at play, the or SK whether association the SK was association a false was a false female reproductive tract. We tested candidates in17 these female reproductive tract. We tested candidatespositive. in these positive. neurons becauseneuronsSPR,theonlyfemalegeneknowntobe because SPR,theonlyfemalegeneknowntobe important forimportant sperm competition, for sperm signalscompetition, through signals them through them Discussion Discussion (Hasemeyer et(Hasemeyer al. 2009; Yanget al.et2009; al. 2009). Yang et Knockdown al. 2009). Knockdown was achievedwas with achieved the ppk-GAL4 with thedriverppk-GAL4 that isdriver specifically that is speciDespitefically overwhelmingDespite overwhelming evidence that evidence the female that genotype the female genotype expressed in thisexpressed subset inof this neurons subset (Grueber of neuronset al. (Grueber2007). etaffects al. 2007). spermaffects competition sperm (Clark competition and Begun (Clark 1998; and Begun Clark 1998; Clark Knockdown femalesKnockdown were comparedfemales were to controls compared in to a sperm controls inet a al. sperm1999; Fiumeraet al. 1999;et al. Fiumera2005, 2007;et al. Civetta2005, 2007;et al. Civetta2008; et al. 2008; competition experimentcompetition identical experiment to the identical DGRP experimentto the DGRP experimentChow et al. 2010),Chow littleet al. is2010), known little about is knownthe molecular about the mech- molecular mech- described above.described Because above. we sought Because to we test sought candidates to test in candidatesanisms that in underlieanisms thesethat underlie female contributions.these female contributions. We sought We sought neurons (ratherneurons than muscle), (rather than we chose muscle), candidate we chose genes candidateto identify genes potentialto identify candidate potential pathways candidate and pathways molecules and in molecules in with clear functionswith clear in neurons. functions Four in neurons. of the neurological Four of the neurologicalthe female thatthe play female a role that in spermplay a role competition. in sperm To competition. do this, To do this, candidates, withcandidates, various levels with variousof signifi levelscance of in signi thefi associ-cance in thewe associ-took advantagewe took of advantage the variation of the captured variation in the captured DGRP in the DGRP ation study (Figureation S4 study), were (Figure chosen S4), for were functional chosen for testing: functionalresource testing: (Mackayresourceet al. (Mackay2012). Theet al. DGRP2012). is Thea collection DGRP is of a collection of Rab2, SK, paraRab2, and, SKRim, .para, and Rim. 192 wild-derived192 wild-derivedinbred lines inbredfrom a lines single from population. a single population.

182 C. Y. Chow,182 M. F.C. Wolfner, Y. Chow, and M. A. F. G. Wolfner, Clark and A. G. Clark Table 2 (continued) Shab Tubulin- 62 66 0.157 ts GAL80 ; 0.084 0.07 0.8

Tubulin- 8,569 7,290 P1 0.4 GAL4/TM3,Sb 0.0 Control Shab sima Tubulin- 62 36 0.5363 ts

GAL80 ; 0.084 0.111 0.8 Tubulin- 8,569 3,882 P1 GAL4/TM3,Sb 0.4 0.0 Control sima spz5 Tubulin- 46 16 0.9049 0.6 GAL80ts; 0.113 0.069 P1 Tubulin- 6,350 1,190 0.3

GAL4/TM3,Sb 0.0 Control spz Tubulin- 62 41 0.4446 ts GAL80 ; 0.084 0.075 0.8

Tubulin- 8,569 4,894 P1 GAL4/TM3,Sb 0.4 0.0 Control spz5 Tubulin- 37 16 0.325 GAL80ts; 0.264 0.249 0.8 P1

Tubulin- 4,702 1,981 0.4 GAL4/TM3,Sb 0.0 Control spz5 CG42796 Tubulin- 62 40 0.2161 ts GAL80 ; 0.084 0.051 0.8

Tubulin- 8,569 2,856 P1 GAL4/TM3,Sb 0.4 0.0 CG42796 Control CG10962 Tubulin- 85 36 0.0707 GAL80ts; 0.3 0.155 0.8 P1 Tubulin- 7,812 2,217 0.4

GAL4/TM3,Sb 0.0 CG10962 Control CG32532 Tubulin- 45 31 0.6465 GAL4/TM3,Sb 0.394 0.375 0.8 P1

6,486 2,936 0.4

0.0 CG32532 Control

18 Table 2 (continued) uif Tubulin- 44 30 0.5673 0.8 GAL80ts; 0.226 0.234 P1 Tubulin- 6,255 2,927 0.4 GAL4/TM3,Sb 0.0 Control uif CG9850 Tubulin- 44 39 0.5681 GAL80ts; 0.226 0.143 0.8 P1 Tubulin- 6,255 4,397 0.4 GAL4/TM3,Sb 0.0 CG9850 Control CG15800 Tubulin- 36 48 0.5083 GAL4/TM3,Sb 0.392 0.333 0.8 P1

4,101 4,349 0.4 0.0 CG15800 Control Rbp6 Tubulin- 41 27 0.8464 0.8 GAL80ts; 0.186 0.196 P1

Tubulin- 4,913 2,780 0.4 GAL4/TM3,Sb 0.0 Control Rbp6 CG32264 nSyb-GAL4 51 25 0.4775 0.134 0.244 0.8

7,160 3,355 P1 0.4 0.0 CG32264 Control CG15765 nSyb-GAL4 51 40 0.7024 0.134 0.13 0.8

7,160 5,534 P1 0.4 0.0 CG15765 Control CG33095 Tubulin- 61 27 0.2967 GAL4/TM3,Sb 0.384 0.269 0.8

7,828 3,014 P1 0.4 0.0 CG33095 Control CG32298 Tubulin- 61 41 0.3175 GAL4/TM3,Sb 0.384 0.263 0.8

7,828 4,452 P1 0.4 0.0 CG33298 Control

19 Table 2 (continued) CG6163 Tubulin- 61 35 0.6127 GAL4/TM3,Sb 0.384 0.301 0.8

7,828 4,078 P1 0.4 0.0 CG6163 Control Zasp66 Tubulin- 29 60 0.0642

GAL4/TM3,Sb 0.298 0.221 0.8 2,402 6,209 P1 0.4 0.0 Control Zasp66 Dh44 ǁ nSyb-GAL4 66 67 0.2965 0.095 0.103 0.8 8,010 8,879 P1 0.4 0.0 Control Dh44

20 Octopamine and Tyramine may play a role in sperm dynamics

Octopamine and tyramine have been previously been shown to be important for sperm storage (Avila 2012). To examine whether they are also important for differential sperm storage/use from two different males, we used fly lines mutant in genes that encode enzymes that synthesize tyramine (tyrosine decarboxylase 2 [tdc2], Cole et al. 2005) and octopamine

(tyramine β-hydroxylase [tβh], Monastirioti et al. 1996). We mated females lacking octopamine

(tβh), or tyramine and octopamine (tdc2) to males with green- or red-labeled sperm (Protamine-

GFP [first male] and Protamine-RFP [second male], Manier et al. 2010), and examined sperm storage 1, 4, and 10 days after the second mating.

One day after the second mating, tβh females (lacking octopamine) have significantly more first male sperm in the seminal receptacle, less second male sperm in the seminal receptacle, and more total sperm compared to controls. By 4 days after the second mating, octopamine mutant females are overall maintaining more sperm in storage compared to controls.

The most striking difference here is that controls have much less sperm remaining in the seminal receptacle at this time point. Ten days after the second mating, controls have almost no sperm remaining in the seminal receptacle, while mutants still retain first male’s sperm. No second male’s sperm is detected in mutants by 10 days after the second mating, so it appears that mutants are also releasing second male sperm faster than controls (Figure 2, Table 3). These results are similar for tdc2 females (lacking tyramine and octopamine; Figure 3, Table 4).

21 Octopamine Tyramine Seminal Receptacle Spermathecae Seminal Receptacle Spermathecae 400 300

300 1st Male 1st Male 200 200 100 100 Sample Sample Mutant (tdc2/Df(2R)42) 0 Mutant (tβh/tβh) 0 400 Control-1 (tdc2/SM5) Control (tβh/FM7) 300 Control-2 (Df(2R)42/Gla) 300 2nd Male 2nd Male Sperm Number Sperm Number Sperm 200 200 100 100

0 0 1day 4day 10day 1day 4day 10day 1day 4day 10day 1day 4day 10day Days after mating Days after mating

Figure 2. Control and tβh mutant females’ sperm storage from 2 different males. ProtamineB- GFP is the first male, ProtamineB-RFP is the second male.

Table 3. Control and tβh mutant females’ sperm storage from 2 different males. ProtamineB- GFP is the first male, ProtamineB-RFP is the second male. Means are represented in the table (variation can be observed in Figure 2). N, number of females (Mutant, Control) NS, not significant *Significant result Seminal Receptacle Spermathecae 1 day 4 day 10 day 1 day 4 day 10 day N (16,19) N (18,16) N (24,32) N (16,19) N (18,16) N (24,32) Mutant 374.3 277.7 206 332.6 197.4 96.7 1st male Control 281.1 70.9 0.6 315.7 263.1 203.7 p-value 4.9e-09* 2.2e-16* 4.1e-10* 0.0983 7.6e-10* 1.8e-14* Mutant 302.5 133.2 0 178.2 73.1 0 2nd male Control 345.9 92.5 6.1 180.1 99.9 42.2 p-value 1.9e-05* 3.1e-05* 3.6e-08* 0.806 5.9e-05* 1.6e-08*

22 Octopamine Tyramine Seminal Receptacle Spermathecae Seminal Receptacle Spermathecae 400 300

0 0 1day 4day 10day 1day 4day 10day 1day 4day 10day 1day 4day 10day Days after mating Days after mating Figure 3. Control and tdc2 mutant females’ sperm storage from 2 different males. ProtamineB- GFP is the first male, ProtamineB-RFP is the second male.

Table 4. Control and tdc2 mutant females’ sperm storage from 2 different males. ProtamineB- GFP is the first male, ProtamineB-RFP is the second male. Means are represented in the table (variation can be observed in Figure 3). N, number of females (Mutant, Control-1, Control-2) NS, not significant *Significant result Seminal Receptacle Spermathecae 1 day 4 day 10 day 1 day 4 day 10 day N(12,9,7) N(14,8,9) N(16,7,11) N(12,9,7) N(14,8,9) N(16,7,11) Mutant 294.3 191.8 165.8 282.6 229.8 163.6 Control-1 307.3 108.9 1.6 265.2 154.1 74.1 Control-2 296.6 109.2 2.3 270.1 124.7 68.7 1st male p-value 0.1935 6.7e-12* 2.2e-16* 0.0366* 1.2e-08* 2.1e-09* (Control-1, 0.7801 1.4e-12* 2.2e-16* 0.2485 2.3e-08* 2.2e-16* Control-2) Mutant 256.6 136.3 0 167.3 87.6 0 Control-1 287.7 84.4 8.6 139.8 76.3 23.3 Control-2 288.7 78.8 8.7 161.1 83 21.4 2nd male p-value 0.0096* 1.8e-07* 0.0186* 0.0791 0.0339* 0.0268* (Control-1, 0.0312* 1.2e-06* 5.8e-05* 0.686 0.5586 0.0035* Control-2)

23 DISCUSSION

Although there is copious evidence that female genotype is important for sperm competition outcome (Clark and Begun 1998; Clark et al. 1999; Chow et al. 2010), very little is known about the specific genes and mechanisms involved in the female contribution. Here, we show that 10 of 29 candidate genes when knocked down in females affect P1 score, the proportion of offspring sired by the first male in doubly mated females.

Many complex interactions occur between males and females, and the genes we identified as involved in sperm competition outcome could be acting at any point in time before, during, or after mating. For example, a candidate gene may be involved in pre-copulatory interactions, such as detecting males’ courtship behaviors or pheromones. A significant effect on

P1 score could also arise from variation in remating behavior, copulation duration, or other nervous system controlled functions such as allocation of sperm into storage, their retention there, or their release from storage. Further tests should be conducted with these candidate genes to determine through which neurons they are acting and if they may have an effect on pre- copulatory behaviors.

Here, we also show that the neuromodulators octopamine and tyramine may play a role in sperm dynamics as there are significant differences in how mutant and control females store sperm from two different types of males. This is a possible physiological mechanism by which females actively modulate sperm use from different males. In the future, it would be interesting to test if the 10 candidate genes we confirmed are important in octopaminergic and/or tyraminergic neurons for sperm competition outcome.

Both of these sets of experiments suggest that one role of the female nervous system is to bias sperm use, suggesting a more active role in sperm competition by females than previously

24 thought. Future studies should focus on how these candidates are affecting sperm competition and through which neurons they are acting. Then, when we have a more comprehensive view of how males and females affect sperm competition outcome separately, we can begin to examine more complex interactions between males and females.

25 REFERENCES

Allen AK and Spradling AC. 2008. The Sf1-related nuclear hormone receptor Hr39 regulates Drosophila female reproductive tract development and function. Development. 135: 311- 321.

Avila FW, Bloch Qazi MC, Rubinstein CD, and Wolfner MF. 2012. A requirement for the neuromodulators octopamine and tyramine in Drosophila melanogaster female sperm storage. Proc. Natl. Acad. Sci. USA 109: 4562–4567.

Chapman T, Liddle LF, Kalb JM, Wolfner MF, and Partridge L. 1995. Cost of mating in Drosophila melanogaster females is mediated by male accessory gland products. Nature 373: 241-244.

Chow CY, Wolfner MF, and Clark AG. 2010. The genetic basis for male x female interactions underlying variation in reproductive phenotypes of Drosophila. Genetics 186: 1355- 1365.

Chow CY, Wolfner MF, and Clark AG. 2013. Large neurological component to genetic differences underlying biased sperm use in Drosophila. Genetics 193: 177-185.

Civetta A, Rosing K, and Fisher J. 2008. Differences in sperm competition and sperm competition avoidance in Drosophila melanogaster. Anim. Behav. 75: 1739–1746.

Clark AG, Aguadé M, Prout T, Harshman LG and Langley CH. 1995. Variation in sperm displacement and its association with accessory gland protein loci in D. melanogaster. Genetics 139:189-201.

Clark AG and Begun DJ. 1998. Female genotypes affect sperm displacement in Drosophila. Genetics 149: 1487-1493.

Clark AG, Begun DJ, and Prout T. 1999. Female x male interactions in Drosophila sperm competition. Science 283: 217-220.

Clark J and Lange AB. 2003. Octopamine modulates spermathecal muscle contractions in Locusta migratoria. J Comp Physiol A 189: 105-114.

Cole SH, Carney GE, McClung CA, Willar SS, Taylor BJ, and Hirsh J. 2005. Two functional but noncomplementaing Drosophila tyrosine decarboxyloase genes: distinct roles for neural tyramine and octopamine in female fertility. J. Biol. Chem. 280: 14948-14955.

Fiumera AC, Dumont BL, and Clark AG. 2005. Sperm competitive ability in Drosophila melanogaster associated with variation in male reproductive proteins. Genetics 169: 243- 257.

26 Fiumera AC, Dumont BL, and Clark AG. 2006. Natural variation in male-induced 'cost-of- mating' and allele-specific association with male reproductive genes in Drosophila melanogaster. Philos Trans R Soc London, Ser B 361: 355-361.

Fiumera AC, Dumont BL, and Clark AG. 2007. Associations between sperm competition and natural variation in male reproductive genes on the third chromosome of Drosophila melanogaster. Genetics 176: 1245-1260.

Greenspan L and Clark, AG. 2011. Associations between Variation in X Chromosome Male Reproductive Genes and Sperm Competitive Ability in Drosophila melanogaster. International Journal of Evolutionary Biology, Article ID 214280. doi:10.4061/2011/214280

Heifetz Y, Lindner M, Garini Y, and Wolfner MF. 2014. Mating regulates neuromodulator ensembles at nerve termini innervating the Drosophila reproductive tract. Curr Biol 24: 731-737.

Hindle S, Afsari F, Stark M, Middleton CA, Evans GJ, Sweeney ST, and Elliott CJ. 2013. Dopaminergic expression of the Parkinsonian gene LRRK2-G2019S leads to non- autonomous visual neurodegeneration, accelerated by increased neural demands for energy. Hum. Mol. Genet. 22: 2129--2140.

Imhof M, Harr B, Brem G, and Schlötterer C. 1998. Multiple mating in wild Drosophila melanogaster revisited by microsatellite analysis. Molecular Ecology 7: 915-917.

Kapelnikov A, Zelinger E, Gottlieb Y, Rhrissorrakrai K, Gunsalus KC, and Heifetz Y. 2008. Mating induces an immune response and developmental switch in the Drosophila oviduct. Proc Natl Acad Sci USA. 105: 13912-13917.

Lee KM, Daubnerova I, Isaac RE, Zhang C, Choi S, Chung J, and Kim YJ. 2015. A Neuronal Pathway that Controls Sperm Ejection and Storage in Female Drosophila. Current Biology 25: 790-797.

Lange AB and da Silva R. Neural and hormonal control of muscular activity of the spermatheca in the locust, Locusta migratoria. Peptides 28: 174-184.

Mack PD, A Kapelnikov, Y Heifetz and Bender M. 2006. Mating-responsive genes in reproductive tissues of female Drosophila melanogaster. Proc Natl Acad Sci USA 103: 10358-10363.

Mackay TFC, Richards S, Stone EA, Barbadilla A, Ayroles JF et al. 2012. The Drosophila melanogaster Genetic Reference Panel. Nature 482: 173–178.

Manier MK, Belote JM, Berben KS, Novikov D, Stuart WT, and Pitnick S. 2010. Resolving mechanisms of competitive fertilization success in Drosophila melanogaster. Science 328: 354–357.

Matthews BJ, Kim ME, Flanagan JJ, Hattori D, Clemens JC, Zipursky SL, and Grueber WB. 2007. Dendrite self-avoidance is controlled by Dscam. Cell 129: 593--604.

Middleton CA, Nongthomba U, Parry K, Sweeney ST, Sparrow JC, and Elliott CJH. 2006. Neuromuscular organization and aminergic modulation of contraction in the Drosophila ovary. BMC Biology 4:17.

Milkman R and Zeitler RR. 1974. Concurrent multiple paternity in natural and laboratory populations of Drosophila melanogaster. Genetics 78: 1191-1193.

Monastirioti M, Linn CE, and White K. 1996. Characterization of Drosophila tyramine beta- hydroxylase gene and isolation of mutant flies lacking ocptopamine. J Neurosci 16:3900- 3911.

Mueller JL, Page JL, and Wolfner MF. 2007. An ectopic expression screen reveals the protective and toxic effects of Drosophila seminal fluid proteins. Genetics 175: 777–783.

Parker GA. 1970. Sperm competition and its evolutionary consequences in the insects. Biol. Rev. 45: 525-567.

Prokupek A, Hoffmann F, Eyun SI, Moriyama E, Zhou M, Harshman L. 2008. An evolutionary expressed sequence tag analysis of Drosophila spermatheca genes. Evolution 62: 2936– 2947.

Prokupek AM, Kachman SD, Ladunga I, and Harshman LG. 2009. Transcriptional profiling of the sperm storage organs of Drosophila melanogaster. Insect Mol. Biol. 18: 465–475.

Prokupek AM, Eyun SI, Ko L, Moriyama EN, and Harshman LG. 2010. Molecular evolutionary analysis of seminal receptacle sperm storage organ genes of Drosophila melanogaster. J. Evol. Biol. 23: 1386–1398.

Prout T and Bundgaard J. 1977. The population genetics of sperm displacement. Genetics 85: 95-124.

Rubinstein CD and Wolfner MF. 2013. Drosophila seminal fluid protein ovulin mediates ovulation through female octopamine neuronal signaling. PNAS 110: 17420-5.

Sitnik JL, Francis C, Hens K, Huybrechts R, Wolfner MF, and Callaerts P. 2014. Neprilysins: An Evolutionarily Conserved Family of Metalloproteases That Play Important Roles in Reproduction in Drosophila. Genetics 196: 781-797.

Swanson WJ, A Wong, MF Wolfner, and Aquadro CF. 2004. Evolutionary expressed sequence tag analysis of Drosophila female reproductive tracts identifies genes subjected to positive selection. Genetics 168: 1457-1465.

28 Wigby S and Chapman T. 2004. Female resisitance to male harm evolves in response to manipulation of sexual conflict. Evolution 58: 1028-1037.