Sexually dimorphic mechanosensory neurons regulate copulation duration and persistence in male Drosophila


 Peripheral sensory neurons are the gateway to the environment across species. In Drosophila, olfactory and gustatory senses are required to initiate courtship, as well as for the escalation of courtship patterns that lead to copulation. To be successful, copulation must last long enough to ensure the transfer of sperm and seminal fluid that ultimately leads to fertilization. The fly genitalia contain sex-specific bristle hairs innervated by mechanosensory neurons. To date, the role of the sensory information collected by these peripheral neurons in male copulatory behavior is unknown. Here, we employed genetic manipulations that allow driving gene expression in the male genitalia as a tool to uncover the role of these genitalia specific neurons in copulation. We found that the sensory information received by the mechanosensory neurons (MSNs) at the male genitalia plays a key role in copulation duration. We confirmed that these MSNs are cholinergic and co-express both fru and dsx. Moreover, our results show that the function of these fru/dsx cholinergic MSNs is required for copulation persistence, which ensures copulation is undisrupted in the presence of environmental stress before sperm transfer is complete.


Introduction
An animal's decision to perform a certain behavior requires the integration of external stimuli with homeostatic regulation. The nervous system collects environmental sensory information and integrates it with the internal status of the animal to generate motor signals that elicit an appropriate behavioral response. Drosophila male courtship is an excellent model to investigate how complex behavior is coordinated by the nervous system. The development of the courtship neural circuit is largely controlled by two sex determination genes, fruitless (fru) and doublesex (dsx) [1][2][3] . The ~2000 fru-expressing neurons, comprising of sensory, integration, and motor neurons, control most aspects of male courtship behavior [4][5][6][7][8][9] . However, the complete courtship neural circuit also requires dsx-expressing neurons, a subset of which co-expresses fru [10][11][12][13][14] .
Successful courtship requires the processing of sensory information received by peripheral sensory neurons 15 . In the antenna, fru olfactory receptor neurons (ORNs) receive and transmit volatile olfactory signals to the antennal lobe in the brain; fru ORNs expressing the receptor Or67d respond to the male pheromone cVA which inhibits mating behavior in males but promotes mating behaviors in females 16 .
Gustatory receptor neurons (GRNs) in the front legs receive and transmit non-volatile gustatory signals to the CNS. For example, fru GRNs expressing the ion channels ppk23 and ppk29 are required to detect inhibitory signals on males and excitatory signals on females during courtship 17 . Dsx neurons expressing the receptor Gr68a respond to the anti-aphrodisiac pheromone CH503 and inhibit male courtship 18 .
Acoustic information from the movement of the female detected by the Johnston's organ neurons (JONs) helps to promote courtship initiation in males 19 . Finally, the visual contribution to the initiation and maintenance of courtship has also been shown in a treadmill-based courtship assay 20 .
When the female is receptive to male courtship, she will slow down and spread her wings, allowing the male to mount. The male initiates copulation by bending his abdomen forward to attach the genitalia.
The male terminalia is divided in two regions: the periphallic structures and the intromittent organ. The periphallic structures form weaker connections to the intromittent organ and can be easily separated by dissection. The periphalilic structures are subdivided into four regions: the epandrial ventral lobe (formerly called: lateral plate) (EVL), the surstylus (formerly called: clasper) (SUR), the epandrial posterior lobe (formerly called posterior lobe) (EPL), and the cercus (formerly called: anal plate) (CER) 21 . Genital coupling, as revealed by a high-resolution electron microscopic time sequence analysis, involves the active movement of these periphalic structures 22 . The surstylus bends medially and is hidden from view.
Within 10 minutes of copulation, the cercus aligns with the female oviscape and achieves genital coupling. Each periphallic region contains an array of stereotypic, species and sex-speci c bristles, and each bristle is innervated by a bipolar mechanosensory neuron 23 . The neural implication of this sensory information collected during copulation is unclear.
Copulation must last long enough to ensure successful sperm transfer, which does not occur until ve minutes after copulation starts 24 . The regulation of copulation time in Drosophila is a complicated process that does not simply depend on the volume or the transfer of sperm and seminal uid, since mutants defective in their synthesis still have normal copulation duration 24 . Both neuronal and nonneuronal factors can in uence mating time. Neuronal regulators include a small cohort of fru neurons that co-express the transcription factor engrail 25 and a subset of 4-5 fru neurons that innervate the male reproductive tissues 26,27 . The latter neurons control copulation duration in response to sperm transfer 26 .
Non-neuronal factors include mutations in the circadian clock gene period, which exhibit longer copulation duration 28 , and environmental factors such as the gut microbiome 29 and environmental stressors 24 . Copulation persistence describes the maintenance of copulation in the presence of stressful stimuli, which peaks within the rst 5 minutes of copulation and decreases over time 24 . This is an important cost-bene t analysis that ensures species survival. The neural circuit for this cost-bene t analysis includes 8 dsx/GABA neurons in the ventral cord that decrease copulation persistence. With opposite action, ventral cord DA neurons increase copulation persistence 24 . However, the neural input required to initiate copulation persistence is unknown. In this study, we show that sensory information received by the mechanosensory neurons (MSNs) at the male genitalia is a novel neuronal regulator of copulation duration and is critical for the maintenance of copulation in the presence of stressful stimuli before sperm transfer is complete.

Results
Discovery of fruitless (fru) neurons that regulate copulation duration To uncover the neuronal regulators of copulation duration, we utilized an intersectional genetic approach, involving both the GAL4/UAS 30 and FLP/FRT 7 expression systems. The details of the intersectional genetic system and how we generated a FLP enhancer trap screen had been described elsewhere 27,31,32 . Brie y, the target gene is downstream of a DNA sequence that contains the binding site (UAS) for the transcription factor GAL4, followed by a DNA sequence that contains a stop codon anked by the FLP recombinase recognition site FRT (UAS-FRT-stop-FRT-target gene). A cell must possess an active promoter for FLP (to remove the stop codon by recombination) and an active promoter for GAL4 (to bind to UAS) to activate expression of the target gene. As an additional tool to restrict gene expression, tsh GAL80 is used to inhibit GAL4 expression speci cally in the ventral cord. From here forward, the genetic nomenclature will list all transgenes that are responsible to drive expression of the target gene (e.g. "X GAL4 ,FLP # >target gene" denotes a y that carries a GAL4 driver controlled by the promoter of X, and the FLP line # from the enhancer trap screen, expressing the target gene). For the relevant neuronal populations, we expressed either GFP (UAS-FRT-stop-FRT-GFP) to visualize the morphology or the neuronal silencer tetanus toxin (UAS-FRT-stop-FRT-TNT) to block neuronal activity. Previously, we identi ed a cluster of ~5 fru 5HT/DA neurons (fru-sAbg-1) in the male abdominal ganglion that innervate various male reproductive tissues and regulates copulation duration 27 . In that study, we observed a reduction in copulation duration when we combined FLP335 with either fru-GAL, that targets most fruitless neurons (Fru GAL4 ,FLP 335 ,tsh GAL80 >TNT), with Th-GAL4, that targets dopaminergic neurons (Th GAL4 ,FLP 335 ,tsh GAL80 >TNT), or with 5HT-GAL4, that targets serotonergic neurons (5HT GAL4 ,FLP 335 ,tsh GAL80 >TNT) 27 . However, neither silencing the subsets of dopaminergic nor serotonergic neurons can recapitulate the copulation phenotype observed when we targeted the broader fruitless circuit. The result suggested that additional fru neurons contributed to the copulation duration phenotype 27 . Indeed, the expression pattern for Fru Gal4 ,FLP 335 ,tsh GAL80 >GFP males extended beyond the accessory glands and seminal vesicles all the way to the genitalia ( Figure 1A). Upon further examination, we were able to trace GFP expression to the male genitalia.
To further con rm the genitalia neurons are implicated in copulation duration control, we screened our enhancer trap FLP library (ref?) to search for other FLP lines that affected copulatory behaviors. We identi ed another FLP line that showed expression at the genitalia and also reduced copulation duration when it is combined with fru-GAL4 to express TNT in males (Fru GAL4 ,FLP 386 ,tsh GAL80 >TNT) ( Figure 1D).
The experimental males (Fru Gal4 ,FLP 386 ,tsh GAL80 >GFP) showed similar expression in fru ORNs, JONs, and GRNs compared to Fru GAL4 ,FLP 335 ,tsh GAL80 >GFP males ( Figure 1B). However, the prominent fru-sAbg-1 neurons in the abdominal ganglion that innervate the male reproductive tissue were noticeably missing, leaving only the sexually dimorphic arborizations (ANN1) in the abdominal ganglion originating from the MSNs in the genitalia. The expression pattern in the peripheral tissues that extend out from the ventral cord of Fru GAL4 ,FLP 386 ,tsh GAL80 >GFP males was highly restricted, showing GFP expression only in the male genitalia ( Figure 1B). Although we still observed expression for fru GRNs, it is less consistent compared to Fru GAL4 ,FLP 335 ,tsh GAL80 >GFP. In summary, three common fru neurons are consistently labeled by both FLP lines: 1) the sexually dimorphic projections in the glomeruli from fru ORNs (DA1, VA1v, VL2a), 2) projections in the antennal mechanosensory motor complex (AMMC) from fru JONs, and 3) projections in the abdominal ganglion (ANN1) from fru MSNs in the genitalia ( Figure 1C, Table 1). The copulation duration of Fru GAL4 ,FLP 386 , tsh GAL80 >TNT was shortened by ~26% compared to the TNT inactive control ( Figure 1D). This difference is comparable to that observed in Fru GAL4 ,FLP 335 ,tsh GAL80 >TNT males. Of the three groups of fru neurons targeted by both FLP lines, we hypothesized that ANN1 from fru MSNs in the genitalia were the most likely to contribute to the copulation duration phenotype since we showed previously that silencing fru ORNs and GRNs did not affect copulation duration 27 . Recently, it was discovered that females sing in copula and that this song in uences the reproductive success of the male 33 . However, the study showed that copulation duration was unaffected by the absence of female singing 33 .

Characterization of fru genitalia mechanosensory neurons (MSNs)
We further characterized the expression pattern in the genitalia terminals using various genetic combinations. The complete fru circuit (fru>GFP) in males show consistent GFP expression in a subset of mechanosensory neurons in the epandrial ventral lobe (10±4, ~50% of EVL bristles), the surstylus (16±2, 64% of SUR bristles), and the cercus (20±2, ~61% of CER bristles) ( Figure 2B, Table 2). Adding FLP 335 or FLP 386 eliminated GFP expression in the epandrial ventral lobe and restricted expression in fru MSNs at the surstylus and the cercus ( Figure 2C, Table 2). Amongst these two lines, Fru GAL4 ,FLP 386 ,tsh GAL80 >GFP revealed a more restrictive expression pattern with 8±3 fru neurons at the surstylus and 4±2 at the cercus ( Table 2). As Fru GAL4 ,FLP 386 ,tsh GAL80 >TNT males resulted in the same copulation duration phenotype, we hypothesized that these 12 neurons (fru-MSNs, Table 2) at the genitalia are responsible for the shortened copulation duration phenotype. Since most sensory neurons in Drosophila are cholinergic, we hypothesized that fru-MSNs are the same. Indeed, using Cha-GAL80 that expresses the GAL4 inhibitor in cholinergic neurons eliminated ANN1 expression that originates from the genitalia MSNs ( Figure 2D).

Fru MSN neurons co-express doublesex (dsx)
To rule out the contributions of other neuronal populations to the copulation duration phenotype, we investigated other genetic combinations that target genitalia neurons more speci cally. Previous research indicated that the genitalia neurons express dsx 11,12,34 . Indeed, we con rmed that the fru-MSN neurons are also dsx positive ( Figure 3A). Replacing fru-GAL4 with dsx-GAL4 in our genetic combination with FLP 335 (Dsx GAL4 ,FLP 335 >GFP) eliminated all expression in the brain ( Figure 3C). In the VNC, Dsx GAL4 ,FLP 335 >GFP showed consistent expression only in the sexually dimorphic arbors originating from the foreleg GRNs and the genitalia neurons ( Figure 3B, D, E). Using this highly restrictive genetic combination, we silenced these neurons by TNT expression and evaluated the post-copulatory behaviors.
Consistent with our hypothesis, copulation duration was still reduced at the same level for Dsx GAL4 ,FLP 335 >TNT compared to the control ( Figure 3G). These results ruled out the involvement of CNS neurons in the copulation duration phenotype and indicate that that the fru neurons responsible for the phenotype are also dsx positive.
The characteristic male-speci c midline crossing of the LAN1 arbors that originated from GRNs on the front legs is regulated by both fru and dsx 35 . These GRNs co-express the ion channel ppk25 that is critical for pheromone detection during early courtship steps 17 . To investigate whether the copulation duration phenotype requires female pheromone perception, we paired Dsx GAL4 ,FLP 335 >TNT males with females whose pheromone-producing cells were ablated genetically via expression of the pro-apoptotic gene hid under an oenocyte-speci c promoter 36 . As expected, the experimental males still exhibited shortened copulation duration compared to the control males, irrespective of whether they were paired with oenocyte-less females (oe-) or their genetic controls (oe+) ( Figure 3H). The copulation duration of the control males (TNT-in) is signi cantly shorter when paired with oe-females compared to control oe+ females ( Figure 3H). However, no signi cant effect was observed in the experimental males (TNT); the lack of signi cance could be due to the variability of the data from the TNT/oe+ pairs. Taken together, these data suggest that the copulation duration phenotype is driven by the silencing of the fru/dsx-MSNs.
Sensory information detected by fru/dsx neurons is required for copulation persistence Next, we asked the biological relevance of a shortened copulation duration as a result of silencing fru/dsx MSNs. In a productive copulation pairing, duration must be long enough for the transfer of sperm and accessory gland uid. In the three genetic combinations presented here, the median copulation duration for the experimental males is 10-14 minutes, which is longer than the minimal time (~8 minutes) necessary for sperm transfer 25,26 . To evaluate if sperm is successfully transferred from the test male to the target female, we quanti ed the number of copulation pairings that resulted in fertilization after a single copulation event between a test male and a virgin Canton-S female. Compared to inactive TNT controls, a signi cantly lower percentage of males where both copulation duration regulating neural clusters (sAbg-1 and MSNs) are silenced (Fru GAL4 ,FLP 335 ,tsh GAL80 >TNT males) could fertilize the virgin females. In contrast, males in which sAbg-1 are not manipulated but MSNs are silenced (Fru GAL4 ,FLP 386 ,tsh GAL80 >TNT and Dsx GAL4 ,FLP 335 >TNT males) have relatively normal levels of fertilization rates compared to their respective inactive TNT controls ( Figure 4A). Therefore, post-copulatory fertility is unaffected by the shortened copulation duration due to the silencing of the fru/dsx-MSNs.
If fertility is unaffected, we wondered if peak copulation persistence is still possible without sensory input from the genitalia. To evaluate if copulation persistence is affected by the silencing of the fru/dsx-MSNs, we applied heat shock as the stress stimulus to copulating ies when copulation persistence is at its peak. We heat-shocked pairs of ies 5 minutes after the onset of copulation and quanti ed the number of pairs that disengaged. While a large number of pairs in which the fru/dsx-MSNs in the male genitalia were silenced by our three genetic combinations (Fru GAL4 ,FLP 335 ,tsh GAL80 >TNT; Fru GAL4 ,FLP 386 ,tsh GAL80 >TNT; Dsx GAL4 ,FLP 335 >TNT) terminated copulation, all pairs with the corresponding control males remained copulating ( Figure 4B). These results indicate that the fru/dsx-MSNs are important in maintaining copulation persistence in response to environmental stress before sperm transfer is complete.

Discussion
Different neuronal populations have been identi ed to control three distinct aspects of male copulatory behaviors: genital coupling, copulation duration, and sperm transfer 26,27,34 . Here, we have characterized the fru/dsx-MSNs at the genitalia and identi ed their novel function in regulating copulation duration and maintaining copulation persistence in the presence of environmental stressors. Although we cannot completely rule out the contribution of pheromone perception in their regulation, the pheromone sensing neurons are not consistently targeted in one of our genetic combination (Fru GAL4 ,FLP 386 ,Tsh GAL80 >GFP) ( Figure 1B).
Retrograde labeling of various genitalia MSNs have shown that most of these neurons project only to the abdominal ganglion except for one neuron from the surstylus (clasper) that projects all the way to the suboesophageal region of the brain 34 . Since we did not observe any arborizations in the brain, the copulation phenotype appears to be regulated by a subset of the fru/dsx − MSNs at the surstylus (claspers) and epandrial ventral lobe (lateral plate) that project speci cally to the abdominal ganglion. In addition, the retrograde labeling experiment showed that the axonal terminals of the genitalia neurons juxtapose the dendrites of the abdominal dsx glutamatergic (dsx/vGlut-Abg) and GABAergic (dsx/GABA-Abg) neurons 34 . Moreover, arti cial mechanical stimulation of the genitalia with a minuten pin activated both dsx/vGlut-Abg and dsx/GABA-Abg neurons 34 . Therefore, we can infer that fru/dsx-MSNs neurons make functional synaptic connections to both dsx/vGlut and dsx/GABA neurons in the abdominal ganglion ( Figure 5). The shortened copulation duration phenotype when fru/dsx-MSNs are silenced is a result of dsx/vGlut-Abg and/or dsx/GABA-Abg not receiving sensory feedback signals from the genitalia. dsx/vGlut-Abg are all motor neurons that innervate the phallic and periphallic musculature responsible for genital attachment 34 . On the other hand, dsx/GABA-Abg is comprised of a heterogeneous population of interneurons with different functions. While some of these interneurons inhibit the activity of dsx/vGlut-Abg to terminate copulation 34 , others have different functional roles, such as regulating copulation persistence 24 . Therefore, sensory signals received from the fru/dsx-MSNs can in uence the relative activity level of dsx/vGlut-Abg and dsx/GABA-Abg neurons with opposing actions and modulate the overall tension of the copulatory muscles. In the absence of sensory information, the baseline muscle strength is su cient to maintain genital coupling long enough for sperm transfer since males with fru/dsx-MSNs silenced has normal fertility ( Figure 4A). However, our results show that the sensory information encoding the male's correct engagement in copulation is necessary to achieve peak copulation persistence before sperm is transferred ( Figure 4B, 5). Sensory information -provided by fru/dsx-MSNs -might be a way to measure the quality of the copulation. Genital coupling will lead to bending of the bristle hairs and activate the fru/dsx-MSNs. Which bristle hair gets stimulated and the strength of the stimulation will depend on whether genital coupling is established and the morphology of the female genitalia. For example, an abnormal amount of pressure received by a bristle hair could send less signal to activate the copulatory muscle innervated by dsx-vGlut-Abg. Similarly, a wrong set of bristles bent could activate the dsx-GABA-Abg neurons that inhibit the dsx-vGlut-Abg. A suboptimal activation decreases the total muscular tension during genital coupling. Copulation persistence is a result of the y's ability to analyze the tradeoff of maintaining copulation during exposure to an environmental stressor. Sensory information that signals the quality of the copulation provides a critical factor in this assessment. The recently published transcriptomic atlas for both the ventral nerve cord 37 and the male genitalia 38 provides new possibilities to develop speci c neural markers that will allow further investigation of the neural mechanism of how sensory information from genital coupling is encoded and processed in the CNS.

Fly strains
The following strains were used in this study: fru-GAL4 7 , dsx-GAL4 11 , UAS>stop>TNTin, UAS>stop>TNT 7 , UAS>stop>mCD8::GFP 9 , FLP 335 and FLP 386 were generated as described in previous study 27 , cha-GAL80 39 , tsh-GAL80 from Julie Simpson, and Canton-S strain from the Bloomington Stock Center, Bloomington, IN. Oenocyte less (oe-) ies and their controls were generated as previously described 36 . Immunohistochemistry Dissection of 3-7 day old adult ies and immunohistochemistry of the adult nervous system were carried out as described previously 40 with some modi cations in primary and secondary antibodies. Dissection of male reproductive organs was performed on Sylgard plate covered with 1xPBS. To obtain male genitalia, the lower half of the abdomen were dissected and xed in 4% PFA at room temperature for 20-30 mins. After which the PFA was replaced with 1xPBS. The genitalia were cut out with a pair of microscissors to ensure a at surface for mounting. Dissected samples were xed and proceeded with immunostaining as described previously 40 . The following primary and secondary antibodies were used in this study: rat polyclonal anti-mCD8 (1:100; Caltag, Burlingame, CA), rabbit anti-GFP (1:600; Invitrogen), mouse anti-GFP (1:500, Life technologies), mouse anti-nc82 (1:20; Hybridoma Bank) 41 , rabbit anti-Dsx M (1:2000) (kindly supplied by Brian Oliver laboratory, NIH), anti-rat IgG conjugated with Alexa Fluor 488 (1:300; Invitrogen), and anti-rabbit IgG conjugated with Alexa Fluor 594 (1:300; Invitrogen).

Microscopy
Images of the ventral cord, the reproductive tissues, and genitalia were acquired using an Olympus Fluoview FV1000 confocal microscope with a 20X objective. ImageJ was used to stitch the overlapping images together. All other images were acquired using a Zeiss LSM 700 confocal microscope. For the genitalia, lambda stacks from 490-600 nm at 20 nm intervals were acquired. The auto uorescence signals were unmixed from the 488 nm signals using the linear unmixing algorithm in the Zen Black software.
Behavioral assays Husbandry Flies were raised on standard cornmeal medium and kept on a 12hr:12hr day:night cycle at 25°C in ambient relative humidity. Each newly eclosed adult was collected and aged for 3-7 days in an isolation vial (16 x 100-mm) supplied with ~2 ml of y food. Virgin, wildtype Canton-S females were aged in groups of 20-40 for 3-7 days. All behavioral experiments were performed at 25°C with ~50% humidity during the rst 3 hrs. after lights on.

Copulation Assay
Copulation assays were performed in 12-well plates (Thermo Scienti c BioLite Multidish). A square glass plate covering four wells was used as the lid. Assays were performed in 4 wells with ~5 ml standard y food to maintain humidity in each well. An experimental male was paired with a virgin female and the courtship behavior was videotaped for at least an hour. If a pair did not copulate within an hour, it was considered unsuccessful. Copulation duration was calculated from the beginning of genital coupling until the male was dismounted from the female.

Fertility Assay
Freshly hatched males were isolated into a glass tube (16x100mm) with 2 ml of food and kept for 4-5 days at standard conditions (25°C, 50% RH). A 6-7 day old CS virgin female was introduced into each tube using an aspirator. After transferring the vials back to standard conditions, the ies were allowed to interact for 1h and observed every 10 mins for successful copulation. Tubes with pairs that did not copulate were discarded. For the rest of the tubes, males were removed by quick anaesthetization and females were allowed to lay eggs for 24h and then discarded. Fertility was recorded after 7 days by observing the presence of progeny.     (Table 2). C) fru

Persistence Assay
MSNs restricted by FLP335. Expression pro le of fru-GAL4, UAS>stop>mCD8::GFP, FLP335, tsh-GAL80 in the male genitalia. Adding FLP335 eliminated GFP expression in the epandrial ventral lobe and restricted expression at the surstylus and the cercus ( Table 2). The auto uorescence and 488 signals were unmixed as described in Methods. D) fru MSNs are cholinergic. Expression pro le of FruGAL4,FLP335,ChaGAL80>GFP in the male ventral nerve cord. Adding the transgene cha-Gal80, which inhibits GAL4 in cholinergic neurons, eliminated ANN1 expression that originated from the genitalia MSNs (compare to Figure 1A  sexually dimorphic LAN1 and ANN1 arbors originated from the GRNs on the front legs and MSNs of the genitalia were the only common expression observed in both GAL4 lines. C-F) Expression pro le of DsxGAL4,FLP335>GFP. The CNS tissues were stained with anti-mCD8 (green) and anti-nc82 (magenta).
No expression was observed in the brain (C), and the ventral nerve cord showed expression for the sexually dimorphic LAN1 arbors that originated from the GRNs in the front legs and the sexually dimorphic ANN1 arbors that originated from the MSNs in the genitalia (D  Proposed model of how sensory information from the MSNs is incorporated into the neural circuit that governs copulation persistence. Genital coupling requires the activation of the dsx/vGlut motor neurons that innervate the copulatory muscles. The resulting muscular tension is proportional to the activity level of the dsx/vGlut motor neurons, which are antagonistically regulated by dsx/GABA interneurons. These copulation regulating neurons are functionally connected to the axonal terminals of sensory neurons Page 23/23 from the genitalia bristles34. We propose that a suboptimal sensory input will reduce the total muscular tension due to less activation of the dsx/vGlut motor neurons or more activation of the dsx/GABA interneurons. Copulation persistence ensures the maintenance of copulation before sperm transfer in the presence of stressful stimulus and is regulated by opposing actions of 8 dsx/GABA and the dopaminergic system in the ventral cord24. This persistence requires optimal sensory input from the genitalia bristles.