Due to the complex behavioural and physiological changes induced by SP, neurons in the central nervous system have been suspected as main targets for SP (Kubli, 1992). To express mSP only in the head, we used an elav FRTstopFRT GAL4 in combination with otdflp that expresses in the head to drive recombination and head-specific expression of mSP from UAS (Figure 1A, Figure 1—figure supplement 1A–F; Haussmann et al., 2008; Asahina et al., 2014; Zaharieva et al., 2015; Nallasivan et al., 2021). To express mSP predominantly in the trunk, we used tshGAL4 (Figure 1B, Figure 1—figure supplement 1G–L; Soller et al., 2006).

When we expressed mSP in the head, females reduced receptivity indistinguishable from mated females, but did not lay eggs, thereby again demonstrating that the two main PMRs can be separated (Figure 1C and D; Haussmann et al., 2013). In contrast, when we expressed mSP in the trunk, females remained receptive but laid eggs in numbers indistinguishable from mated females (Figure 1C and D).

Moreover, tshGAL4 is expressed in fru, dsx, ppk genital tract sensory neurons (Figure 1E–H). Since mSP expression with tshGAL4 does not affect receptivity, these genital tract neurons unlikely are direct targets for SP (Haussmann et al., 2013). Taken together, these results indicate the presence of SP target neurons in the brain and ventral nerve cord (VNC) for the reduction of receptivity and induction of egg laying, respectively.

Expression of mSP from UAS via GAL4 inserts in fru and dsx genes induces a robust reduction in receptivity and increase in egg laying (Rezával et al., 2012; Haussmann et al., 2013). To identify SP target neurons, we thought to dissect the broad expression pattern of complex SPR, fru, and dsx genes spanning 50–80 kb by identifying regulatory DNA fragments in the enhancer regions that drive UAS mSP in a subset of neurons. For these experiments, we analysed 22, 27, and 25 GAL4 lines from the VDRC and Janelia tiling GAL4 projects (Pfeiffer et al., 2008; Jenett et al., 2012; Kvon et al., 2014; Figure 2A–C).

Strikingly, in SPR, fru, and dsx genes, we identified only one regulatory region in each gene (SPR8, fru11/12, and dsx24) that reduced receptivity and induced egg laying through GAL4 UAS expression of mSP (Figure 2D and E). In addition, we identified one line (SPR12) in the SPR gene that induced egg laying but did not reduce receptivity, consistent with previous results that SP regulation of receptivity and egg laying can be split (Haussmann et al., 2013).

All of these lines expressed in subsets of neurons in the central brain and the VNC in distinct, but reduced patterns compared to the expression of the SPR, fru, and dsx genes (Yapici et al., 2008; Rideout et al., 2010; Zhou et al., 2014; Figure 2F–O). Moreover, these lines showed prominent labelling of abdominal ganglion neurons in the VNC (Figure 2K–O). In addition, all of these lines except SPR12 are also expressed in genital tract sensory neurons (Figure 2—figure supplement 1A–E).

From all the 74 lines that we have analysed for PMRs from SPR, fru, and dsx genes, we also analysed expression in genital tract sensory neurons as they had been postulated to be the primary targets of SP (Yapici et al., 2008; Häsemeyer et al., 2009; Yang et al., 2009; Rezával et al., 2012). Apart from PMR-inducing lines SPR8, fru11, fru12, and dsx24, that showed expression in genital tract sensory neurons, we identified three lines (SPR3, SPR 21, and fru9), which also robustly expressed in genital tract sensory neurons but did not induce PMRs from expression of mSP (Figure 3, Figure 2—figure supplement 2).

Since genital tract sensory neurons have been postulated to induce the SP response, we tested previously identified split-GAL4 (SPSN-1: VT058873 ∩ VT003280/FD6 and SPSN-2: VT58873 ∩ VT033490) lines, which upon neuronal inhibition reduced receptivity and induced egg laying (Feng et al., 2014), for their capacity to induce the SP response upon expression of mSP. Both lines reduced receptivity and induced egg laying upon expression of mSP (Figure 3A and B). However, since genital tract neuron expressing SPR3 and SPR21 lines did not induce PMRs upon expression of mSP, the SP response induced by SPSN1 and 2 split-GAL4 mSP expression could originate from other neurons.

Expression analysis of these two lines revealed that in addition to expression in genital tract sensory neurons (Figure 3C, D, G, and H), they also showed expression in the brain and VNC (Figure 3K, L, O, and P). Intriguingly, the brain neurons labelled in SPSN-1 resembled the neurons identified by SPR8 ∩ FD6 (Figure 4G).

A screen aiming to identify neurons involved in the control of receptivity and egg laying by expression of the rectifying potassium channel Kir2.1 identified six enhancer GAL4 driver lines (FD1-6) (Feng et al., 2014). FD1-6 are expressed in diverse subsets of neurons in the brain and the VNC; in particular, they show common expression in the abdominal ganglion with projections to the central brain. The lines expressing in FD1-5 neurons have been termed SAG (secondary ascending abdominal ganglion neurons) neurons that are also interconnected with MIP sensing neurons (Jang et al., 2017). Since enhancer lines identified in SPR, fru, and dsx genes are prominently expressed in the abdominal ganglion, we tested whether mSP expression from these FD1-6 lines induced PMRs.

From these six lines, one robustly suppressed receptivity and induced egg laying (FD6/VT003280), while two lines only induced egg laying (FD3/VT4515 and FD4/V000454) similar to controls from mSP expression (Figure 4). Again, all three lines also expressed in subsets of neurons in the central brain and VNC, particularly in the abdominal ganglion (Zhou et al., 2014). In addition, FD3 and FD4 did not express in genital tract sensory neurons, in contrast to FD6 (Feng et al., 2014). A SAG split-GAL4 (VT050405/FD1 AD and VT007068/FD2 DBD) line did not show a response to expression of mSP and virgin females, for example, they mated and did not lay eggs (Figure 4).

To further restrict the expression to fewer neurons, we intersected the expression patterns of those lines that induced robust reduction of receptivity and increase of egg laying using split-GAL4 (SPR8, fru11/12, dsx, and FD6; for further experiments we used dsxGAL4-DBD, because dsx24 is less robust and fru11 and fru12 were made into one fragment) that activates the UAS reporter when GAL4 is reconstituted via dimerisation of activation (AD-GAL4) and DNA binding (GAL4-DBD) domains (Luan et al., 2006; Figure 5A).

Again, the intersection of SPR8 with fru11/12, dsx or FD6, and fru11/12 with dsx or FD6 expression robustly reduced receptivity and increased egg laying upon expression of mSP (Figure 5B and C). Accordingly, we termed these SP response-inducing neurons SPRINz, though the exact identity in the split-GAL4 intersection population needs to be determined.

When we further analysed the expression of these split-GAL4 intersections in the brain, we found that each combination first showed very restricted expression, but second, that none of these combinations labelled the same neurons (Figure 5D–H). For dsx neurons, split-GAL4 intersections correspond to a subset of dPC2l (SPR8 ∩ dsx) and dPCd-2 (fru11/12 ∩ dsx) neurons (Deutsch et al., 2020; Schretter et al., 2020; Nojima et al., 2021). These results suggest the SP targets interneurons in the brain that feed into higher processing centres from different entry points likely representing different sensory input.

In the VNC, we found expression in the abdominal ganglion with all split-GAL4 combinations (Figure 5I–M). In particular, the intersection of dsx with SPR8 or fru11/12 showed exclusive expression in the abdominal ganglion, while the other combinations also expressed in other cells of the VNC. Altogether, these data suggest that the abdominal ganglion harbours several distinct types of neurons involved in directing PMRs (Oliveira-Ferreira et al., 2023).

In the female genital tract, these split-Gal4 combinations show expression in genital tract neurons with innervations running along oviduct and uterine walls (Figure 5—figure supplement 1A–J). In addition, SPR8 ∩ fru11/12 and SPR8 ∩ dsx were also expressed in the spermathecae (Figure 5—figure supplement 1A and B).

Next, we tested whether PMRs induced by mSP expression in the SPR8 ∩ dsx, fru11/12 ∩ dsx or SPR8 ∩ fru11/12 rely on SPR. Expression of mSP in dsx ∩ SPR8 and dsx ∩ fru11/12 neurons in SPR mutant females did not reduce receptivity or induce egg laying (Figure 6A and B, see also Figure 5A and B), while a partial response was observed for SPR8 ∩ fru 11/12 induced mSP expression in SPR mutant females, which is consistent with presence of additional receptors for SP (Haussmann et al., 2013).

Since SP is transferred during mating to females and enters the hemolymph (Haussmann et al., 2013), we wanted to test whether SPR is required in these neurons for inducing PMRs after mating. For SPR RNAi in dsx ∩ fru11/12 and SPR8 ∩ fru 11/12 neurons, no reduction, or a partial reduction, of receptivity was observed, respectively, while SPR RNAi in dsx ∩ SPR8 neurons turned virgin females unreceptive (Figure 6A). Expression of mSP in dsx ∩ fru11/12 neurons in the context of SPR RNAi partially reduced receptivity, again suggesting additional receptors for SP (Haussmann et al., 2013).

Strikingly, however, SPR RNAi in these neurons prevented egg laying independent of whether SP was delivered by mating or when tethered to the membrane of these neurons (Figure 6B).

These results demonstrate that neurons identified by split-GAL4 intersected expression of SPR8 with dsx or fru11/12, or fru11/12 with dsx are genuine SP targets as they rely on SPR and PMRs are induced by SP delivered through mating.

The analysis of ppkGAL4 neurons in SP-insensitive Nup54 alleles revealed a hierarchy of trunk neurons that dominate over central brain neurons (Nallasivan et al., 2021). To focus on the role of central brain neurons, we generated a UAS mSP line with a stop cassette (UAS FRTstopFRT mSP) that allows us to restrict expression of mSP to the head in the presence of otdflp, which only expresses in the head (Figure 7A), but not in the trunk (Asahina et al., 2014; Nallasivan et al., 2021).

In combination with the intersectional approach, we now can restrict mSP expression to few central brain neurons, or alternatively activate or silence these neurons (Figure 7B). Expression of mSP in SPR8 ∩ dsx, fru11/12 ∩ dsx, or SPR8 ∩ fru11/12 neurons in the central brain significantly reduced receptivity, but oviposition was only substantially induced in SPR8 ∩ dsx brain neurons (Figure 7C and D). In fru11/12 ∩ dsx or SPR8 ∩ fru11/12, PMR inducing neurons from the VNC could be required to potentiate the response.

These results clearly demonstrate a role for brain neurons in the SP response. However, we noticed that the flipase approach can result in false negatives as fruflp inserted in the same position in the endogenous locus as fruGAL4 does not induce a response with UAS FRTstopFRT mSP in contrast to fruGAL4-induced expression of mSP. In contrast, the same experiment with dsxGAL4 and dsxflp results in a positive SP response indistinguishable from mated females (Haussmann et al., 2013).

Next, we tested whether neuronal activation or inhibition would induce a PMR. Strikingly, conditional activation of SPR8 ∩ dsx, fru11/12 ∩ dsx, or SPR8 ∩ fru11/12 brain neurons with TrpA1 in adult females completely inhibited receptivity and induced egg laying comparable to mated females (Figure 7C and D). In contrast, inhibition of these neurons with tetanus toxin (TNT) did not alter the virgin state, for example, receptivity was not reduced and egg laying was not induced (Figure 7C and D).

When we analysed split-GAL4 combinations of SPSN (VT058873, the common line in the SPSN1 and 2 lines) with SPR8, fru11/12, and dsx, we observed full response to mSP expression for the intersection with SPR8 and fru11/12, and a partial response for the SPSN ∩ dsx intersection (Figure 8A and B). Intriguingly, all of these split-Gal4 combinations expressed in few neurons in the brain, the VNC and genital tract neurons, except for VT058873 ∩ fru11/12, which did not express in genital tract neurons (Figure 8C–N).

We then restricted expression of mSP and induction of neuronal activity to the head with these split-GAL4 combinations using FRTstop cassettes and otdflp. In this set-up, we can induce PMRs from mSP expression or neuronal activation from TrpA1 expression with the VY058873 ∩ SPR8 and dsx combination, but not with the fru11/12 combination (Figure 8O and P). For the VT058873 ∩ fru11/12 intersection, PMR inducing neurons likely reside in the VNC.

We then analysed co-expression of SPR, dsx, and fru with SPSN originating split-GAL4 enhancer lines from CG31637 (FD6), ocelliless (VT05573), and Gyc76C (VT033490) in the single-cell brain atlas (Li et al., 2022). CG31637 co-expressed in many cells with SPR and fru, but only a few cells with dsx (Figure 8—figure supplement 1A–C). Expression of ocelliless with SPR and fru is broad, while only one neuron expressed with and Gyc76C in the brain (Figure 8—figure supplement 1D, E, G, and H). Expression of ocelliless with dsx is restricted to two neurons, and no overlap was detected with Gyc76C in the brain (Figure 8—figure supplement 1F and I).

Expression of UASmSP using a GAL4 driven by a promoter fragment of the ppk gene can also induce PMRs (Figure 9A and B; Häsemeyer et al., 2009; Yang et al., 2009). The complement of neurons labelled with ppkGAL4 consists of at least two populations including prominently sensory neurons, but also eight interneurons in the central brain (Nallasivan et al., 2021). These brain neurons show severe developmental defects in SP-insensitive Nup54 mutant alleles, but they receive inhibitory input from sensory neurons (Nallasivan et al., 2021).

To evaluate whether ppkGAL4 neurons are part of the previously identified expression patterns, we intersected them by crossing GAL4-AD lines SPR8, SPR12 and fru11/12 and the pan-neural nSybAD with a ppk GAL4-DBD line containing the previously used 3 kb promoter fragment (Grueber et al., 2003; Seidner et al., 2015; Riabinina et al., 2019). Surprisingly, none of these split-GAL4 combinations reduced female receptivity or increased egg laying (Figure 9A and B).

Few GFP-expressing neurons were detected in the brain for the nSyb ∩ ppk and the fru11/12 ∩ ppk intersection (Figure 9C–F) or abdominal ganglion (Figure 9G–J). For the nSyb ∩ ppk and the SPR8 ∩ ppk intersection, we detected GFP expression in genital tract sensory neurons (Figure 9K, L, O, and P), but not for the other combinations (Figure 9M, N, Q, and R).

Inhibiting or activating neurons with these split-Gal4 combinations did not reduce receptivity or induce egg laying (Figure 9S–V). How exactly ppk neurons labelled with ppkGAL4 impact on PMRs, however, needs to be further evaluated in follow-up studies. Moreover, if genetical tract neurons were SP target sites, an SP response would have been expected for the nSyb ∩ ppk intersection, which we did not observe.

A number of additional split-GAL4 combinations with restricted expression have been identified that play a role in female reproductive behaviours (Wang et al., 2020a; Wang et al., 2020b; Wang et al., 2021). These lines express in a subset of dsx expressing neurons (pC1-SS1), in oviposition descending neurons (oviDN-SS1 and 2), in oviposition excitatory neurons (oviEN-SS1 and 2), in oviposition inhibitory neurons (oviIN-SS1 and 2), and in vaginal plate opening neurons (vpoDN-SS1, also termed ovipositor extrusion/rejection behaviour neurons, because Drosophila does not have vaginal plates like e.g. seen in Hemiptera; Aigaki et al., 1991; Soller et al., 2006). When we analysed these lines for a response to mSP expression, receptivity was reduced from mSP expression in oviEN-SS2, oviN-SS1, and vpoDN-SS1 neurons, but no egg laying was induced from mSP expression in any of these neurons (Figure 10A and B).

In genital tract neurons, OviDN-SS1s, OviEN-SS1, OviIN-SS1, and vpoDNs express, but OviDN-SS1s and OviEN-SS1 express weakly (Figure 10C and J).

Both inhibitory and activating neurons have been attributed to impact on PMRs (Kvitsiani and Dickson, 2006; Yapici et al., 2008; Rezával et al., 2012). These neurons seem to be part of intersecting circuitry as general inhibition of ppkGAL4 neurons by TNT only partially blocks the SP response in contrast to inhibition of ppkGAL4 neurons in the brain alone (Nallasivan et al., 2021).

When we inhibited neuronal activity by expression of TNT (Sweeney et al., 1995), we observed a significant reduction of receptivity for all split-Gal4 combinations, though only partially for inhibition in fru11/12 ∩ FD6 neurons. Likewise, all split-Gal4 combinations induced a significant increase in egg laying (Figure 11A and B). Ablation of these neurons by expression of apoptosis-inducing reaper and hid genes essentially replicated the results from neuronal inhibition indicating that SPR target neurons are modulatory and are not part of motor circuits because females laid eggs and performed normally in receptivity assays (Figure 11C and D).

To evaluate the composition of the intersected expression patterns into inhibitory and activating neurons, we also expressed the Bacillus halodurans sodium channel (NaChBac) (Feng et al., 2014) to activate all of the intersected neurons. Here, we found a significant reduction of receptivity for four of the five split-GAL4 combinations, though only partially for activation of SPR8 ∩ dsx neurons (Figure 11E). Activating fru11/12 ∩ FD6 neurons did not reduce receptivity (Figure 11E). Likewise, we found the same pattern for the induction of egg laying (Figure 11F). Four of the five split-GAL4 combinations induced a significant increase which was only partial in SPR8 ∩ dsx neurons, and no egg laying was induced by activating fru11/12 ∩ FD6 neurons.

Essentially, these results are consistent with previous findings that inhibitory neurons prevail (Nallasivan et al., 2021), possibly as input from trunk neurons as found for ppk expressing neurons.

With the split-GAL4 approach, we identified five distinct neuronal sub-types that can induce PMRs. To find out whether these neurons receive input from distinct entry points in the brain and to identify the target neurons of these mSP-responsive neurons, we used the retro- and trans-Tango technique to specifically activate reporter gene expression in up- and down-stream neurons (Talay et al., 2017; Sorkaç et al., 2023; Figure 12A–O).

In the brain, the retro-Tango analysis did not identify primary sensory neurons, but higher order neurons in the central brain in all five split-GAL4 combinations (Figure 12A–E). In addition, neurons in the suboesophageal ganglion were marked from SPR8 intersections with dsx and FD6, and in dsx ∩ fru11/12. In dsx ∩ fru11/12, neurons in the optic lobe (medulla) were marked. In addition, a strong signal was observed in all five split-GAL4 combinations in the mushroom bodies (Figure 12A–E). Although mushroom bodies are dispensable for PMRs (Fleischmann et al., 2001), their connection to SP target neurons indicates an experience-dependent component of PMRs.

The trans-Tango analysis identified a subset of neurons with cell bodies in the suboesophageal ganglion with projections to the pars intercerebralis for SPR8 ∩ dsx and fru11/12 ∩ dsx neurons (Figure 12K and L). For SPR8 ∩ fru11/12 and SPR8 ∩ FD6 neurons, common target neurons were found in the antennal mechanosensory and motor centre (AMMC) region with a single neuron identified near the mushroom body region (Figure 12M and N; Ishimoto and Kamikouchi, 2021). For fru11/12 ∩ FD6, no obvious targets were identified in the central brain (Figure 12O).

In the VNC, the trans-Tango analysis showed post-synaptic targets within the abdominal ganglion with all five split-GAL4 combinations indicating an interconnected neuronal network (Figure 12—figure supplement 1A–O), which needs to be elaborated in detail. In the genital tract, no post-synaptic targets were detected, indicating that these are afferent neurons integrating sensory input (Figure 12—figure supplement 1P–AD).

Taken together, circuitries identified via retro- and trans-Tango place SP target neurons at the interface of sensory processing interneurons connecting to two commonly shared post-synaptic processing neuronal populations in the brain. Hence, our data indicate that SP interferes with sensory input processing from multiple modalities that are canalised to higher order processing centres to generate a behavioural output.

Early characterisation of the SP signalling cascade demonstrated induction of PMRs from various other sources than mating, including transgenic secretion from the fat body, expression as membrane-tethered form on neurons or injection of synthetic peptide into the hemolymph (Chen et al., 1988; Aigaki et al., 1991; Schmidt et al., 1993; Nakayama et al., 1997). Likewise, SP is detected in the haemolymph after mating at a PMR inducing concentration (Haussmann et al., 2013). Moreover, PMRs are induced faster when SP is injected compared to induction by mating (Haussmann et al., 2013). This delay, however, is not attributed to sperm binding of SP as it is unchanged after mating with spermless males. These results suggest that SP reaches its targets through entering the circulatory system to target neurons and contrasts a previously proposed model favouring genital tract neurons as SP sensors from the lumen of the genital tract (Häsemeyer et al., 2009; Yang et al., 2009; Rezával et al., 2012). We previously observed binding of radiolabelled SP to various sites in the nervous system including afferent nerves, but these signals likely reflect binding to broadly expressed SPR rather than binding to SPRINz (Ottiger et al., 2000; Ding et al., 2003; Yapici et al., 2008; Haussmann et al., 2013).

In further support of the internalisation model, we identified GAL4 drivers that express mSP in genital tract neurons, but do not induce PMRs. Also, SPR12 does not express in genital tract neurons, but induces egg laying by expression of mSP. Moreover, expression of mSP predominantly in the trunk (including all genital tract sensory neurons) only induces egg laying, but does not change receptivity. Likewise, expression of mSP specifically in the brain (SPR8 ∩ dsx) can reduce receptivity and induce egg laying indistinguishable from mated females.

A ppkGAL4 line generated by P-element mediated transformation can induce PMRs by expression of UAS mSP (Grueber et al., 2003). The same promoter fragment fused to a GAL4 DBD and inserted by phiC31 integration into a landing site intersected with pan-neural nSyb AD line (Seidner et al., 2015; Riabinina et al., 2019), however, does not induce an SP response despite being expressed in genital tract neurons. We found that the ppkGAL4 expresses in a few neurons in the brain and VNC (Nallasivan et al., 2021), but this expression is absent in nSyb ∩ ppk intersection. Likely, the ppkGAL4 construct is inserted in a locus that contains an enhancer that drives expression in SP target neurons.

These results are in strong favour of SP entering the hemolymph to target neurons in the VNC for inducing egg laying, and in the central brain for reducing receptivity and inducing egg laying (Haussmann et al., 2013).

Reduction of receptivity and induction of egg laying are both induced by the same critical concentration of injected SP (Schmidt et al., 1993; Haussmann et al., 2013) initially suggesting a simple on/off system for PMRs likely initiated from a small population of neurons. However, such a model would not allow us to split the SP response into individual PMR components by expression of mSP.

Here, we identified several GAL4 drivers, which can induce only egg laying (SPR12, FD3, FD4, and tsh GAL4), but do not reduce receptivity, and others that can only reduce receptivity (oviEN-SS2, oviIN-SS1, and vpoDN-SS1), but do not induce egg laying. Strikingly, tshGAL4, which expresses predominantly in the trunk, only affects egg laying, suggesting a role for the abdominal ganglion in egg laying. Moreover, dsx and all of the SBRINz split-GAL4 combinations affect egg laying and express in the abdominal ganglion (Rezával et al., 2012; Zhou et al., 2014). Hence, this neuronal structure has a key role in regulating egg laying. Since more than a single neuronal population seems to direct egg laying, further high-resolution mapping is required to identify individual neuronal population within the abdominal ganglion (Jang et al., 2017; Oliveira-Ferreira et al., 2023).

Since tshGAL4 only induces egg laying, neurons in the brain must direct reduction of receptivity. Through intersectional expression in combination with head-specific expression of otdflp, we could express mSP only in the brain by FLP-mediated brain-specific excision of a stop cassette. We observed a significant reduction in receptivity for all five intersections tested, but for four, the response is only partial, likely due to the inefficiency of FLP-mediated recombination.

Moreover, brain neurons can also induce egg laying when SPR8 is intersected with dsx, and to some extent also from SPR8 intersection with fru11/12. Due to the inefficiency of FLP-mediated recombination, however, this is likely an underestimate and solving this issue requires development of more robust tools.

In any case, however, our results show that PMRs can be induced from mSP expression from several sites, suggesting interference with processing of sensory information at the level of interneurons. In particular, SPR8 ∩ fru11/12 neurons resemble auditory AMMC-B2 neurons involved in processing of information of the male love song (Yamada et al., 2018). Likewise, SPR8 ∩ dsx neurons seem to overlap with dimorphic dsx pCL2 interneurons that are part of the 26 neurons constituting the pC2 neuronal population involved in courtship song sensing, mating acceptance and ovipositor extrusion for rejection of courting males (Kimura et al., 2015; Deutsch et al., 2019; Wang et al., 2020a). The SPR8 ∩ FD6 neurons resemble dopaminergic fru P1 neurons involved in courtship and the fru11/12 ∩ dsx neurons seem to overlap with dsx pCd and neuropeptide F neurons involved in courtship (Zhang et al., 2021). In females, pC1d neurons have been linked to aggression (Deutsch et al., 2020; Schretter et al., 2020). The fru11/12 ∩ FD6 neurons resemble a class of gustatory pheromone sensing neurons (Sakurai et al., 2013). Although we likely have not identified all SP sensing neurons, our resources will provide a handle to future exploration of the details of this neuronal circuitry incorporating SP signalling for inducing PMRs.

We have identified distinct SP response-inducing neurons (SPRINz) in the central brain and the VNC. Since these five different SP response-inducing neuronal populations in the central brain converge into two target sites, our data suggest a model (Figure 12P), whereby SP signalling interferes with integration of sensory input. Independent interference with different sensory modalities opts for the female to counteract male manipulation at the level of perception of individual sensory cues to adapt to varying physiological and environmental conditions to maximise reproductive success.

Flies were kept on standard cornmeal-agar food (1% industrial-grade agar, 2.1% dried yeast, 8.6% dextrose, 9.7% cornmeal, and 0.25% Nipagin, all in [w/v]) in a 12-hour light: 12-hour dark cycle. Propionic acid was omitted from fly food as acidity affects egg laying (Gou et al., 2014). Genetic crosses were done in vials and kept at low density to ensure larvae were not competing for food and if necessary, additional live yeast was added. For all behavioural assays, virgin and mated Canton-S were used as controls. Virgin females, for example, from crosses of GAL4 with UASmSP, were collected after emergence within a 5-hour window and well-fed with live yeast sprinkled on food for maximum egg production and allowed to sexually mature (3–5 days).

To recombine second chromosome inserts for split-GAL4AD (attP40) and third chromosome split-GAL4DBD (attP2), standard genetic crossing schemes were used and final stocks were balanced with CyO and TM3 Sb (combined from ST and CT stock, see Key Resources Table). Split-Gal4AD and DBD combination lines were then crossed to UASmSP. For meiotic recombination, final stocks were validated by behavioural analysis for UAS mSP, for flp with eFeG UASCD8GFP to monitor GFP expression and for otdflp UASstopTrpA and otdflp UASstopTNT by crossing to elavGAL4 and monitored by lethality.

For enhanced recombination with flp, virgin females were transferred to 30°C after eclosion and kept for 5 days at this temperature before performing the behavioural assays. For induction of neuronal activity by temperature-sensitive TrpA1, females were kept at 30°C.

To make UAS FRTstopFRT mSP, a gBlock (IDT) stop cassette with the FRT sequences used in the eFeG plasmid (Haussmann et al., 2008) was inserted into NotI cut pUAST-GGTmSP (gift from T. Aigaki) by Gibson assembly. In the stop-cassette, the FRT sequence is followed by a GFP with a 3’UTR from ewg containing polyA site 1 from intron 6 (Haussmann et al., 2011). Flies were transformed by P-element-mediated transgenesis and inserts on each chromosome were established that show a robust PMR with dsxflp indistinguishable from mated females.

Females were examined for the main post-mating behaviours receptivity and oviposition as described previously and as follows (Soller et al., 1999; Soller et al., 2006). To generate mated females, one female and three males were added to fly vials and observed until mating and males were removed after mating. For receptivity tests, mature 3–7-day-old virgin or mated females were added to fly vials (95 mm length and 24 mm diameter) containing Canton S males with an aspirator and observed for 1 hour, generally three females and seven males. For these experiments, males were separated from females at least 1 day before the experiment. Receptivity tests were done in the afternoon with virgins, or 5–24 hours after mating for controls. For oviposition, females were placed individually in fly vials in the afternoon and the number of eggs laid was counted the next day. Receptivity and oviposition tests were tested were done blinded.

Sample size was based on previous studies, non-blinded and not predetermined by statistical methods (Soller et al., 1997; Haussmann et al., 2013; Nallasivan et al., 2021). Behavioural data are representatives of at least three replicates that were performed on three different days. Statistical analysis of behavioural experiments was performed using GraphPad Prism 9 (GraphPad by Dotmatics, RRID:SCR_002798) using one-way ANOVA followed by pairwise comparisons with Tukey’s test.

For the analysis of adult neuronal projection from UAS CD8GFP, UAS H2BYFP,UASmyrGFP, lexAopNLStomato, or QUAS mtdtomato3xHA expressing brains, VNCs or genital tracts, tissues were dissected in PBS (137 mM NaCl, 10 mM phosphate, 2.7 mM KCl, pH 7.4), fixed in 4% (w/v in PBS) paraformaldehyde for 15 minutes, washed three times in PBST (PBS with 1% BSA and 0.3% Triton-X100), then once in PBS for 10 minutes, mounted in Vectashield (Vector Labs) and visualised with confocal microscopy using a Leica TCS SP8. If signals were weak, antibody in-situ stainings were done as described previously (Haussmann et al., 2008) for validation using rat anti-HA (MAb 3F10, 1:20; Roche), rabbit anti-GFP (Molecular Probes, 1:100) and visualised with Alexa Fluor 488 (1:250; Molecular Probes or Invitrogen), Alexa Fluor 546 (1:250; Molecular Probes or Invitrogen), or Alexa Fluor 647 (1:250; Molecular Probes or Invitrogen). For imaging, tissues were mounted in Vectashield (Vector Labs).

Adult tissues were scanned using a Leica SP8 confocal microscope equipped with a set of fluorescent filters and hybrid detector (HyD). Adult brains were scanned using a 40× HC PL APO 40×/1.30 lens with oil, 1024 × 1024 resolution and 0.96 µm Z-step. VNC and genital tracts were scanned using a HC PL APO CS2 20×/0.75 with oil, 1024 × 1024 resolution and 0.96 µm Z-step. Images were obtained using Leica Application Suite X (LAS X) imaging acquisition software. Raw data files were in LIF format and were processed using FIJI RRID:SCR_002285.

For high-resolution mapping, neurons were identified in the virtual fly brain based on registered GAL4 expression and traces retrieved for modelling (Scheffer et al., 2020; Phelps et al., 2021; Galili et al., 2022).