Obp56h Modulates Mating Behavior in Drosophila melanogaster

Social interactions in insects are driven by conspecific chemical signals that are detected via olfactory and gustatory neurons. Odorant binding proteins (Obps) transport volatile odorants to chemosensory receptors, but their effects on behaviors remain poorly characterized. Here, we report that RNAi knockdown of Obp56h gene expression in Drosophila melanogaster enhances mating behavior by reducing courtship latency. The change in mating behavior that results from inhibition of Obp56h expression is accompanied by significant alterations in cuticular hydrocarbon (CHC) composition, including reduction in 5-tricosene (5-T), an inhibitory sex pheromone produced by males that increases copulation latency during courtship. Whole genome RNA sequencing confirms that expression of Obp56h is virtually abolished in Drosophila heads. Inhibition of Obp56h expression also affects expression of other chemoreception genes, including upregulation of lush in both sexes and Obp83ef in females, and reduction in expression of Obp19b and Or19b in males. In addition, several genes associated with lipid metabolism, which underlies the production of cuticular hydrocarbons, show altered transcript abundances. Our data show that modulation of mating behavior through reduction of Obp56h is accompanied by altered cuticular hydrocarbon profiles and implicate 5-T as a possible ligand for Obp56h.

sensilla (Vosshall et al. 1999;Clyne et al. 1999), which recognize volatile airborne odorants (Hallem and Carlson 2006); ionotropic receptors (Irs) expressed in coeloconic sensilla, which detect a diverse array of chemicals, including water-soluble compounds (Benton et al. 2009); and odorant-binding proteins (Obps). Odorantbinding proteins are soluble proteins secreted into the perilymph that surrounds the dendrites of olfactory sensory neurons. They are the first components of the chemosensory system to interact with airborne chemicals and facilitate the transport of hydrophobic odorants to their membrane-bound receptors (Wojtasek and Leal 1999;Xu et al. 2005). In contrast to extensive information about the molecular response profiles of membrane-bound gustatory and olfactory receptors, relatively little functional information is known about Obps.
Obps were first identified as pheromone-binding proteins in the antennae of the male silk moth, Bombyx mori (Vogt and Riddiford 1981), where pH-induced conformational changes mediate binding and release of the pheromone (Wojtasek and Leal 1999). In the silk moth group Antheraea, two pheromone binding proteins showed preferential binding to specific components of an acetate and aldehyde pheromone blend (Maida et al. 2003). In Drosophila sechellia, Obp57d and Obp57e have been implicated in host plant preference in Drosophila by affecting the perception of octanoic and hexanoic acids (Matsuo et al. 2007;Matsuo 2012).
The D. melanogaster genome encodes a family of 51 Obp genes (Hekmat-Scafe et al. 2002), which has evolved through gene duplication and subsequent subfunctionalization (Vieira et al. 2007). Overall, Obp genes are structurally diverse, with an average amino acid identity of 10-15%, but can range from 4 to 60% (Zhou 2010). Several Obp family members show distinct expression patterns in the antenna (McKenna et al. 1994;Shanbhag et al. 2001), including OS-E (Obp83b), OS-F (Obp83a), lush (Obp76a), PBPRP-2 (Obp19d), and PBPRP-5 (Obp28a), and eight have been identified in antennal extracts by high-performance liquid chromatography and mass spectrometric analyses (Anholt and Williams 2010). Despite the genetic divergence of Obp genes, they are often highly correlated at the level of gene expression (Zhou et al. 2009).
Several studies have documented the role of D. melanogaster Obps in olfactory behavior. Natural variation in Obp genes is associated with variation in olfactory responses to benzaldehyde and acetophenone (Wang et al. 2007Arya et al. 2010). Obp-dependent odorant recognition appears to be combinatorial. Behavioral responses to 16 ecologically relevant odorants tested across 17 knockdown Obp RNAi lines revealed that some Obp genes had altered behavioral responses to multiple odorants, and some odorants had altered behavioral responses in several Obp knockdown lines (Swarup et al. 2011). This suggests that individual odorants may interact with multiple Obps, and individual Obps may interact with multiple odorants.
There is increasing evidence that Obps have diverse pleiotropic functions in D. melanogaster not limited to olfaction. First, expression of Obp genes is not restricted to olfactory tissues; for example, Obp8a is expressed in the male accessory gland (Arya et al. 2010;St. Pierre et al. 2014). Second, expression of Obp genes is genetically correlated with expression of other genes that are enriched for diverse gene ontology categories including synaptic transmission, detection of signals regulating tissue development and apoptosis, postmating behavior and oviposition, and nutrient sensing (Arya et al. 2010). Third, different physiological and social conditions modulate expression of Obp genes (Zhou et al. 2009). Fourth, there is direct evidence that Obp genes are associated with other traits, including gustatory responses to tastants (Swarup et al. 2014) and lifespan (Arya et al. 2010).
Here, we show that RNAi-mediated suppression of the expression of Obp56h reduces copulation latency, and this behavioral effect is accompanied by an alteration in the composition of CHCs, notably a reduction in the male sex pheromone 5-tricosene (5-T). Using RNA-seq analysis, we observe a number of differentially expressed genes, including Or19b, and several genes associated with lipase activity. Our results suggest that Obp56h may be associated with pheromone production and affect social recognition via pheromone perception.
All stocks were reared on cornmeal/molasses/agar medium and maintained under standard culture conditions (25°, 12:12 hr light/ dark cycle; lights on at 6:00 AM) in an environmentally controlled walk-in incubator.

Behavioral assays
All behavioral assays were performed on F 1 individuals from crosses of UAS-RNAi Obp lines to Tub-GAL4 and/or Dll-GAL4 lines. F 1 individuals with CyO or TM3 balancer genotypes were discarded and not assessed. CO 2 was used as an anesthetic; however, anesthesia exposure was withheld 24 hr prior to behavioral assays. All behavioral assays were conducted in a behavioral chamber (25°) between 8:00 AM and 11:00 AM. We assessed copulation latency, and phototaxis and geotaxis as additional sensorimotor behaviors. Unless otherwise specified, we used one-way fixed effect ANOVA models of the form Y = m + G + e where Y is the phenotype, m is the overall mean, G is the genotype, and e is the within-genotype residual variance; and/or t-tests to evaluate significant differences in behavior among genotypes. All statistical analyses were conducted using SAS (SAS Institute Inc. 2011) software.
Copulation latency: To assess mating behavior, we paired five males and five virgin females aged 3-7 d together in a vial and recorded copulation latency for 30 min. Once a pair engaged in copulation, they were removed from the vial with a mouth aspirator and the remaining flies were observed for the remaining time. We performed at least 40 replicates with five males and five females each per genotype.

Geotaxis:
We assessed geotaxis behavior on individual flies by measuring distance traveled upwards following a sudden disturbance. Flies were placed in 25 mm · 150 mm glass vials (Pyrex-Corning flat bottom) with a ruler marking 5-mm increments from 0, indicating the lowest position, to 24, indicating the highest position. Each fly was tapped to the bottom of the vial, and the distance traveled upwards was scored based on the highest point reached in 5 sec. Twenty individual flies were assayed each day for 3 d, creating a total sample size of 60 per sex per genotype.
Phototaxis: We assessed phototaxis behavior using a "countercurrent apparatus" (Benzer 1967). Each replicate per genotype consisted of $50 3-to 7-d-old flies of the same sex; we performed three replicates per sex and genotype across 3 d. Flies were darkadapted for 30 min prior to performing the assay in a dark room. To assess phototaxis, we tapped flies to the bottom of the first start tube and placed the apparatus horizontally with the distal tubes 5 cm away from a 15-W fluorescent light. The flies were given 15 sec to reach the distal tube. We repeated this procedure seven more times, so that flies could choose to go toward the light a maximum of eight times. At the end of each trial, we collected all flies into the start tubes, removed the start tubes from the apparatus and froze them at 280°for $30 min before counting the number of flies in each tube. The phototaxis score was analyzed by ANOVA according to the factorial mixed model Y = m + G + S + G · S + R(G · S) + e, where Y is the observed value, m is the overall mean, and G, S, and R denote genotype, sex, and replicate, respectively, and e is the residual experimental error. Genotype and sex are fixed effects and replicate is random.

Cuticular hydrocarbon analysis
Cuticular hydrocarbon analysis was performed as described previously (Dembeck et al. 2015). We performed two separate experiments, one with Dll-GAL4 · UAS-Obp56hRNAi and Dll-GAL4 · control F 1 virgin males, and one with Tub-GAL4 · UAS-Obp56hRNAi and Tub-GAL4 · control F 1 virgin males. All males were collected at eclosion and placed in mixed sex groups with five males and five females of the same genotype for 3 d prior to collection for CHC analysis. The flies were separated into three replicate samples per line, with five flies per replicate. To ensure cuticular lipid contamination did not occur, a fresh paper tissue was placed on the carbon dioxide pad and the flies were handled with acetone-washed titanium forceps at each round of sorting. All samples were stored in 2-ml glass auto-injection vials with a Teflon cap and were flash frozen and stored at 230°until cuticular lipid extraction.
Cuticular lipids were extracted from each sample using 200 ml of hexane containing an internal standard (IS, 1 mg n-C32) with gentle swirling for 5 min. The flies were briefly extracted a second time with 100 ml of hexane (free of internal standard). After each wash the extract was transferred to a 300 ml conical glass insert. The extract was dried using a gentle stream of high-purity N 2 and resuspended in 50 ml of hexane. The samples were immediately processed using gas chromatography or stored at 4°(no longer than 1 d) until processing.
The cuticular lipid extracts were analyzed using an Agilent 7890A gas chromatograph with a DB-5 Agilent capillary column (20 m · 0.18 mm · 180 mm) and a flame ionization detector (FID) for quantification. We introduced 1 ml of sample using an Agilent 7683B autoinjector into a 290°inlet operated in splitless mode. The split valve was turned on after 1 min. The oven temperature program was as follows: 50°for 1 min, increased at 20°/min to 150°, and increased at 5°/min to 300°followed by a 10-min hold. Hydrogen was used as the carrier gas at constant flow (average linear velocity = 35 cm/sec) and the FID was set at 300°. Compound identifications were based on a previous GC-MS analysis (Dembeck et al. 2015). All chromatograms were analyzed using Agilent ChemStation software. The data were represented as proportions by dividing each peak area by the total sum of all integrated peaks. We analyzed differences in CHCs between Obp56h-RNAi knockdown flies and controls using t-tests (SAS 9.3). Principal component analysis was conducted on the correlation matrix of the proportions of CHCs quantified in each sample in JMP v.10.

Gene expression analysis
We used RNA-seq to quantify differences in gene expression in heads and bodies of males and females of Dll-GAL4 · UAS-Obp56h and Dll-GAL4 · control F 1 individuals. F 1 individuals with CyO or TM3 balancer genotypes were discarded. Flies were aged for 5-6 d in a mixed sex environment at a density of $20 in a vial. Flies were flash frozen over dry ice between 8:00 AM and 11:00 AM and 30 heads and bodies per sex and genotype were manually dissected and collected over 3 d in a randomized design, with four biological replicates per sex, genotype, and tissue.
We extracted total RNA with Trizol with the Quick-RNA MiniPrep kit (Zymo Research; R1055). rRNA was depleted using the Ribo-Zero Gold rRNA Removal Kit (Human/Mouse/Rat) (Illumina) with 5 mg total RNA input. Depleted mRNA was fragmented and converted to first strand cDNA using SuperScript III Reverse Transcriptase (Thermo Fisher Scientific). During the synthesis of second strand cDNA, dUTP instead of dTTP was incorporated to label the second strand cDNA (Thermo Fisher Scientific). cDNA from each RNA sample was used to produce barcoded cDNA libraries using NEXTflex DNA Barcodes (Bioo Scientific) with an Illumina TrueSeq compatible protocol. Library size was selected using Agencourt Ampure XP Beads (Beckman Coulter) and centered on 250 bp with average insert size around 130 bp. Second strand DNA was digested with Uracil-DNA glycosylase before amplification to produce directional cDNA libraries. Libraries were quantified using Qubit dsDNA HS Kits (Life Technologies) and Figure 1 Effect of Obp56h-RNAi knockdown on copulation latency. Five males and five virgin females aged 3-7 d were placed together in a vial and copulation latency was recorded for 30 min. We performed at least 40 replicates per genotype (i.e., 200 males and 200 females total per genotype). Red bars denote Dll-GAL4/Obp56h-RNAi and blue bars denote Dll-GAL4/Control F 1 genotypes; red and blue stacked bars denote pairs of flies with different male and female genotypes. Ã P , 0.05; NS, not significant.
Bioanalyzer (Agilent Technologies) to calculate molarity. They were then diluted to equal molarity and requantified, and 32 libraries were pooled. Pooled library samples were quantified to calculate final molarity and finally denatured and diluted to 14 pM. Pooled library samples were clustered on an Illumina cBot and sequenced on an Illumina Hiseq2500 using 125-bp single-read v4 chemistry on each of two lanes.
The quality of the RNA-seq data was assessed using FASTQC (Andrews 2010). Following assessment that the data were of high quality, adapter sequences were trimmed using Cutadapt (Martin 2011). Ribosomal reads were filtered against a database of the most common ribosomal sequences using fast BWA alignment BWA-0.7.10 (Li and Durbin 2009). The remaining reads were aligned to the Dmel_r5.57_FB2014_03 genome and transcriptome using STAR_2.4.1d (Dobin et al. 2013). All individual RNA-seq samples had .12 million reads after standard filtering by quality scores and after filtering out of residual rRNA sequences. Read counts for each gene in each sample were computed using HTSeq (v0.6.1p1) (Anders et al. 2015). R software was used for further quality assessment and statistical analysis (R-Core-Team 2012). The EDASeq package was used to plot principal components (Risso et al. 2011), and one replicate sample (HRNAiF1) was identified as a technical outlier, removed, and the remaining 31 samples were used for analysis. The edgeR package was used to calculate differential expression analysis for pairwise comparisons between the control and RNAi sample for sex and tissue as well as the interaction between genotype and tissue for each sex (Robinson et al. 2010). Biological pathway and gene ontology enrichment analyses were performed using DAVID (Huang et al. 2009).
Data availability RNA-seq data have been deposited in the Gene Expression Omnibus database under accession numbers GSM1959750-GSM1959781.

RESULTS AND DISCUSSION
RNAi knockdown of Obp56h reduces copulation latency Previously we investigated the functions of Drosophila Obps in olfaction by measuring responses of 17 Obp-RNAi lines to 16 odorants (Swarup et al. 2011). Quantification of expression of mRNA targets showed a major reduction in the expression of Obp56h. Subsequent behavioral studies using these Obp RNAi lines indicated that suppression of Obp56h expression could influence mating behavior. To further explore the role of Obp56h in mating behavior, we obtained a UAS-RNAi knockdown line targeting Obp56h, Obp56h KK111996 , and its co-isogenic control with an empty integration site (y,w 1118 ; P{attP,y + ,w 39 }) from the Vienna Drosophila Stock Center (http:// stockcenter.vdrc.at). We obtained two GAL4 driver strains from the Bloomington Drosophila Stock Center (http://flystocks.bio.indiana.edu/): a ubiquitously expressed tubulin-GAL4 driver line (y 1 w Ã ; P{tubP-GAL4} LL7/TM3, Sb 1 ) and a Dll-GAL4 driver, which has more restricted expression, including in the antennae, labium, legs, and wings (P{w[+mW.hs]=GawB}Dll md23 /CyO).
To assess whether reduced copulation latency from RNAi knockdown of Obp56h could be in part due to differences in chemical communication, we quantified CHC profiles of Dll-GAL4/Obp56h-RNAi knockdown and Dll-GAL4/Control males. We detected 42 CHCs Figure 3 Effects of Obp-RNAi knockdown with a Dll-GAL4 driver on CHC profiles. Cuticular hydrocarbon analysis was performed as described previously (Dembeck et al. 2015). (A) Proportion of 42 CHCs in Dll-GAL4/Obp56h-RNAi (red bars) and Dll-GAL4/Control (blue bars) F 1 males. ÃÃÃ P , 0.001, ÃÃ P , 0.01, Ã P , 0.05. NI, not identified. (B) Principal component biplots for PC1 and PC2 for Dll-GAL4/Obp56h-RNAi (red circles) and Dll-GAL4/Control (blue circles) F 1 males. The principal components analysis is a linear transformation used to reduce the dimensionality of the multivariate dataset. PC1 captures the variation in the data that can be attributed to genotype (control vs. RNAi knockdown) seen by the clustering of the samples into two distinct groups. (C) PC1 and PC2 eigenvectors. The eigenvectors are composed of the weights of each original variable in the linear combinations that define PC1 and PC2. The plot indicates which of the original variables are most strongly correlated with PC1 and PC2. The percent of variance explained by each PC is indicated on the x-and y-axes of panels (B) and (C).
( Figure 3). The two major male CHC sex pheromones, 7-T and 7-P, were not different between the two genotypes. However, 10 (23.8%) CHCs were significantly altered between the RNAi knockdown and the control (Figure 3). Eight of the ten significantly different CHCs increased relative to the control and are all n-alkanes (n-C21-n-C29, except n-C27). Of the two that decreased, one was a minor, unidentified Figure 4 Effects of Obp-RNAi knockdown with a Tub-GAL4 driver on CHC profiles. Cuticular hydrocarbon analysis was performed as described previously (Dembeck et al. 2015). (A) Proportion of 42 CHCs in Tub-GAL4/Obp56h-RNAi (red bars) and Tub-GAL4/Control (blue bars) F 1 males. ÃÃÃ P , 0.001, ÃÃ P , 0.01, Ã P , 0.05. NI, not identified. (B) Principal component biplots for PC1 and PC2 for Tub-GAL4/Obp56h-RNAi (red circles) and Tub-GAL4/Control (blue circles) F 1 males. The principal components analysis is a linear transformation used to reduce the dimensionality of the multivariate dataset. PC1 captures the variation in the data that can be attributed to genotype (control vs. RNAi knockdown) seen by the clustering of the samples into two distinct groups. (C) PC1 and PC2 eigenvectors. The eigenvectors are composed of the weights of each original variable in the linear combinations that define PC1 and PC2. The plot indicates which of the original variables are most strongly correlated with PC1 and PC2. The percent of variance explained by each PC is indicated on the x-and y-axes of panels (B) and (C). compound, the other was 5-T. 5-T is an inhibitory pheromone, produced primarily in males and only in small quantities in females, that is thought to delay the initiation of courtship in D. melanogaster and may serve to decrease the probability of male-male courtship in nature (Ferveur and Sureau 1996;Ferveur 1997;Waterbury et al. 1999). 5-T is one of the most volatile D. melanogaster CHCs and may be detected through olfaction rather than through contact (Ferveur and Sureau 1996;Ferveur 1997;Waterbury et al. 1999). The Dll-GAL4/Obp56h-RNAi knockdown males had about 19% less 5-T than the control males. To replicate these observations, we also determined CHC profiles of Tub-GAL4/Obp56h-RNAi knockdown and Tub-GAL4/Control males (Figure 4). With the ubiquitously expressed tubulin driver line 5-T was also reduced in Tub-GAL4/ Obp56h-RNAi knockdown males, this time by 32%. The alterations in CHC profiles were similar between the two drivers: nine of the 10 significantly changed CHCs using the Dll-GAL4 driver also changed when reduction in Obp56h expression was driven by Tub-GAL4. Thus, interference with chemosensory input through Obp56h, and possibly other functions of Obp56h, resulted in systemic alterations in CHC biosynthesis.
Genome-wide changes in gene expression caused by RNAi knockdown of Obp56h Understanding which genes are coregulated when Obp56h expression is reduced by RNAi knockdown can give insights into the biological processes through which Obp56h affects mating behavior. Therefore, we performed RNA-seq analysis for Dll-GAL4/Control and for Dll-GAL4/Obp56h-RNAi males and females, separately for heads and bodies (Supplemental Material, Table S1).
Based on the 17,055 FlyBase IDs indicated in Table S1, we performed gene ontology enrichment analyses (Huang et al. 2009) for genes with differential expression between the Obp56h-RNAi and control genotypes in heads and bodies (Table S3). The most enriched categories in heads and female bodies comprised genes associated with immune/defense responses, which may participate in removal of xenobiotics, including odorants. In addition, and consistent with changes in CHCs, genes associated with the gene ontology terms of lipase, triglyceride lipase activity, and phospholipase activity were also enriched in the bodies of both males and females. Four genes with decreased triglyceride lipase activity and phospholipase activity in male bodies (CG11598, CG6271, CG6277, CG6283) are interesting since lipases modify lipids and fatty acids, which are precursors of insect CHCs (Howard and Blomquist 2005;van der Goes van Naters and Carlson 2007). Decreases in expression of genes inferred to have lipase activity could provide a mechanistic basis for the altered CHC profiles.
The promoter of Obp56h expresses lacZ in approximately five sensilla on each third antennal segment, in the pharyngeal organs and in the dorsal organ, the terminal organ, and the ventral pits of the third instar larvae (Galindo and Smith 2001). This Obp, therefore, may function in both olfactory and gustatory systems. It is of interest that expression of Or19b is down-regulated in Obp56h-RNAi male heads, especially since Or19b is expressed in trichoid sensilla (Couto et al. 2005), which appear specialized for the detection of pheromones (Ha and Smith 2006;Ronderos and Smith 2010). Or19b is thus a plausible candidate receptor for 5-T or another unknown Obp56h ligand.

ACKNOWLEDGMENTS
This work was supported by National Institutes of Health (NIH) grants R01 GM076083 and R01 GM059469 to T.F.C.M. and R.R.H.A. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.