High-resolution comparative single-cell transcriptomics of doublesex -expressing neurons reveals evolutionary conservation and diversity of sexual circuits in Drosophila

High-resolution comparative single-cell transcriptomics of doublesex-expressing neurons reveals evolutionary conservation and diversity of sexual circuits in Drosophila | bioRxiv /* */ /* */ <!-- <!-- /*! * yepnope1.5.4 * (c) WTFPL, GPLv2 */ (function(a,b,c){function d(a){return"[object Function]"==o.call(a)}function e(a){return"string"==typeof a}function f(){}function g(a){return!a||"loaded"==a||"complete"==a||"uninitialized"==a}function h(){var a=p.shift();q=1,a?a.t?m(function(){("c"==a.t?B.injectCss:B.injectJs)(a.s,0,a.a,a.x,a.e,1)},0):(a(),h()):q=0}function i(a,c,d,e,f,i,j){function k(b){if(!o&&g(l.readyState)&&(u.r=o=1,!q&&h(),l.onload=l.onreadystatechange=null,b)){"img"!=a&&m(function(){t.removeChild(l)},50);for(var d in y[c])y[c].hasOwnProperty(d)&&y[c][d].onload()}}var j=j||B.errorTimeout,l=b.createElement(a),o=0,r=0,u={t:d,s:c,e:f,a:i,x:j};1===y[c]&&(r=1,y[c]=[]),"object"==a?l.data=c:(l.src=c,l.type=a),l.width=l.height="0",l.onerror=l.onload=l.onreadystatechange=function(){k.call(this,r)},p.splice(e,0,u),"img"!=a&&(r||2===y[c]?(t.insertBefore(l,s?null:n),m(k,j)):y[c].push(l))}function j(a,b,c,d,f){return q=0,b=b||"j",e(a)?i("c"==b?v:u,a,b,this.i++,c,d,f):(p.splice(this.i++,0,a),1==p.length&&h()),this}function k(){var a=B;return a.loader={load:j,i:0},a}var l=b.documentElement,m=a.setTimeout,n=b.getElementsByTagName("script")[0],o={}.toString,p=[],q=0,r="MozAppearance"in l.style,s=r&&!!b.createRange().compareNode,t=s?l:n.parentNode,l=a.opera&&"[object Opera]"==o.call(a.opera),l=!!b.attachEvent&&!l,u=r?"object":l?"script":"img",v=l?"script":u,w=Array.isArray||function(a){return"[object Array]"==o.call(a)},x=[],y={},z={timeout:function(a,b){return b.length&&(a.timeout=b[0]),a}},A,B;B=function(a){function b(a){var a=a.split("!"),b=x.length,c=a.pop(),d=a.length,c={url:c,origUrl:c,prefixes:a},e,f,g;for(f=0;f<d;f++)g=a[f].split("="),(e=z[g.shift()])&&(c=e(c,g));for(f=0;f<b;f++)c=x[f](c);return c}function g(a,e,f,g,h){var i=b(a),j=i.autoCallback;i.url.split(".").pop().split("?").shift(),i.bypass||(e&&(e=d(e)?e:e[a]||e[g]||e[a.split("/").pop().split("?")[0]]),i.instead?i.instead(a,e,f,g,h):(y[i.url]?i.noexec=!0:y[i.url]=1,f.load(i.url,i.forceCSS||!i.forceJS&&"css"==i.url.split(".").pop().split("?").shift()?"c":c,i.noexec,i.attrs,i.timeout),(d(e)||d(j))&&f.load(function(){k(),e&&e(i.origUrl,h,g),j&&j(i.origUrl,h,g),y[i.url]=2})))}function h(a,b){function c(a,c){if(a){if(e(a))c||(j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}),g(a,j,b,0,h);else if(Object(a)===a)for(n in m=function(){var b=0,c;for(c in a)a.hasOwnProperty(c)&&b++;return b}(),a)a.hasOwnProperty(n)&&(!c&&!--m&&(d(j)?j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}:j[n]=function(a){return function(){var b=[].slice.call(arguments);a&&a.apply(this,b),l()}}(k[n])),g(a[n],j,b,n,h))}else!c&&l()}var h=!!a.test,i=a.load||a.both,j=a.callback||f,k=j,l=a.complete||f,m,n;c(h?a.yep:a.nope,!!i),i&&c(i)}var i,j,l=this.yepnope.loader;if(e(a))g(a,0,l,0);else if(w(a))for(i=0;i (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0];var j=d.createElement(s);var dl=l!='dataLayer'?'&l='+l:'';j.src='//www.googletagmanager.com/gtm.js?id='+i+dl;j.type='text/javascript';j.async=true;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-M677548'); Skip to main content Home About Submit ALERTS / RSS Search for this keyword Advanced Search New Results High-resolution comparative single-cell transcriptomics of doublesex -expressing neurons reveals evolutionary conservation and diversity of sexual circuits in Drosophila View ORCID Profile Justin Walsh , View ORCID Profile Ian P. Junker , View ORCID Profile Yu-Chieh David Chen , Yen-Chung Chen , Helena Gifford , Yun Ding doi: https://doi.org/10.1101/2025.05.06.652433 Justin Walsh 1 Department of Biology, University of Pennsylvania , Philadelphia, PA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Justin Walsh For correspondence: yding19{at}sas.upenn.edu Ian P. Junker 1 Department of Biology, University of Pennsylvania , Philadelphia, PA, USA 2 Department of Biological and Biomedical Sciences, Vanderbilt University , Nashville, TN, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Ian P. Junker Yu-Chieh David Chen 3 Department of Biology, New York University , New York, NY, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Yu-Chieh David Chen Yen-Chung Chen 3 Department of Biology, New York University , New York, NY, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Helena Gifford 1 Department of Biology, University of Pennsylvania , Philadelphia, PA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Yun Ding 1 Department of Biology, University of Pennsylvania , Philadelphia, PA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site For correspondence: yding19{at}sas.upenn.edu Abstract Full Text Info/History Metrics Supplementary material Preview PDF Abstract Understanding how the cellular and molecular composition of neural circuits change during evolution is essential for deciphering how behavior evolves. Male sexual behaviors in Drosophila species are remarkably diverse, and the underlying sexual circuits are specified by sex determination genes. Here, we employed single-cell transcriptomics to systematically characterize and compare neuron cell types that express the sex determination gene doublesex in adult males across Drosophila species with divergent sexual behaviors. High-resolution profiling to delineate cellular diversity revealed a largely conserved set of cell types across four species, with minimal evolutionary gain or loss and potentially more prevalent changes in cell type abundance. In-depth comparisons between D. melanogaster and D. yakuba showed that transcriptomic conservation and differentially expressed genes between species are highly cell-type-specific, suggesting that cell types evolve as highly independent units. We identified widespread species differences in gene expression, particularly in the neuromodulatory signaling pathway, while preserving a conserved circuit layout in sexual identity and neurotransmitter properties. We further generated a female dataset in D. melanogaster to define sex differences in cell types and examine species divergence in relation to sex differences. Finally, we reported marker gene combinations that uniquely define each cell type, providing a foundational resource that enables the design of cell -type-specific genetic reagents. Overall, our study provides fundamental insights into the cellular diversity of sexual circuits and how evolution shapes cell types and gene expression in behavioral adaptations. Introduction Even among closely related species, behaviors may vary drastically, and this behavioral variation largely arises from evolutionary changes in the structure and function of the underlying homologous neural circuits ( Katz & Harris-Warrick, 1999 ; Katz, 2011 ; Roberts et al., 2022 ). These changes can occur in the presence and absence of specific cell types, the abundance of a particular cell type, or gene expression that influences neuron anatomy and physiology. Facilitated by the unprecedented capacity of singl e-cell RNA sequencing (scRNA-seq) to define molecular cell types, recent comparative studies have begun to shed light on species conservation and differences in cell types ( Shami et al., 2020 ; Khrameeva et al., 2020 ; Schafer et al., 2022; Geirsdottir et al., 2019 ; Wei et al., 2022 ; Lee & Benton, 2023 ; Wang et al., 2024 ). However, existing studies often focus on distantly related species and/or have limited resolution to capture fine-scale changes in cell types. Particularly, since the basic cellular organization of the nervous system tends to be largely conserved among closely related species (Katz & Harris -Warrick, 1999, Katz, 2011 ), subtle changes in the nervous system may result in major behavioral differences but go undetected in existing datasets. With few exceptions of relatively simple organisms (e.g., Zhao et al., 2008 ; Pollo et al., 2024 ; Toker et al., 2024 ), it remains unclear how much neuronal cell types differ among closely related species with divergent behaviors at a fine-grained level. Furthermore, it is unknown whether certain cell types exhibit greater divergence across species and how gene expression evolves at the level of individual cell types. Drosophila species, with their rich behavioral diversity and experimental accessibility, serve as a powerful system to study behavioral evolution through species comparisons ( Prieto-Godino et al., 2017 ; Seeholzer et al., 2018 ; Ding et al., 2019 ; Auer et al., 2020 , 2022 ; Takagi et al. 2024 ; Ye, et al., 2024 ; Coleman et al., 2024 ; Li et al., 2024 ). Drosophila sexual behaviors present a unique opportunity as the genes and neural circuits underlying these behaviors have been well studied in the model species D. melanogaster and these behaviors have rapidly diversified across species (Spieth et al., 1974; Yamamoto & Koganezawa, 2013 ; Shirangi et al., 2016; Ding et al., 2019 ; Shen et al., 2023 ; Lillvis et al., 2024 ). For example, Drosophila males perform complex and diversified courtship rituals such as singing a species -specific courtship song by vibrating wings ( Ewing & Bennet-Clark 1968 ; Greenspan & Ferveur, 2000 ). Species also vary in their responses to sensory cues that are visual, auditory, and chemosensory ( Ferveur, 2005 ; Billeter et al., 2009 ; Gleason et al., 2012 ; Seeholzer et al., 2018 ; Khallaf et al., 2021 ). For example, the sex pheromone 7-tricosene is male-specific in D. melanogaster , where it suppresses male courtship, but is sexually monomorphic in D. yakuba and instead promotes male courtship ( Coleman et al., 2024 ). These sexual behaviors are orchestrated by two key transcription factors in the sex determination pathway: doublesex ( dsx ) and fruitless ( fru ). The dsx gene is spliced into sex-specific isoforms that regulate the development and function of sexual circuits ( Villella & Hall, 1996 ; Rideout et al., 2007 ; Sanders & Arbeitman, 2008 ; Rideout et al., 2010 ; Shirangi et al., 2016). Approximately 900 dsx + neurons, representing about 1% of the adult male central nervous system, are distributed in anatomically and functionally distinct neuronal clusters. For instance, the male-specific TN1 cluster in the ventral nerve cord (VNC) is a group of motor patterning neurons responsible for generating courtship song (Shirangi et al., 2016); the sexually dimorphic pC1 cluster in the brain integrates sensory cues and contains functionally heterogeneous populations that encode sexual and aggressive states (Kimura et al., 2008; von Philipsborn et al., 2011 ; Deutsch et al., 2020 ; Schretter et al., 2020 ). Given the expanding knowledge of their roles in sexual behaviors–which evolve rapidly across species– dsx + neurons provide an excellent inroad for uncovering the cellular and molecular mechanisms underlying the evolution of complex behaviors and species-specific adaptations. Indeed, we have previously linked the lineage-specific loss of sine song (a type of courtship song) in D. yakuba to the loss of a neuronal subtype within the TN1 cluster ( Ye, et al., 2024 ). In this study, we overcome the common challenge of cellular resolution in comparative scRNA-seq studies by profiling genetically labeled dsx + neurons across Drosophila species. This effort resulted in a scRNA-seq atlas comprising 46,922 dsx + neurons, with each dsx + neuron represented an average of 52.1 times. The dataset included four species within the D. melanogaster subgroup: D. melanogaster (19.1x coverage), D. yakuba (20.1x), D. santomea (6.0x), and D. teissieri (7.1x) ( Supplemental Table 1 )—a group of closely related species that exhibit a wide range of differences in sexual behaviors and are emerging models for comparative studies of the nervous system and behavior ( Stern et al., 2017 ; Ding et al., 2019 ; Coleman et al., 2024 ; Ye, et al., 2024 ; Li et al., 2024 ). In total, we molecularly defined 84 cell types, all systematically mapped to and hierarchically organized by known dsx + neuronal clusters, providing a comprehensive characterization of cellular diversity within this neuronal population of broad interests. We found that sexual circuits evolve through widespread cell-type-specific gene expression changes on top of a largely conserved organization in cell types, sex identity, and neurotransmitter properties. We also generated a scRNA-seq dataset in adult D. melanogaster females to characterize female dsx + cell types and explore the relationship between species and sex differences. Finally, we compiled a list of marker gene combinations that uniquely target each cell type and hosted our datasets on a Shiny app interface that allows user-defined tasks such as gene expression inquiries and molecular marker detection across cell types, species, and sexes. Altogether, our study offers both novel insights and useful resources for understanding and dissecting the function and evolution of neural circuits. Results High-resolution characterization of cell type diversity within dsx+ neurons We genetically labeled dsx + neurons using CRISPR/Cas9-mediated genome editing ( Ye et al., 2024 ) and fluorescently sorted them from adult male brains and VNCs of four Drosophila species that diverged within the past 12 million years ( Figure 1A ; Tamura et al., 2004 ; Bachtrog et al., 2006 ). In total, we generated three replicates for D. melanogaster and D. yakuba, including one previously published replicate for each species ( Ye, et al., 2024 ). We primarily focused on the comparison of cell types and gene expression between these two species of common interests ( Stern et al., 2017 ; Ding et al., 2019 ; Ye, et al., 2024 ; Coleman et al., 2024 ). We also included one sample each for D. santomea and D. teissieri to provide a general perspective of cell type conservation across species and phylogenetic context for gene expression changes we identified between D. melanogaster and D. yakuba ( Figure 1A ). Download figure Open in new tab Figure 1. Identification of parental clusters in scRNAseq data in four species. (A) Left: schematic representation of the cell bodies of dsx+ neurons in the brain and VNC, color-coded by parental clusters, and a phylogeny of species included in this study (divergence times shown in millions of years ago, based on Tamura et al., 2004 ; Bachtrog et al., 2006 ). Gray dots represent dsx+ neurons that are not identified in our scRNA-seq data. Right: UMAP representations of scRNA-seq data from four species, color-coded to match the dsx+ parental clusters on the left. The top row shows single species (non-integrated) data, while the bottom row displays the integrated data separated by species. (B) Heatmap showing the similarity (AUROC) of parental clusters between D. melanogaster and D. yakuba . (C) Dotplot illustrating the average expression and percentage of cells expressing three marker genes for each parental cluster. Genes highlighted in red were used to identify parental clusters via split-GAL4. All genes shown are conserved across all four species. (D) Barplot displaying the number of marker genes conserved across species, including only the top 20 marker genes for each parental cluster for each species. Marker genes from all nine parental clusters are shown. Starting with D. melanogaster scRNA-seq data, we performed an initial clustering with the goal of identifying previously defined dsx + neuronal clusters ( Lee et al., 2002 ; Kimura et al., 2008; Rideout et al., 2007 ; Sanders & Arbeitman 2008 ; Robinett et al., 2010 ; Rideout et al., 2010 ), which we refer to as “parental clusters” ( Figure 1A ). Upon identifying unique marker genes, we used a split-GAL4 genetic intersectional approach, pairing dsx hemi-drivers with hemi-drivers against marker genes for each parental cluster. This strategy, combined with previously known marker genes such as Hox genes, allowed us to validate the fidelity of the clustering and successfully assign parental cluster identities: aDN ( tey +); pC1 ( Optix +); pC2l (TfAP-2 +, tsh -); pC2m ( sv +); pCd-1 ( unc-4 +); pCd-2 ( Gad1 +); TN1 ( TfAP-2 +, tsh +); TN2 ( ara +, abd-A -, Abd-B -), ANm ( abd-A + and/or Abd-B +) ( Figure 1B and 1C and Supplemental Figures 1 and 2A ). We did not identify cell clusters corresponding to the single-neuron pairs pMN1, pMN3, pLN, and sLG and the TN2 single-neuron pairs prA, prC, msB (ara partially labeled TN2), which together comprise less than 2% of all dsx + neurons (cell bodies shown in gray color in Figure 1A ). As these unidentified neurons exhibit larger soma, we speculate that they might be lost during sample preparation due to sensitivity to experimental procedures or stringency of cell sorting. Download figure Open in new tab Figure 2. Subclustering analyses of the parental clusters from the integrated four-species dataset. (A) UMAP representations of the subclustering of parental clusters. Different colors represent molecularly distinct cell types. (B) Representative dotplots showing the average expression and percentage of cells expressing two marker genes for each subcluster within pC1 (top) and pCd-2 (bottom). (C) Barplot displaying the number of marker genes conserved across species, including only the top 20 marker genes for each subcluster for each species. Marker genes from all subclusters are shown. Following this initial clustering, which faithfully recapitulated existing knowledge of dsx + clusters, we transferred the identities of parental clusters in D. melanogaster to the other three species using the expression of the above marker genes. Furthermore, we used MetaNeighbor ( Crow et al., 2018 ) to ensure our transferred parental identities were accurate (i.e., had the highest AUROC scores) ( Figure 1B , Supplemental Figure 3 ). This high degree of species consensus justified the further integration of data from all the four species into one dataset for identifying and comparing cell types on a common clustering space ( Figure 1C ). Overall, the marker genes for parental clusters were well-conserved across species. The majority of the top 20 markers for each parental cluster in a species were also marker genes for the same cluster in the other three species ( Figure 1D ). To resolve the cellular diversity within each parental cluster, we extracted each parental cluster from the integrated dataset for subclustering and identified a total of 84 subclusters that we refer to as cell types ( Figure 2A ). The aDN, pCd-2, and TN2 clusters resolved into 2, 3, and 6 subclusters, respectively ( Figure 2A ), each likely representing a cell type corresponding to a single-neuron pair, most of which are distinguishable by a single marker gene ( Figure 2B and Supplemental Figure 4 ). Larger parental clusters that likely originate from a single hemilineage origin, such as pC1 and pC2 ( Ren et al., 2016 ) — whose heterogeneity has been widely appreciated but remains poorly understood—resolved into several subclusters, each typically representing a group of neurons ( Figure 2B and Supplemental Figure 4 ). While some of these subclusters are distinguishable by a single marker (e.g., pC1_6), others (e.g., pC1_1 and pC1_3) are harder to define and often display continuity in clustering space ( Figure 2A , 2B ). This continuity may reflect the biological nature of these cell types defined by the quantitative expression levels of multiple genes. The ANm cluster, as expected given its diverse hemilineage origins ( Sanders & Arbeitman, 2008 ; Rideout et al., 2010 ), resolved into a large number (50) of more discrete subclusters, many of which are well-defined by one or two marker genes, revealing extensive cellular diversity ( Figure 2A ). Based on described hemilineage markers ( Bossing et al., 1996 ; Schmid et al., 1999 ; Lacin et al., 2019 ; Allen et al., 2020 ), we annotated the hemilineage origin for 25 of the 50 ANm subclusters, representing at least 16 distinct hemilineages ( Supplemental Figure 2 ; Supplemental Table 2 ). Similar to the marker genes for parental clusters, these subcluster marker genes are largely shared across four species ( Figure 2C ). Unless stated otherwise, all subsequent analyses were conducted on these subclusters. Overall conservation of dsx+ cell types across species The systematic characterization of cell types set the stage for species comparison. Focusing on D. melanogaster and D. yakuba , where biological replicates were available, we assessed species differences in the relative abundance of cell types using scCODA, a Bayesian model designed to detect differences in cell numbers across groups based on scRNA-seq data ( Büttner et al., 2021 ). To ensure the accuracy of our comparison, we manually quantified the number of cells in each dsx+ parental cluster across species by immunostaining (excluding ANm due to its high density of dsx + neurons, making precise counting difficult) ( Supplemental Table 3 ). We then scaled the scRNA-seq cell count for each cell type accordingly so that the size of each parental cluster aligned with the immunostaining results. Among the 84 molecularly distinct subclusters, the only species-specific one was TN1_4 ( Figure 3A and 3B ). This subcluster corresponds to one we previously identified ( Ye, et al. 2024 ), which is a TN1 subtype associated with the lineage-specific loss of sine song in D. yakuba males. We also defined 12 species-biased subclusters that varied in cell number by 1.3-to 6.8-fold. Of these, ten were in the ANm (ANm_4/8/17/18/24/25/32/33/45/48, all D. yakuba -biased), the other two were TN1_2 ( D. melanogaster -biased) and pC1_6 ( D. yakuba -biased) ( Figure 3A and 3B ). Each of the three ANm subclusters with annotated hemilineage origins (ANm-25/32/45) originates from a distinct hemlineage ( Supplemental Figure 2B ). These subclusters are potential candidates of more variable cell types across species for future validation and functional investigation. Download figure Open in new tab Figure 3. Cell type number comparison between D. melanogaster and D. yakuba . ( A ) UMAP representations of the subclustering of parental clusters, color-coded by species-biases in cell number within each subcluster. ( B ) Barplot showing the normalized number of cells (N) per fly in each subcluster, separated by species and grouped by parental cluster identity shown at the bottom. Numbers were plotted on a log2 scale with an offset of 1 (log2(N+1)) along the Y-axis to accommo-date a wide range of values and account for zeros. Of the 83 subclusters shared between D. melanogaster and D. yakuba , 82 were present in all four species. A small subcluster in the ANm (ANm_47) was not detected in D. santomea , which may reflect a true species difference or could be due to its lower coverage of dsx+ neurons. We did not perform scCODA analysis for D. santomea and D. teissieri due to the lack of biological repeats. Nevertheless, the comparison across all four species strongly indicates that species differences in the presence or absence of cell types are rare. Cell type-specific transcriptomic conservation and divergence across species A neuron’s morphology and function, to a great extent, are the product of its gene expression. Therefore, we asked if some cell types are more similar or different in gene expression profile across species, potentially revealing circuit nodes with more constrained or divergent functions. For each subcluster, we used Spearman rank correlation of gene expression across species to provide a continuous measure for the degree of transcriptomic conservation ( Kim et al., 2018 ; Skinnider et al., 2019 ). Subclusters with higher across-species correlation scores indicate greater transcriptomic conservation, whereas those with lower scores indicate greater transcriptomic divergence. We found that the correlation scores between D. melanogaster and D. yakuba , ranging from 0.60 to 0.73, varied substantially even among cell types within the same parental cluster ( Figure 4A ). Download figure Open in new tab Figure 4. Variation across subclusters in transcriptomic conservation between D. melanogaster and D. yakuba . (A) UMAP representations of the subclustering of parental clusters, color-coded by transcriptomic correlation scores measured by Spearman rank correlation scores between D. melanogaster and D. yakuba . (B) Transcriptomic correlation scores of subclusters grouped by parental cluster; no significant difference was observed across parental clusters (Kruskal Walllis; χ 2 = 10.793; p = 0.213). (C) Box and whisker plot showing transcriptomic correlation scores of subclusters grouped by central nervous system location, no significant difference was observed (t-test; t = 0.486, df = 39.2, p = 0.629). (D) Box and whisker plot showing transcriptomic correlation scores grouped by whether or not the subcluster expresses fru , no significant difference was observed (t-test; t = -0.4297, df = 72.6, p = 0.767). (E) Box and whisker plot showing transcriptomic correlation scores when including different sets of genes-all genes with median values greater than zero, only marker genes, or only transcription factors (TFs) (ANOVA; F = 82.28, df = 2, p < 0.001). We examined whether the heterogeneity in transcriptomic conservation among subclusters could be caused by possible technical characteristics. First, we tested whether the across -species correlation score simply reflected how well the cell type can be molecularly defined. Cell types that are less well-defined might lead to misassignment of cell types, which may artificially cause lower correlation scores. To check this, we calculated the mean silhouette score ( Tang et al., 2020 ), a measure of clustering quality, and found no association ( Supplemental Figure 5A ). Similarly, the correlation score was not associated with the number of cells in the subcluster ( Supplemental Figure 5B ), or the variation across replicates Supplemental Figure 5C ). We did find a significant positive association with the number of genes expressed within a subcluster ( Supplemental Figure 5D ). However, this did not seem to be a technical artifact because randomly downsampling the gene set to 1,000 for 1,000 iterations did not alter the association (Spearman rank correlation between the mean of bootstrap iterations and all expressed genes: r = 0.999, p < 0.001), raising the possibility that cell types with higher transcriptional activity may exhibit greater transcriptomic conservation. Using a generalized linear model, we found that the number of expressed genes explains 31.8% of the variance in the correlation scores among subclusters. We further examined whether the heterogeneity could be explained by several biological characteristics. First, we asked whether specific parental clusters show greater conservation or divergence. We found that the variance of correlation scores across parental clusters was not homogenous (Bartlett test; K 2 = 16.482; p = 0.0356); however, when investigating the non-homogeneity of variance by performing pairwise Bartlett tests across parental groups, there were no significant differences between any two parental clusters. We further performed a non-parametric Kruskal-Wallis test and found that the correlation scores were not different among parental clusters ( Figure 4B ). Second, the scores also did not differ between subclusters located in the brain versus the VNC ( Figure 4C ). Third, we tested whether they varied as a function of fru gene expression fru , given that fru is a key transcription factor in the development and function of sexual circuits in males ( Kimura et al., 2005 , 2008; Cachero et al., 2010 ; Rideout et al., 2010 ; Robinett et al., 2010 ) and again found no difference ( Figure 4D ). Together, these results suggest that transcriptomic conservation and divergence are highly cell - type-specific rather than driven by broader biological characteristics. The five subclusters with the lowest across-species correlation scores comprised four ANm subclusters (ANm_12/19/39/40) and pC2l_3. These cell types of greater transcriptomic divergence are potential candidates for functionally divergent cell types. The five most conserved subclusters include three ANm subclusters (ANm_17/23/25), pC1_6, and pC2l_6. Notably, pC1_6 also appeared more abundant in D. yakuba , highlighting independent evolutionary modes between transcriptomic pattern and cell type abundance. Finally, we investigated how genes crucial to cell identity contribute to transcriptomic conservation across species. Performing Spearman rank correlations using either only marker genes or only transcription factors for Spearman rank correlations resulted in significantly higher correlation scores, with the highest observed when using only transcription factors ( Figure 4E ). This pattern suggests stronger evolutionary constraints on the expression of genes essential for cell identity, particularly transcription factors. Differentially expressed genes (DEGs) across species are highly cell type-specific We identified DEGs for each subcluster between D. melanogaster and D. yakuba and assessed their cell-type specificity. Ambient RNA in single-cell suspensions can introduce contamination, leading to false-positive signals that appear as lowly expressed genes ( Macosko et al., 2015 ; Lun et al., 2019; Caglayan et al., 2022 ), artificially inflating the breadth of gene expression. Since fast-acting neurotransmitters typically show minimal co-transmission within a given cell type ( Lacin et al., 2019 ; Shih et al., 2019 ; Davis et al., 2020 ), we used the co-expression of their marker genes ( Gad1 , VGlut , and VAChT ) as a proxy for potential ambient RNA contamination ( Allen et al., 2020 ) to explore optimal criterion of gene expression. When defining gene expression as non-zero read(s) in ≥10% of cells within a given subcluster, 20.2% (17 of 84) of subclusters appeared to release more than one neurotransmitter. In these cases, one neurotransmitter marker was expressed at a much higher level than the others, suggesting contamination. Alternatively, defining gene expression using a Bayesian mixture model by scMarc o ( Chen et al., 2023 ), where a gene is considered expressed if it has >50% probability of being detected, yielded results that aligned more closely with expectations: 82 of 84 subclusters were shown to release at least one fast-acting neurotransmitter, with only two subclusters releasing more than one neurotransmitter, indicating limited false-negatives and false-positives. Based on this approach to filter out genes that were not expressed in either species, we further identified DEGs using a threshold of a log fold change of three and a corrected p -value below 0.05. Following these guidelines, 504 genes are DEGs in at least one subcluster. These genes are mostly differentially expressed in a cell type-specific manner, with many genes being differentially expressed in only a small subset of subclusters where they are expressed ( Figure 5A ), regardless of whether they are sparsely or broadly expressed ( Figure 5B ). Download figure Open in new tab Figure 5. DEG analyses between D. melanogaster and D. yakuba . (A) Barplot showing the number of subclusters in which a gene is expressed, color-coded by the number of those subclusters in which the gene is differentially expressed and the direction of the bias. Each bar represents a single gene. (B) Box and whisker plot showing the percentage of subclusters in which a gene is differentially expressed, binned by the number of subclusters in which the gene is expressed. (C) Left: UMAP representation showing the expression of fru in pC1 in D. melanogaster . Right: Violin plots showing the expression levels of fru in both D. melanogaster and D. yakuba in each of the six pC1 subclusters. The expression level of fru is not significantly different across species in any of the subclusters. (D) fru expression in pC2l plotted similarly to panel C. (E) Barplot showing the number of subclusters in which neuropeptides (left) and neuropeptide receptors (right) are expressed, including only those that are differentially expressed in at least one subcluster, color-coded by the direction of bias. (F) Violin plots showing the expression levels of Dh31 and Tk in both D. melanogaster and D. yakuba in all the subclusters where expression levels are significantly different. between the two species. (G) Species differences in the gene expression patterns of Dh31 and Tk in the TN2-5 cell type. Left: schematic denotes the cell body location of TN2-5 and the appropriate imaging region (open black box). Right: representative confocal images of in situ HCR against dsx , Dh31 , and Tk mRNA. Images presented as maximum intensity projections of z-stacks covering the cell body of TN2-5. Scale bar: 20µm. Of the 504 DEGs, 13.3% (67 genes) exhibited higher expression in one species for certain subclusters and lower expression in others, again highlighting the cell-type specificity of gene expression evolution. The remaining genes displayed consistently higher or lower expression in one species across DEG subclusters. For these genes, we observed more DEGs with higher expression in D. melanogaster (60.1%, 303 genes) than in D. yakuba (26.6%, 134 genes). This bias cannot be explained by the percentage of reads mapped to the genome, as mapping rates were similar between species ( Supplemental Table 1 ). It also cannot be readily explained by species differences in genome annotation quality, as species-specific misannotations would be expected to have a broad effect on cell types that express the gene. Here, the bias remained even when considering only genes that were not DEGs in the vast majority (>95%) of expressed subclusters (71.8% of genes consistently higher in D. melanogaster and 22.9% in D. yakuba ). Therefore, this species bias might be biological. Across species comparison of specific genes or gene groups Conserved sexual identity defined by fru Changes in the spatiotemporal regulation of sex determination genes are thought to be a major mechanism driving the evolution of sex-specific traits, with many examples reported in morphological adaptations (e.g., Williams et al., 2008 ; Tanaka et al., 2011 ; Kijimoto et al., 2012 ; Kunte et al., 2014 ; Luo & Baker, 2015 ). However, how often such processes occur in the nervous system and contribute to behavioral evolution remains unclear. Like dsx , the sex determination gene fru plays a central role in defining the sex-specific fate of neurons and acts as a master regulator of sexual circuit development and function ( Goodwin & Hobert, 2021 ). The dsx+ neurons include both fru+ and fru- populations with known functional differences ( Rideout et al., 2007 ; Kimura et al., 2008; Sanders & Arbeitman, 2008 ; Robinette et al., 2010; Koganezawa et al., 2016 ; Coleman et al., 2024 ). For example, fru + and fru -populations within the parental cluster pC1 exhibit behavioral specialization in courtship versus aggression ( Koganezawa et al., 2016 ) and distinct sensory responses to sexual pheromones ( Coleman et al. 2024 ). Consistent with fru ’s prominent role in specifying cell fate, fru is a marker gene for 25 of the 84 subclusters, well above the average of 9.2 subclusters for all marker genes. Additionally, fru expression, both in presence/absence and quantitative levels, is highly spatially organized in UMAP space (e.g. pC1 and pC2l; Figure 5C , Figure 5D , and Supplemental Figure 6 ). While cell type-specific changes in fru expression might seem like a plausible evolutionary mechanism for major functional modifications, fru is not differentially expressed in any subcluster between D. melanogaster and D. yakuba ( Figure 5C , Figure and 5D ), and its expression patterns are highly similar across all four species ( Supplemental Figure 6 ). This suggests that fru -specified sexual identity may be highly conserved across species, and that the gain or loss of fru sexual identity may not be a common mode of evolutionary changes in the nervous system. Conserved neuronal properties of fast-acting neurotransmitters We annotated and compared subclusters that release the fast-acting neurotransmitters acetylcholine, GABA, and glutamate. Except for the highly diverse ANm cluster, all subclusters within each parental cluster release the same fast-acting neurotransmitter, and neurotransmitter production was consistent across all four species ( Supplemental Figure 7 ). Consistent with previous studies ( Zhou et al., 2014 ; Nojima et al., 2021 ; Imoto et al., 2024 ), pC1, pC2l, pC2m, pCd-1, TN1, and TN2 are cholinergic ( VAChT/VGAT +), pCd-2 is GABAergic ( Gad1 +), and aDN is glutamatergic ( VGlut +). Among the 50 ANm subclusters, 28 are cholinergic, 13 are GABAergic, seven are glutamatergic, and two lack expression of marker genes for these fast-acting neurotransmitters. Expression levels of these marker genes did not differ between D. melanogaster and D. yakuba in any subcluster. Conserved neuronal properties of monoamines Monoaminergic neurons are known to regulate sexual behaviors ( Lee & Hall, 2001 ; Certel et al., 2007 , 2010; Yilmazer et al., 2016 ; Zhang et al., 2016 ; Jois et al., 2018 ). To annotate monoaminergic neurons, we used Vmat , which encodes a monoamine vesicular transporter, as a marker gene ( Greer et al., 2005 ). Across all four species, Vmat is expressed in two ANm subclusters: ANm_14 is serotonergic ( SerT +) and did not express maker genes of fast-acting neurotransmitters; and ANm_20 is glutamatergic ( DAT +) ( Supplemental Figure 8 ). The serotonergic ANm_14 corresponds to a previously described neuron group that is necessary for successful copulation in D. melanogaster males ( Yilmazer et al., 2016 ). Expression levels of Vmat , SerT, and DAT did not differ between D. melanogaster and D. yakuba . Pervasive gene expression changes in neuropeptides and neuropeptide receptors Neuropeptide signaling pathways, along with multiple terms related to synaptic connectivity and axon guidance, are among the enriched GO terms for DEGs. Here, we specifically focused on the gene expression patterns of neuropeptides and neuropeptide receptors, given their well-established functions in reproductive and courtship behaviors (e.g. Castellanos et al., 2013 ; Lee et al., 2015 ; Sellami & Veenstra, 2015 ; Liu et al., 2019 ; Wu et al., 2019 ) and the recognized importance of neuromodulation in behavioral evolution ( Stoop, 2012 ; Taghert & Nitabach, 2012 ; Van Den Pol, 2012 ; Nässel & Zandawala, 2022 ; Jiang et al., 2024 ). Of the 23 neuropeptides expressed in at least one subcluster, nine are differentially expressed between D. melanogaster and D. yakuba in at least one subcluster ( Figure 5E ). Three of these genes– Dh31 , Dh44 , and Tk –have been implicated in sexual behaviors, including male courtship ( Asahina et al., 2014 ; Shankar et al., 2015 ; Lee et al., 2015 ; Lin et al., 2022 ; Kim et al., 2024 ). Notably, all three genes are DEGs only in a subset of the subclusters where they are expressed, with expression biased toward D. melanogaster in some subclusters and towards D. yakuba in others ( Figure 5D ). Of the 34 neuropeptides receptors expressed in at least one subcluster, 25 are differentially expressed in at least one subcluster, with 14 whose gene expressions are biased toward different species depending on the subcluster ( Figure 5F ). These 25 genes include CCKLR-17D3 and SIFaR , both of which play roles in male courtship ( Wu et al., 2019 , Sellami & Veenstra, 2015 ) and are DEGs in a small subset of expressed subclusters. DEG analysis revealed that the cell type TN2-5 expresses the neuropeptide Tk in D. melanogaster and Dh31 in D. yakuba ( Figure 5F ). We leveraged this intriguing pattern as a case for DEG validation. Based on molecular markers, we first registered TN2-5 to one of the two mesothoracic TN2 neuron pairs previously referred to as MsA (Robinette et al., 2010; Figure 5G and Supplementary Figure 9 ). Using multiplexed in situ Hybridization Chain Reaction (HCR; Choi et al., 2018 ), we then confirmed that one pair was consistently Tk + and Dh31 -in D. melanogaster and Tk - and Dh31 + in D. yakuba (the other pair was Tk - and Dh31 - in both species) ( Figure 5G ). Further scRNA-seq comparisons across four species showed that D. yakuba , D. santomea , and D. teissieri shared the same Tk- and Dh31 + pattern in TN2-5 ( Supplementary Figure 9 ). This raised the possibility that the shift in Tk and Dh31 expression co-occurred at the divergence between D. melanogaster and the common ancestor of the other three species. Defining sex differences in dsx+ cell types in D. melanogaster While many dsx + parental clusters are male-specific or more abundant in males, female D. melanogaster have about 360 dsx + neurons, many of which regulate key aspects of female sexual behaviors ( Lee et al., 2002 ; Kimura et al., 2008; Robinette et al. 2010; Rideout et al., 2010 ; Zhou et al., 2014 ; Deutsch et al., 2020 ; Schretter et al., 2020 ; Mezzera et al., 2020 ; Wang et al., 2020). However, the diversity of dsx+ cell types in females and the sex differences in dsx+ cell types–particularly in the ANm, where the majority of dsx + neurons are found in females ( Lee et al., 2002 ; Rideout et al., 2010 ; Deutsch et al., 2020 ) –remain mostly uncharacterized. To address this, we generated a scRNA-seq dataset of dsx + neurons in adult female D . melanogaster to define and compare cell types between sexes. This dataset not only provides foundational knowledge of female sexual circuits and sexual dimorphism but also enables our further exploration of the relationship between species- and sex-differences (see next session). The female data include 3,802 high quality dsx+ cells from adult female D. melanogaster , resulting in 10.8x coverage. To ensure comparability, we downsampled our male D. melanogaster data to the same coverage and integrated them for clustering to define sex-specific and sex-biased cell types. Using the same marker genes for males ( Supplemental Figure 1 ), we successfully identified each parental cluster in the integrated male/female dataset ( Figure 6A ). Consistent with previous studies ( Lee et al., 2002 ; Rideout et al., 2010 ; Deutsch et al., 2020 ), TN1, and TN2 are male-specific, while pC1, pC2l, and pC2m are male-biased ( Figure 6A and 6B ). We further resolved cell types by extracting each parental cluster for subclustering ( Figure 6C ) and defined sex-specific, and sex-biased cell types as those subclusters consisting of over 95% or 75% of cells from one sex, respectively. We note that this classification is based on transcriptomic patterns; it may not fully reflect the precise developmental origins. Among the 81 subclusters resolved, we classified 47 as sex-unbiased, 26 as male-specific, 2 as female-specific, 3 as male-biased, and 3 as female-biased ( Figure 6D and 6E ). The majority (89.7%) of male-specific and male-biased cell types are located in the brain, whereas all female-specific and female-biased ones are found in the ANm. Download figure Open in new tab Figure 6. Comparison of dsx + neurons across sexes in D. melanogaster . (A) UMAP representation of scRNA-seq data from D. melanogaster males and females, color-coded by parental cluster. (B) UMAP representation from panel A but color-coded by male and female cells. (C) UMAP representations of the subclustering of parental clusters shown in panel A . (D) UMAP representations of the subclusters, color-coded by sex-biases in cell number within each subcluster. (E) Barplot showing the normalized number of cells (N) per fly in each subcluster, separated by sex, and color-coded by sex-biases. Numbers were plotted on a log2 scale with an offset of 1 (log2(N+1)) along the Y-axis to accommodate a wide range of values and account for zeros. (F) Schematic representation showing our approach to match cell type identities across datasets by matching cell barcodes present in each dataset, color-coded by subcluster identity in the four species dataset. Only pC1 is shown as an example. (G) Box and whisker plot showing transcriptomic correlation grouped by whether the cell type is sex-unbiased across males and females, male-specific, or male-biased, no significant difference was observed (ANOVA; F = 0.0004, df = 2, p = 0.751). No clear association between sex differences and species differences Males and females may experience different fitness consequences from the same evolutionary changes ( Chapman et al., 2003 ; Arnqvist & Rowe, 2005 ; Lessells, 2006 ; Tregenza et al., 2006 ; Cox & Calsbeek, 2009 ). However, it remains unclear whether sex-unbiased cell types in sexual circuits are more evolutionarily constrained than sex-specific ones. Additionally, it is unknown whether sex-specific cell types are subject to stronger positive sexual selection. Therefore, we examined whether male-specific and male-biased dsx + cell types are evolutionarily more divergent than the sex-unbiased ones. To transfer the ‘male-specific’, ‘male-biased’ and ‘sex-unbiased’ labels from the male/female dataset to the four-species dataset for species comparisons, we cross-referenced the cell type identities of male cells present in both datasets ( Figure 6F and Supplemental Table 4 ). Of the 79 subclusters in the male/female dataset (excluding the two female-specific ones), 68 had clear one-to-one match in the four-species dataset, with >80% of cells aligning to a single subcluster. The remaining 11 had more complex relationships, often due to more-to-one matches in the four-species dataset. We tested whether male-specific cell types showed greater divergence between D. melanogaster and D. yakuba in abundance or transcriptomic patterns. Among the 26 male-specific cell types, three were either species-specific (TN1_4) or species-biased (pC1_6 and TN1_2), and one (pC2l_3) was among the five subclusters with the lowest transcriptomic correlation scores across species. However, male-specific cell types were not significantly overrepresented among species-specific or species-biased cell types (Fisher’s exact test; odds ratio = 1.338; p = 1) or among those with the five lowest transcriptomic correlations (odds ratio = 1.542; p = 1). Additionally, transcriptomic correlation scores did not differ among male-specific, male-biased, and sex-unbiased cell types ( Figure 6G ). Overall, we found no clear association between sex differences and species divergence. Resources for cell type-specific marker combinations and gene expression inquires Gaining genetic access to discrete cell types is crucial for investigating the organization, function, and evolution of the nervous system. Exploiting our datasets across species and between sexes, we used scMarco– a graphical user interface that binarizes gene expression in scRNA-seq data using a Bayesian mixture model approach ( Chen et al., 2023 ) – to identify optimal combinations of molecular markers for each cell type ( Supplemental material ). This approach informs genetic intersection strategies for cell-type-specific labeling and functional manipulations, a powerful technique that has been successfully demonstrated ( Tirian & Dickson, 2017 ; Luan et al., 2020 ; Chen et al., 2023 ; Ye, et al., 2024 ). For each subcluster defined in both the four-species dataset and the male/female dataset, we assessed individual marker genes (for genes uniquely expressed in just one cell-type) and pairs of marker genes by two key scores: (1) the clarity score (ranging from 0 to 1), which quantifies how uniquely a gene or gene pair labels a given subcluster, and (2) the bimodal score, which represents the likelihood that a gene is bimodally expressed in that subcluster. We report all genes and gene pairs with a clarity score of 1 for each of the four species individually and for male and female D . melanogaster ( Supplemental material ). We report all genes and gene pairs with clarity scores of 1, so that users can select the most convenient marker genes based on existing split-GAL4 lines or by the ease of making new lines. In the four species dataset, depending on the species, there are between 23 and 30 subclusters that can be labeled by just one gene (in addition to dsx ), and 15 (male) or 2 (female) such subclusters in the male/female dataset. These subclusters have NA values in the columns referring to “gene 2” in the Supplemental material . Users can cross-reference these marker genes or gene pairs with the D. melanogaster gene-specific split-GAL4 database ( https://splitgal4.org/ ) to identify existing driver lines or design new lines for targeted genetic manipulations ( Chen et al., 2023 ). Many split-GAL4 lines can be generated by simple genetic crosses ( Li et al., 2023 ). Users may further intersect the split-GAL4 with dsx FLP ( Rezával et al., 2014 ) to achieve labeling specificity as needed. Additionally, both the four-species and the male/female datasets are publicly accessible through a Shiny app ( https://apps.yenchungchen.com/dsx_neurons ), an interactive web portal designed to help users identify and evaluate optimal molecular markers for cell types of interests while also enabling gene expression queries across cell types, species, and sexes. These resources will facilitate both precise genetic manipulations in D. melanogaster and comparative studies across species. Discussion A prerequisite for understanding behavioral evolution is to decipher how changes in the molecular identity of cell types shape divergent circuit properties and behaviors. In this study, we generated a comprehensive cell type atlas of male dsx + neurons across four species of Drosophila ( Figure 1 and Figure 2 ). Our species comparison revealed key aspects of neural circuit evolution, including cell types, transcriptomic conservation, and gene expression ( Figures 3 -5 ). We also generated a cell type atlas for dsx + neurons in female D. melanogaster , defined sex differences in dsx+ cell types, and tested whether male-specific cell types exhibit distinct evolutionary patterns ( Figure 6 ). Finally, we systematically identified combinations of molecular markers that uniquely label each cell type, enabling precise cell type-specific functional characterization in future studies. By isolating relevant neurons associated with specific behaviors of rapid evolution, our study provided the previously lacking granularity in cell type evolution. Among the 84 defined cell types, ranging from one to about 23 neurons per hemisphere, we identified all but two across all four Drosophila species. We note that the appropriate level of granularity in defining cell types is an open question ( Zeng, 2022 ; Özel & Desplan, 2025 ). Here, we refined cell types by first identifying biologically meaningful parental clusters and then iteratively subdividing each until further division no longer yielded strong marker genes. More cell types may still be embedded in our dataset. For instance, several pC1 and pC2l subclusters contain both fru + and fru -populations, suggesting potential heterogeneity. Clustering strategies, such as species selected for data integration, may also influence the precise outcome of cell type identification. Nevertheless, our study provides a necessary baseline for investigating the cellular basis of behavioral evolution. The high-resolution datasets generated here allow continued exploration tailored to specific research questions and cell types of interest, such as adjusting clustering parameters, performing clustering on unintegrated or differently integrated datasets, combining scRNA-seq data of earlier developmental stages, and incorporating additional neuronal modalities such as neuron anatomy informed by light microscopy and EM connectome. In contrast to the overall conservation in the presence and absence of cell types, species differences in cell type abundance appear to be common, accompanied with widespread transcriptomic changes in a highly cell-type-specific manner. 12 of 84 cell types, spanning diverse functional populations, potentially vary in abundance between D. melanogaster and D. yakuba . At the transcriptional level, we observed substantial variation in transcriptomic conservation even among cell types within the same parental cluster. This heterogeneity also aligns with the cell-type-specificity of DEGs, as most of the 490 DEGs we identified were differentially expressed in only a small subset of cell types across species. Interestingly, the heterogeneity among cell types cannot be readily explained by various features we examined, such as cluster size, parental cluster identity, or sex specificity. These findings together suggest that the evolutionary processes shaping transcriptomic divergence, including “developmental system drift” and adaptive changes of new functions, operate in a highly cell-type-specific manner. At the level of individual cell types, our study revealed conserved and evolvable aspects of gene expression in sexual circuits. Surprisingly, despite long-standing appreciation of cis -regulatory changes of master regulators as key drivers of phenotypic evolution ( Wittkopp & Kalay, 2012 ), fru –a central orchestrator of male sexual circuits ( Goodwin & Hobert, 2021 ) as well as a previously implicated genetic locus of species divergence ( Gleason & Ritchie, 2004 ; Lagisz et al., 2012 ; Parker et al., 2014 ) –did not stand out as an obvious player. Both the composition of fru + and fru -populations in dsx + neurons and the quantitative expression of fru across cell types appear highly conserved. (we note that a substantial proportion of fru + neurons do not express dsx , thus species differences in fru expression that were not investigated here may exist). This observation aligns with recent work revealing a conserved functional subdivision of P1 neurons–a subset of pC1 neurons that acts as a core node of the male courtship circuit–based on fru expression ( Coleman et al., 2024 ). Similarly, the conservation of dsx + cell types across species, along with the absence of quantitative species differences in dsx expression levels, suggests that dsx -defined sexual identity is largely conserved. Other conserved features include greater transcriptomic conservation for genes crucial to cell identity (i.e., transcriptional factors and molecular marker genes) and a fully conserved pattern of neurotransmitter identities (i.e., fast-acting neurotransmitters and monoamines) across all dsx+ cell types in all four species. Given these conserved patterns, the widespread gene expression differences observed between D. melanogaster and D. yakuba– representing 10.2% of all genes expressed in the dataset–likely arise disproportionately from genes outside these fundamental properties. Particularly, genes involved in neuropeptide signaling, which have been broadly associated with behavioral evolution ( Katz, 2011 ; Schoofs et al., 2017 ), are significantly overrepresented in DEGs. Strikingly, 39.1% of neuropeptide genes and 73.5% neuropeptide receptor genes expressed in dsx + neurons, including five with established roles in regulating sexual behaviors, are differentially expressed in at least one cell type. These findings highlight the remarkably flexible and modular nature of their gene expression patterns in evolution. Together, our study revealed a dynamic evolutionary landscape of widespread cell-type-specific modifications built upon a largely conserved organization of male sexual circuits in cell types, sexual identity, and neurotransmitter identity. The profound functional diversity among cell types in the nervous system, coupled with the highly cell-type-specific mode of evolution, urgently calls for cell type-specific genetic tools for functional characterization. This need becomes even more critical for studying evolutionary changes, as comparing truly homologous neurons or neuronal groups across species is indispensable for appropriate functional interpretation. While transferring GAL4 and split-GAL4 enhancer drivers that label specific neurons in D. melanogaster into closely related species has proven to be a fruitful approach in comparative studies (e.g., Stern et al., 2017 ; Seeholzer et al., 2018 ; Ding et al., 2019 ; Luan et al., 2020 ; Ohashi et al., 2023 ; Ye, et al., 2024 ; Li et al., 2024 ; Coleman et al., 2024 ), this approach can be constrained by factors such as species differences in enhancer activity and the lack of well-characterized reagents in D. melanogaster . Unlike enhancer-based genetic tools, gene-specific tools that leverage native regulatory elements have been shown to reliably recapitulate the expression patterns of targeted genes ( Diao et al., 2015 ; Lee et al., 2018 ; Chen et al., 2023 ). By identifying molecular markers and marker combinations that represent specific cell types, our scRNA-seq datasets provide a roadmap for designing gene-specific genetic tools that target individual cell types within and across species. Finally, we have uncovered many potential evolutionary changes, including alterations in cell type abundance, cell types exhibiting pronounced transcriptomic divergence between species, and genes with species-specific expression patterns, inviting future validation and functional characterization. As a case of validation, we demonstrated that the single-neuron cell type TN2-5 has undergone an evolutionary switch in neuropeptide identity—expressing Tk in D. melanogaster and expressing Dh31 in D. yakuba . While the function of TN2-5 is currently uncharacterized, its male-specificity and mesothoracic location suggest that it might be involved in generating male-specific wing behavior, such as courtship song and agnostic song in aggression ( Jonsson et al., 2011 ). Interestingly, in D. melanogaster , Tk signaling suppresses courtship and enhances aggression ( Asahina et al., 2014 ; Shankar et al., 2015 ); while gut - derived Dh31 signaling promotes courtship ( Lin et al., 2022 ). Future research may uncover the behavioral role of TN2-5 and the phenotypic consequences of this evolutionary switch in neuropeptide identity, now made possible by identifying molecular markers for cell-type-specific manipulation. Given the widespread evolutionary turnover in the expression of neuropeptide signaling genes, a broadly relevant question is whether the behavioral states associated with these genes have diverged across species or if the gene-behavioral state associations remain conserved while evolutionary changes in gene expression have repurposed the same cell type to associate with a distinct behavioral state. In conclusion, using dsx as a molecular handle, we characterized the cellular and molecular diversity of sexual circuits with unprecedented detail both within and across species. Our findings on the conservation and divergence of cell types and gene expression provide a crucial foundation for understanding the fundamental principles governing neural circuit organization and evolution in behavioral adaptations. Methods Fly strains and husbandry We housed all flies on a cornmeal-agar-yeast medium (Fly Food B, Bloomington Recipe, Lab Express) in a humidified incubator at 23°C on a 12 hour light and dark cycle. We list the genotypes of fly lines used in this study in Supplemental Table 5 . Generation of genetic reagents We digested the plasmid pBac {3XP3::EYFP, attP} ( Stern et al., 2017 ) with NaeI and inserted a DNA cassette containing 10XUAS-nls::tdTomato-SV40, PCR amplified from the plasmid pJFRC105-UAS-nls::tdTomato (gift of Gerry Rubin’s lab, Janelia) into the NaeI-digested pBac {3XP3::EYFP, attP} by Gibson Assembly. We injected the resulting plasmid into D. teissieri wild-type flies (DSSC # 14021-0257.01) to generate the transgenic line pBac {UAS-nls::tdTomato, 3XP3::YFP}. We also injected the plasmid pJFRC105-10XUAS-nls::tdTomato into the D. santomea line 2253 ( Stern et al., 2017 ) to create the transgenic line w; ; pJFRC105-10XUAS-nls::tdTomato (2283). Dissection, cell dissociation, library preparation, and sequencing We genetically labeled dsx + neurons by crossing females carrying dsx -GAL4 (GAL4 was introduced into the endogenous dsx locus by CRISPR/Cas9-mediated genome editing) ( Ye et al., 2024 ) with males carrying UAS-nls::tdTomato. The resultant dsx- GAL4>UAS-nls::tdTomato males were sorted within 12 hours of eclosion and housed in groups of 15–20 until dissection at 4–8 days old. We dissected the CNS (brain and ventral nerve cord) from approximately 100 flies per sample. Our cell dissociation method was adapted from previously described protocols ( Li et al., 2022 ). In short, we dissected the flies in ice-cold Schneider’s medium for one hour before transferring all tissue into 500 ul of dispase (1mg/1ml Liberase DH dissolved in Schneider’s medium). We digested the tissue in 500 µL dispase (1 mg/mL Liberase DH in Schneider’s medium) for 45 minutes at room temperature, then washed it three times with ice-cold Schneider’s medium. We dissociated the tissue in 500 µL ice-cold 1X PBS with 2% BSA by pipetting it 100 times with a pipette and 20 times with a 25-gauge needle, both coated with FBS to prevent cell adhesion. We transferred the solution into a flow cytometry tube before sorting out nls::tdTomato+ ( dsx + neurons) cells with a FACS Aria Cell Sorter (using a 100 µm nozzle and 1.25 drop precision). To verify the yield, we counted the number of nls::tdTomato+ cells using a disposable hemocytometer and a Leica M165 fluorescent scope. Upon confirmation, we transported the cells to the Center for Applied Genomics at the Children’s Hospital of Philadelphia’s Research Institute where all library preparation and sequencing was performed using the 10x Genomics Single Cell 3’ Library and Gel Bead Kit following the manufacturer’s recommendation protocol and the libraries were sequenced on an Illumina NovaSeq 6000. The amount of time between the start of cell dissociation and the start of cell sorting was about 70 min, sorting the cells lasted about 30 min, and the time between the end of cell sorting and the start of the 10x Genomics protocol was about 30 min. We prepared two replicates for both D. melanogaster and D. yakuba (in addition to using one published replicate per species from Ye et al., 2024 ) and one sample each for D. teissieri and D. santomea . scRNA-seq data mapping and filtering We used the CellRanger software (v7.1.0) “count” command to map sequencing reads to Drosophila melanogaster (Assembly GCA_000001215.4), Drosophila yakuba (Assembly GCA_016746365.2), Drosophila santomea (GCA_016746245.2), and Drosophila teissieri (GCA_016746235.2) genomes accordingly. We analyzed expression matrices with Seurat v5.0.1 ( Hao et al., 2024 ) and applied default parameters for all Seurat functions unless otherwise specified. To ensure high-quality single-cell data, we filtered the datasets by removing cells with greater than 15% mitochondrial transcripts, fewer than 200 genes detected, or expressing more than 4000 unique genes. To identify orthologous genes between species, we performed three reciprocal best hits BLAST analyses, pairing D. yakuba , D. santomea , and D. teissieri with D. melanogaster using default settings ( Camacho et al., 2009 ). We identified 11,739 orthologous genes shared among all four species and retained only these genes in the datasets, including the dataset of D. melanogaster males and females (described below). scRNA-seq data clustering and integration We used the functions “NormalizeData” and “ScaleData” to normalize and scale the individual species’ datasets, followed by principal components analyses (PCA). We determined the number of statistically significant principal components using Jackstraw analysis. We used the functions “runUMAP”, “FindNeighbors”, and “FindClusters” to reduce the dimensionality and perform clustering. We assessed clustering quality using the R package scclustereval v0.0.0.9 ( Tang et al., 2020 ). After clustering, we removed clusters where the majority of cells lacked expression of dsx or elav (a neuronal marker gene), retaining a total of 17,168 D. melanogaster , 18,067 D. yakuba , 6,344 D. teissieri , and 5,343 D. santomea cells. See Supplemental Table 1 for a summary of all sequencing and filtering data. We initially analyzed each species data separately to confirm the presence of all nine parental clusters in each species (see below). After identifying the nine parental clusters in each species, we integrated the datasets across species using the Seurat functions “SelectIntegrationFeatures”, “FindIntegrationAnchors”, and “IntegrateData”. Following integration, we performed PCA and Jackstraw analysis, reduced dimensionality, and conducted clustering. We confirmed the presence of all nine parental clusters in the integrated dataset and extracted each parental cluster into a separate Seurat fil e for further clustering analyses using the same protocol described above. Parental clusters annotations To identify each parental cluster, we first classified neurons as either from the brain or the VNC based on the expression of tsh (a VNC marker gene; Röder et al., 1992 ) and the four Hox genes Antennapedia ( Antp ), Ultrabithorax ( Ubx ), abdominal-A ( abd-A ), and Abdominal-B ( Abd-B ), which follow an anterior-to-posterior expression pattern in the VNC ( Allen et al., 2020 ). In the brain, we considered the expected size for each parental cluster and identified good marker genes for clusters defined in UMAP space. We then used split-GAL4, pairing a dsx hemi-driver with a hemi-driver for each marker gene, to drive the expression of fluorescent proteins (GFP or tdTomato) in D. melanogaster . We performed immunostaining against the fluorescent proteins to determine the dsx + neurons labeled by each marker gene. These maker genes included tey , Optix , TfAP-2 , sv, unc-4, Gad1 , whose intersection with dsx labeled aDN, pC1, pC2l, pC2m, pCd-1, and pCd-2, respectively. In the VNC, the marker gene TfAP-2 labeled TN1 neurons ( Ye, et al., 2024 ). We further identified neurons that were Antp + and/or Ubx + but abd-A - and Abd-B -, hypothesizing that these were TN2 given that neurons in the ANm should express one or both of abd-A and Abd-B . A split-GAL4 intersection between ara –a marker gene for these neurons–and dsx confirmed their identity as TN2. We scored all VNC clusters that were not TN1 or TN2 but expressed abd-A and/or Abd-B as ANm. Immunostaining was performed as previously described ( Ye et al., 2024 ) and imaged on a Leica 350 DMi8 microscope with a TCS SP8 Confocal System at 40x with optical sections at 0.4-0.8 µm intervals. All the genes mentioned above were shared marker genes identified across independent clustering analyses for each of the four species, permitting us to transfer parental cluster identities from D. melanogaster to D. yakuba , D. santomea , and D. teissieri . Beyond relying on these shared marker genes, we further validated the transferability of our annotations by measuring the similarity of parental clusters across species using MetaNeighbor ( Crow et al., 2018 ) to calculate the mean AUROC scores for each pairwise comparison. Identification of marker genes To identify marker genes for all parental clusters and subclusters, we used the Seurat function “FindAllMarkers” with the “assay” parameter set to “RNA”. This function employs Wilcoxon signed-rank tests to identify marker genes. We performed this analysis for all four species combined as well as for each species individually. Designating parental cluster- or subcluster-level gene expression We defined a gene as expressed in a given subcluster if the scMarco ( Chen et al., 2023 ) bimodal score for that gene in the subcluster exceeded 0.5. Using this criterion, we annotated hemilineage markers, fru , neurotransmitter markers, neuropeptides, and neuropeptide receptors. We also quantified the number of genes expressed in each subcluster and identified genes to include in our DEG analyses (see below). Predicting hemilineage identity We used previously identified hemilineage markers ( Bossing et al., 1996 ; Schmid et al., 1999 ; Lacin et al., 2019 ; Allen et al., 2020 ) to annotate the hemilineage origins of ANm subclusters. We considered a marker gene expressed in a given subcluster if the scMarco bimodal score exceeded 0.5 for D. melanogaster cells within the integrated dataset. See Supplemental Table 2 for a full list of hemilineage annotations and the marker genes used. Defining species differences in cell number within subclusters between D. melanogaster and D. yakuba For each species, we manually quantified the number of cells in each dsx+ parental cluster in dsx -GAL4>UAS-myrGFP males and females by immunostaining as previously described ( Ye et al., 2024 ). We counted cells from confocal images using the Fiji/ImageJ plugin “Cell Counter.” We excluded the ANm cluster due to the high density of dsx+ neurons, which made precise counting challenging, and therefore assumed equal numbers of ANm dsx + cells in species comparison. Using these data, we scaled the cell count for each cell type to align the size of each parental cluster with immunostaining measurements. After scaling, we applied a Bayesian modeling approach with the Python package scCODA v0.1.2 ( Büttner et al., 2021 ) and default parameters to identify species-biased subclusters. Spearman rank correlation analysis across species To assess transcriptomic conservation between D. melanogaster and D. yakuba subclusters, we calculated Spearman rank correlations based on the median expression levels of genes with non-zero median expression. We performed this analysis using three gene sets: all genes, all marker genes, and transcription factors. For each subcluster, we also conducted 1,000 iterations of randomly downsampling the gene set to 1,000 genes (selected from genes with non-zero median expression in at least one species) and calculated the Spearman rank correlation between D. melanogaster and D. yakuba . Additionally, we computed the Spearman rank correlation for each subcluster between replicates within each species, considering all genes with non-zero median expression values. When grouping Spearman rank correlation scores by parental cluster, we found that although the data were normally distributed (Shapiro-Wilk, W = 0.985, p = 0.462), the variance across parental clusters was not homogenous (Bartlett test; K 2 = 16.482; p = 0.0356). Therefore, we conducted a nonparametric Kruskal Wallis test. Furthermore, we conducted pairwise Bartlett tests across parental clusters but found no significant differences after correcting for multiple comparisons via a Bonferroni adjustment. DEG analysis between D. melanogaster and D. yakuba To identify differentially expressed genes (DEGs) between D. melanogaster and D. yakuba , we used the Seurat function “FindAllMarkers” with the “assay” parameter set to “RNA” and the active Ident set to species after subsetting out only the D. melanogaster and D. yakuba cells from the full integrated dataset. For each subcluster, we included genes expressed in at least one of the two species defined by scMarco to perform DEG analysis. For each subcluster, we compared the expression of each gene in D. melanogaster cells to its expression in D. yakuba cells. To ensure that we took a relatively conservative approach, we classified genes as differentially expressed only if they met a log fold change threshold greater than three and had Bonferroni-corrected p -values less than 0.05. We performed Gene Ontology (GO) term enrichment analysis using PANTHER (Mi et al., 2015), looking for enriched “biological process” terms among the DEGs against all genes e expressed in at least one subcluster. Validating differential gene expression of Tk and Dh31 in TN2-5 neurons between species To determine the identity of TN2-5 neurons, we performed immunostaining against Antp (DHSB 8C11, 1:20) and Ubx (DHSB FP3.38, 1:20) in dsx -GAL4>UAS-myrGFP male flies (anti-GFP, Abcam #13970, 1:600). We identified the two prothoracic TN2 neurons as Antp- and Ubx- , corresponding to TN2-3 and TN2-4; the two metathoracic ones as Antp - and Ubx +, corresponding to TN2-1 and TN2-6; and the two mesothoracic ones as Antp + and Ubx -, corresponding to TN2-2 and TN2-5. To validate the DEG genes Tk and Dh31 , we performed multiplexed HCR with three probe sets detecting dsx , Tk , and Dh31 mRNA in male flies. For Tk and Dh31 , D. melanogaster sequences were provided for probe design (Molecular Instrument), with the request of designing probes against conserved gene regions. The same probe set was used for D. melanogaster and D. yakuba , producing robust and unambiguous signals, albeit a bit weaker in D. yakuba . HCR was performed following a previously described protocol ( Duckhorn et al., 2022 ). See Supplementary Table 5 for probe and hairpin information. Generating scRNA-seq data of dsx + neurons in female D. melanogaster To prepare the female dataset, we followed the same dissection, cell dissociation, library preparation, sequencing, data mapping, and filtering protocols as described above, resulting in 3,802 high quality dsx + cells. Because males have approximately 2.5 times the number of dsx + neurons as females, we randomly downsampled the male D. melanogaster dataset to 2.5 times the number of female cells (9,502 cells). We then integrated the downsampled male dataset with the female dataset and performed clustering usi ng the protocol described above. Using the same marker genes as before, we identified all nine parental clusters in the male/female integrated dataset. We extracted each parental cluster individually for subclustering. To match subclusters between our two datasets (four-species versus male/female D. melanogaster ), we cross-referenced the subcluster identities of male cells present in both datasets using their unique barcode IDs. We considered subclusters to have a one-to-one match if over 80% of the cells from a given subcluster in the male/female dataset aligned to a single subcluster from the four-species dataset. When subclusters did not have a clear one-to-one match, we considered subclusters from the male/female dataset to contain all subclusters from the four-species dataset with over 10% matching cells. See Supplemental Table 4 . Defining sex differences in cell number Since lacked biological replicates for female D. melanogaster , we could not use scCODA to identify sex-specific or sex-biased subclusters. Instead, we defined sex-specific subclusters as those consisting of over 95% cells from one sex and sex-biased subclusters as those with over 75% cells from one sex. Using marker gene lists to identify genes that label subclusters For both the four-species and the male/female datasets, we report pairs of genes that uniquely label each subcluster. Although we report all pairs of genes with clarity scores of 1, we encourage users to be selective when choosing genes by using genes with the highest possible percentage of cells expressing that gene and checking gene expression patterns in scMarco. Author contribution J.T.W and Y.D. conceived the study and wrote the paper. I.P.J. and H.G. performed in situ HCR and immunostaining and collected the confocal imaging data. J.T.W. generated the scRNA-seq data and performed analyses. Y.D. generated genetic reagents. Y.-C.D.C. and Y.-C.C. generated the resources for cell type-specific marker combinations. Acknowledgement We thank Lihua Li and Hongjie Li for their valuable input during the development of our scRNA-seq pipeline. We thank Troy Shirangi’s lab for contributing to sample dissection. We thank Nancy Zhang for valuable feedback on statistical analysis. We are immensely grateful for the generosity of the fly research community. Particularly, we thank Haluk Lacin, Yufeng Pan, Chris Doe, Gary Struhl, Joshua Kavaler, Nobert Perrimon, Ben Ewen-Campen, Ruth Lehmann, and Jing Wang for generously sharing genetic reagents and antibodies of marker genes. Although many experiments using these resources were not directly presented here, they have greatly informed our initial identification of dsx + parental clusters before we took a split-GAL4 strategy for final validation. We also thank Rory Coleman, Troy Shirangi, Dawn Chen, Stephen Goodwin, Megan Goodwin, and Aaron Allen for helpful comments on the manuscript. We thank Jennifer Jakubowski, Shifu Tian, and Andrew Morschauser for their help with FACS. We thank Molly Gallagher and Diana Slater at the Center for Applied Genomics at the Children’s Hospital of Philadelphia’s Research Institute for help with library preparation and sequencing. Y.D. was supported by NIH grant R35GM142678. Y.-C.C. was supported by New York University (MacCracken Fellowship), by a NYSTEM institutional training grant (Contract #C322560GG), and by a Scholarship to Study Abroad from the Ministry of Education, Taiwan. Y.-C.D.C. is supported by the NIH National Eye Institute (K99EY035757). Funder Information Declared NIH , R35GM142678 NIH National Eye Institute , K99EY035757 References ↵ Allen A.M. , Neville M.C. , Birtles S. , Croset V. , Treiber C.D. , Waddell S. & Goodwin S.F . ( 2020 ) A single-cell transcriptomic atlas of the adult Drosophila ventral nerve cord . Elife 9 , e54074 . OpenUrl CrossRef PubMed ↵ Arnqvist G. & Rowe L . ( 2005 ) Sexual conflict . Princeton university press . ↵ Asahina K. , Watanabe K. , Duistermars B.J. , Hoopfer E. , González C.R. , Eyjólfsdóttir E.A. , Perona P. & Anderson D.J . ( 2014 ) Tachykinin-expressing neurons control male-specific aggressive arousal in Drosophila . Cell 156 , 221 – 35 . OpenUrl CrossRef PubMed ↵ Auer T.O. , Álvarez-Ocaña R. , Cruchet S. , Benton R. & Arguello J.R . ( 2022 ) Copy number changes in co-expressed odorant receptor genes enable selection for sensory differences in drosophilid species . Nature Ecology & Evolution 6 , 1343 – 53 . OpenUrl CrossRef PubMed ↵ Auer T.O. , Khallaf M.A. , Silbering A.F. , Zappia G. , Ellis K. , Álvarez-Ocaña R. , Arguello J.R. , Hansson B.S. , Jefferis G. , Caron S.J.C. , Knaden M. & Benton R . ( 2020 ) Olfactory receptor and circuit evolution promote host specialization . Nature 579 , 402 – 8 . OpenUrl CrossRef PubMed ↵ Bachtrog D. , Thornton K. , Clark A. & Andolfatto P . ( 2006 ) Extensive introgression of mitochondrial DNA relative to nuclear genes in the Drosophila yakuba species group . Evolution 60 , 292 – 302 . OpenUrl CrossRef PubMed Web of Science ↵ Billeter J.-C. , Atallah J. , Krupp J.J. , Millar J.G. & Levine J.D . ( 2009 ) Specialized cells tag sexual and species identity in Drosophila melanogaster . Nature 461 , 987 – 91 . OpenUrl CrossRef PubMed Web of Science ↵ Bossing T. , Udolph G. , Doe C.Q. & Technau G.M . ( 1996 ) The embryonic central nervous system lineages of Drosophila melanogaster : I. Neuroblast lineages derived from the ventral half of the neuroectoderm . Developmental Biology 179 , 41 – 64 . OpenUrl CrossRef PubMed Web of Science ↵ Büttner M. , Ostner J. , Müller C.L. , Theis F.J. & Schubert B . ( 2021 ) scCODA is a Bayesian model for compositional single-cell data analysis . Nature Communications 12 , 6876 . OpenUrl CrossRef PubMed ↵ Cachero S. , Ostrovsky A.D. , Yu J.Y. , Dickson B.J. & Jefferis G.S.X.E . ( 2010 ) Sexual dimorphism in the fly brain . Current Biology 20 , 1589 – 601 . OpenUrl CrossRef PubMed Web of Science ↵ Caglayan E. , Liu Y. & Konopka G . ( 2022 ) Neuronal ambient RNA contamination causes misinterpreted and masked cell types in brain single-nuclei datasets . Neuron 110 , 4043 – 56 .e5. OpenUrl CrossRef PubMed ↵ Camacho C. , Coulouris G. , Avagyan V. , Ma N. , Papadopoulos J. , Bealer K. & Madden T.L . ( 2009 ) BLAST+: architecture and applications . BMC Bioinformatics 10 , 421 . OpenUrl CrossRef PubMed ↵ Castellanos M.C. , Tang J.C. & Allan D.W . ( 2013 ) Female-biased dimorphism underlies a female-specific role for post-embryonic Ilp7 neurons in Drosophila fertility . Development 140 , 3915 – 26 . OpenUrl Abstract / FREE Full Text ↵ Certel S.J. , Savella M.G. , Schlegel D.C. & Kravitz E.A . ( 2007 ) Modulation of Drosophila male behavioral choice . Proceedings of the National Academy of Sciences 104 , 4706 – 11 . OpenUrl Abstract / FREE Full Text ↵ Chapman T. , Arnqvist G. , Bangham J. & Rowe L . ( 2003 ) Sexual conflict . Trends in Ecology & Evolution 18 , 41 – 7 . OpenUrl CrossRef ↵ Chen Y.-C.D. , Chen Y.-C. , Rajesh R. , Shoji N. , Jacy M. , Lacin H. , Erclik T. & Desplan C . ( 2023 ) Using single-cell RNA sequencing to generate predictive cell-type-specific split-GAL4 reagents throughout development . Proceedings of the National Academy of Sciences 120 , e2307451120 . OpenUrl CrossRef PubMed ↵ Choi H.M.T. , Schwarzkopf M. , Fornace M.E. , Acharya A. , Artavanis G. , Stegmaier J. , Cunha A. & Pierce N.A . ( 2018 ) Third-generation in situ hybridization chain reaction: multiplexed, quantitative, sensitive, versatile, robust . Development 145 . ↵ Coleman R.T. , Morantte I. , Koreman G.T. , Cheng M.L. , Ding Y. & Ruta V . ( 2024 ) A modular circuit coordinates the diversification of courtship strategies . Nature 635 , 142 – 150 OpenUrl CrossRef PubMed ↵ Cox R.M. & Calsbeek R . ( 2009 ) Sexually antagonistic selection, sexual dimorphism, and the resolution of intralocus sexual conflict . The American Naturalist 173 , 176 – 87 . OpenUrl CrossRef PubMed Web of Science ↵ Crow M. , Paul A. , Ballouz S. , Huang Z.J. & Gillis J . ( 2018 ) Characterizing the replicability of cell types defined by single cell RNA-sequencing data using MetaNeighbor . Nature Communications 9 , 884 . OpenUrl CrossRef PubMed ↵ Davis F.P. , Nern A. , Picard S. , Reiser M.B. , Rubin G.M. , Eddy S.R. & Henry G.L . ( 2020 ) A genetic, genomic, and computational resource for exploring neural circuit function . Elife 9 , e50901 . OpenUrl CrossRef PubMed ↵ Deutsch D. , Pacheco D. , Encarnacion-Rivera L. , Pereira T. , Fathy R. , Clemens J. , Girardin C. , Calhoun A. , Ireland E. , Burke A. , Dorkenwald S. , McKellar C. , Macrina T. , Lu R. , Lee K. , Kemnitz N. , Ih D. , Castro M. , Halageri A. , Jordan C. , Silversmith W. , Wu J. , Seung H.S. & Murthy M . ( 2020 ) The neural basis for a persistent internal state in Drosophila females . Elife 9 . ↵ Diao F. , Ironfield H. , Luan H. , Diao F. , Shropshire W.C. , Ewer J. , Marr E. , Potter C.J. , Landgraf M. & White B.H . ( 2015 ) Plug-and-play genetic access to Drosophila cell types using exchangeable exon cassettes . Cell Reports 10 , 1410 – 21 . OpenUrl CrossRef PubMed ↵ Ding Y. , Lillvis J.L. , Cande J. , Berman G.J. , Arthur B.J. , Long X. , Xu M. , Dickson B.J. & Stern D.L . ( 2019 ) Neural evolution of context-dependent fly song . Current Biology 29 , 1089 – 99 .e7. OpenUrl CrossRef PubMed ↵ Yamamoto D Duckhorn J.C. , Junker I.P. , Ding Y. & Shirangi T.R. ( 2022 ) Combined in situ hybridization chain reaction and immunostaining to visualize gene expression in whole-mount Drosophila central nervous systems . In: Behavioral Neurogenetics (ed. by Yamamoto D ), pp. 1 – 14 . Springer US , New York, NY . ↵ Ewing A.W. & Bennet-Clark H.C . ( 1968 ) The courtship songs of Drosophila . Behaviour 31 , 288 – 301 . OpenUrl CrossRef ↵ Ferveur J.-F . ( 2005 ) Cuticular hydrocarbons: their evolution and roles in Drosophila pheromonal communication . Behavior genetics 35 , 279 – 95 . OpenUrl CrossRef PubMed Web of Science ↵ Geirsdottir L. , David E. , Keren-Shaul H. , Weiner A. , Bohlen S.C. , Neuber J. , Balic A. , Giladi A. , Sheban F. , Dutertre C.-A. , Pfeifle C. , Peri F. , Raffo-Romero A. , Vizioli J. , Matiasek K. , Scheiwe C. , Meckel S. , Mätz-Rensing K. , van der Meer F. , Thormodson F.R. , et al. , ( 2019 ) Cross-species single-cell analysis reveals divergence of the primate microglia program . Cell 179 , 1609 – 22 .e16. OpenUrl CrossRef PubMed ↵ Gleason J.M. , Pierce A.A. , Vezeau A.L. & Goodman S.F . ( 2012 ) Different sensory modalities are required for successful courtship in two species of the Drosophila willistoni group . Animal Behaviour 83 , 217 – 27 . OpenUrl CrossRef ↵ Gleason J.M. & Ritchie M.G . ( 2004 ) Do quantitative trait loci (QTL) for a courtship song difference between Drosophila simulans and D. sechellia coincide with candidate genes and intraspecific QTL? Genetics 166 , 1303 – 11 . OpenUrl Abstract / FREE Full Text ↵ Goodwin S.F. & Hobert O . ( 2021 ) Molecular mechanisms of sexually dimorphic nervous system patterning in flies and worms . Annual Review of Cell and Developmental Biology 37 , 519 – 47 . OpenUrl CrossRef PubMed ↵ Greenspan R.J. & Ferveur J.F . ( 2000 ) Courtship in Drosophila . Annual Review of Genetics 34 , 205 – 32 . OpenUrl CrossRef PubMed Web of Science ↵ Greer C.L. , Grygoruk A. , Patton D.E. , Ley B. , Romero-Calderon R ., Chang H.Y. , Houshyar R. , Bainton R.J. , DiAntonio A. & Krantz D.E. ( 2005 ) A splice variant of the Drosophila vesicular monoamine transporter contains a conserved trafficking domain and functions in the storage of dopamine, serotonin, and octopamine . Journal of neurobiology 64 , 239 – 58 . OpenUrl CrossRef PubMed Web of Science ↵ Hao Y. , Stuart T. , Kowalski M.H. , Choudhary S. , Hoffman P. , Hartman A. , Srivastava A. , Molla G. , Madad S. , Fernandez-Granda C. & Satija R . ( 2024 ) Dictionary learning for integrative, multimodal and scalable single-cell analysis . Nature Biotechnology 42 , 293 – 304 . OpenUrl CrossRef PubMed ↵ Imoto K. , Ishikawa Y. , Aso Y. , Funke J. , Tanaka R. & Kamikouchi A . ( 2024 ) Neural-circuit basis of song preference learning in fruit flies . iScience 27 , 110266 . OpenUrl CrossRef PubMed ↵ Jiang X. , Sun M. , Chen J. & Pan Y . ( 2024 ) Sex-Specific and state-dependent neuromodulation regulates male and female locomotion and sexual behaviors . Research 7 , 0321 . OpenUrl CrossRef PubMed ↵ Jois S. , Chan Y.B. , Fernandez M.P. & Leung A.K . ( 2018 ) Characterization of the sexually dimorphic fruitless neurons that regulate copulation duration . Frontiers in Physiology 9 , 780 . OpenUrl CrossRef PubMed ↵ Jonsson T. , Kravitz E.A. & Heinrich R . ( 2011 ) Sound production during agonistic behavior of male Drosophila melanogaster . Fly 5 , 29 – 38 . OpenUrl CrossRef PubMed ↵ Katz P.S . ( 2011 ) Neural mechanisms underlying the evolvability of behaviour . Philosophical Transactions of the Royal Society B: Biological Sciences 366 , 2086 – 99 . OpenUrl CrossRef PubMed ↵ Katz P.S. & Harris-Warrick R.M . ( 1999 ) The evolution of neuronal circuits underlying species-specific behavior . Current Opinion in Neurobiology 9 , 628 – 33 . OpenUrl CrossRef PubMed Web of Science ↵ Khallaf M.A. , Cui R. , Weißflog J. , Erdogmus M. , Svatoš A. , Dweck H.K.M. , Valenzano D.R. , Hansson B.S. & Knaden M . ( 2021 ) Large-scale characterization of sex pheromone communication systems in Drosophila . Nature Communications 12 , 4165 . OpenUrl CrossRef PubMed ↵ Khrameeva E. , Kurochkin I. , Han D. , Guijarro P. , Kanton S. , Santel M. , Qian Z. , Rong S. , Mazin P. , Sabirov M. , Bulat M. , Efimova O. , Tkachev A. , Guo S. , Sherwood C.C. , Camp J.G. , Pääbo S. , Treutlein B. & Khaitovich P . ( 2020 ) Single-cell-resolution transcriptome map of human, chimpanzee, bonobo, and macaque brains . Genome Researc 30 , 776 – 89 . OpenUrl CrossRef ↵ Kijimoto T. , Moczek A.P. & Andrews J . ( 2012 ) Diversification of doublesex function underlies morph-, sex-, and species-specific development of beetle horns . Proceedings of the National Academy of Sciences 109 , 20526 – 31 . OpenUrl Abstract / FREE Full Text ↵ Kim D.-H. , Jang Y.-H. , Yun M. , Lee K.-M. & Kim Y.-J . ( 2024 ) Long-term neuropeptide modulation of female sexual drive via the TRP channel in Drosophila melanogaster . Proceedings of the National Academy of Sciences 121 , e2310841121 . OpenUrl CrossRef PubMed ↵ Kim T. , Chen I.R. , Lin Y. , Wang A.Y.-Y. , Yang J.Y.H. & Yang P . ( 2018 ) Impact of similarity metrics on single-cell RNA-seq data clustering . Briefings in Bioinformatics 20 , 2316 – 26 . OpenUrl Kimura K.-i ., Hachiya T. , Koganezawa M. , Tazawa T. & Yamamoto D. ( 2008a ) Fruitless and doublesex coordinate to generate male-specific neurons that can initiate courtship . Neuron 59 , 759 – 69 . OpenUrl CrossRef PubMed Web of Science ↵ Kimura K.-I. , Ote M. , Tazawa T. & Yamamoto D . ( 2005 ) Fruitless specifies sexually dimorphic neural circuitry in the Drosophila brain . Nature 438 , 229 – 33 . OpenUrl CrossRef PubMed Web of Science Kimura K. , Hachiya T. , Koganezawa M. , Tazawa T. & Yamamoto D . ( 2008b ) Fruitless and doublesex coordinate to generate male-specific neurons that can initiate courtship . Neuron 59 , 759 – 69 . OpenUrl CrossRef PubMed Web of Science ↵ Koganezawa M. , Kimura K. & Yamamoto D . ( 2016 ) The neural circuitry that functions as a switch for courtship versus aggression in Drosophila males . Current Biology 26 , 1395 – 403 . OpenUrl CrossRef PubMed ↵ Kunte K. , Zhang W. , Tenger-Trolander A. , Palmer D. , Martin A. , Reed R. , Mullen S. & Kronforst M . ( 2014 ) doublesex is a mimicry supergene . Nature 507 , 229 – 32 . OpenUrl CrossRef PubMed Web of Science ↵ Lacin H. , Chen H.-M. , Long X. , Singer R.H. , Lee T. & Truman J.W . ( 2019 ) Neurotransmitter identity is acquired in a lineage-restricted manner in the Drosophila CNS . Elife 8 , e43701 . OpenUrl CrossRef PubMed ↵ Lagisz M. , Wen S.Y. , Routtu J. , Klappert K. , Mazzi D. , Morales-Hojas R. , Schäfer M.A. , Vieira J. , Hoikkala A. , Ritchie M.G. & Butlin R.K . ( 2012 ) Two distinct genomic regions, harbouring the period and fruitless genes, affect male courtship song in Drosophila montana . Heredity 108 , 602 – 8 . OpenUrl CrossRef PubMed ↵ Lee D. & Benton R . ( 2023 ) Comparative single-cell transcriptomic atlases reveal conserved and divergent features of drosophilid central brains . bioRxiv , 2023.11.22.568274. ↵ Lee G. & Hall J.C . ( 2001 ) Abnormalities of male-specific FRU protein and serotonin expression in the CNS of fruitless mutants in Drosophila . Journal of Neuroscience 21 , 513 – 26 . OpenUrl Abstract / FREE Full Text ↵ Lee G. , Hall J.C. & Park J.H . ( 2002 ) Doublesex gene expression in the central nervous system of Drosophila melanogaster . Journal of Neurogenetics 16 , 229 – 48 . OpenUrl CrossRef PubMed Web of Science ↵ Lee K.-M. , Daubnerová I. , Isaac R.E. , Zhang C. , Choi S. , Chung J. & Kim Y.-J . ( 2015 ) A neuronal pathway that controls sperm ejection and storage in female Drosophila . Current Biology 25 , 790 – 7 . OpenUrl CrossRef PubMed ↵ Lee P.-T. , Zirin J. , Kanca O. , Lin W.-W. , Schulze K.L. , Li-Kroeger D. , Tao R. , Devereaux C. , Hu Y. , Chung V. , Fang Y. , He Y. , Pan H. , Ge M. , Zuo Z. , Housden B.E. , Mohr S.E. , Yamamoto S. , Levis R.W. , Spradling A.C. , Perrimon N. & Bellen H.J . ( 2018 ) A gene-specific T2A-GAL4 library for Drosophila . Elife 7 , e35574 . OpenUrl CrossRef PubMed ↵ Lessells C.M . ( 2006 ) The evolutionary outcome of sexual conflict . Philosophical Transactions of the Royal Society B: Biological Sciences 361 , 301 – 17 . OpenUrl CrossRef PubMed ↵ Li H. , Janssens J. , De Waegeneer M. , Kolluru S.S. , Davie K. , Gardeux V. , Saelens W. , David F.P.A. , Brbić M. , Spanier K. , Leskovec J. , McLaughlin C.N. , Xie Q. , Jones R.C. , Brueckner K. , Shim J. , Tattikota S.G. , Schnorrer F. , Rust K. , Nystul T.G. , et al. , ( 2022 ) Fly Cell Atlas: A single-nucleus transcriptomic atlas of the adult fruit fly . Science 375 , eabk2432 . OpenUrl CrossRef PubMed ↵ Li M. , Chen D.S. , Junker I.P. , Szorenyi F.I. , Chen G.H. , Berger A.J. , Comeault A.A. , Matute D.R. & Ding Y . ( 2024 ) Ancestral neural circuits potentiate the origin of a female sexual behavior in Drosophila . Nature Communications 15 , 9210 . OpenUrl CrossRef PubMed ↵ Li S.A. , Li H.G. , Shoji N. , Desplan C. & Chen Y.D . ( 2023 ) Protocol for replacing coding intronic MiMIC and CRIMIC lines with T2A-split-GAL4 lines in Drosophila using genetic crosses . STAR Protocols 4 , 102706 . OpenUrl CrossRef PubMed ↵ Lillvis J.L. , Wang K. , Shiozaki H.M. , Xu M. , Stern D.L. & Dickson B.J . ( 2024 ) Nested neural circuits generate distinct acoustic signals during Drosophila courtship . Current Biology 34 , 808 – 24 .e6. OpenUrl CrossRef PubMed ↵ Lin H.-H. , Kuang M.C. , Hossain I. , Xuan Y. , Beebe L. , Shepherd A.K. , Rolandi M. & Wang J.W . ( 2022 ) A nutrient-specific gut hormone arbitrates between courtship and feeding . Nature 602 , 632 – 8 . OpenUrl CrossRef PubMed ↵ Liu W. , Ganguly A. , Huang J. , Wang Y. , Ni J.D. , Gurav A.S. , Aguilar M.A. & Montell C . ( 2019 ) Neuropeptide F regulates courtship in Drosophila through a male-specific neuronal circuit . Elife 8 , e49574 . OpenUrl CrossRef PubMed ↵ Luan H. , Diao F. , Scott R.L. & White B.H . ( 2020 ) The Drosophila split Gal4 system for neural circuit mapping . Frontiers in Neural Circuits 14 . Lun A.T.L. , Riesenfeld S. , Andrews T. , Dao T.P. , Gomes T. , Marioni J.C. & participants in the 1st Human Cell Atlas J. (2019) EmptyDrops: distinguishing cells from empty droplets in droplet-based single-cell RNA sequencing data . Genome Biology 20 , 63 . ↵ Luo S.D. & Baker B.S . ( 2015 ) Constraints on the evolution of a doublesex target gene arising from doublesex’s pleiotropic deployment . Proceedings of the National Academy of Sciences 112 , E852 – E61 . OpenUrl Abstract / FREE Full Text ↵ Macosko E.Z. , Basu A. , Satija R. , Nemesh J. , Shekhar K. , Goldman M. , Tirosh I. , Bialas A.R. , Kamitaki N. , Martersteck E.M. , Trombetta J.J. , Weitz D.A. , Sanes J.R. , Shalek A.K. , Regev A. & McCarroll S.A. ( 2015 ) Highly parallel genome-wide expression profiling of individual cells using nanoliter droplets . Cell 161 , 1202 – 14 . OpenUrl CrossRef PubMed ↵ Mezzera C. , Brotas M. , Gaspar M. , Pavlou H.J. , Goodwin S.F. & Vasconcelos M.L . ( 2020 ) Ovipositor extrusion promotes the transition from courtship to copulation and signals female acceptance in Drosophila melanogaster . Current Biology 30 , 3736 – 48 .e5. OpenUrl CrossRef PubMed Mi H. , Muruganujan A. , Casagrande J.T. & Thomas P.D . ( 2013 ) Large-scale gene function analysis with the PANTHER classification system . Nature Protocols 8 , 1551 – 66 . OpenUrl CrossRef PubMed ↵ Nässel D.R. & Zandawala M . ( 2022 ) Endocrine cybernetics: neuropeptides as molecular switches in behavioural decisions . Open Biology 12 , 220174 . OpenUrl CrossRef PubMed ↵ Nojima T. , Rings A. , Allen A.M. , Otto N. , Verschut T.A. , Billeter J.-C. , Neville M.C. & Goodwin S.F . ( 2021 ) A sex-specific switch between visual and olfactory inputs underlies adaptive sex differences in behavior . Current Biology 31 , 1175 – 91 .e6. OpenUrl CrossRef PubMed ↵ Ohashi T.S. , Ishikawa Y. , Awasaki T. , Su M.P. , Yoneyama Y. , Morimoto N. & Kamikouchi A. ( 2023 ) Evolutionary conservation and diversification of auditory neural circuits that process courtship songs in Drosophila . Scientific Reports 13 , 383 . OpenUrl CrossRef PubMed ↵ Özel M.N. & Desplan C . ( 2025 ) Does a cell’s gene expression always reflect its function? Nature 638 , 899 – 900 . OpenUrl CrossRef PubMed ↵ Parker D.J. , Gardiner A. , Neville M.C. , Ritchie M.G. & Goodwin S.F . ( 2014 ) The evolution of novelty in conserved genes; evidence of positive selection in the Drosophila fruitless gene is localised to alternatively spliced exons . Heredity 112 , 300 – 6 . OpenUrl CrossRef PubMed Web of Science ↵ Pollo S.M. , Liu H. , Leon Coria A. , Rosin N. , Labit E. , Biernaskie J. , Finney C.A. & Wasmuth J.D. ( 2024 ) A single-cell transcriptomics atlas for the parasitic nematode Heligmosomoides bakeri : Extrapolating model organism information to non-model systems . bioRxiv , 2024.02. 27.582282. ↵ Prieto-Godino L.L. , Rytz R. , Cruchet S. , Bargeton B. , Abuin L. , Silbering A.F. , Ruta V. , Dal Peraro M. & Benton R . ( 2017 ) Evolution of acid-sensing olfactory circuits in Drosophilids . Neuron 93 , 661 – 76 .e6. OpenUrl CrossRef PubMed ↵ Ren Q. , Awasaki T. , Huang Y.-F. , Liu Z. & Lee T . ( 2016 ) Cell class-lineage analysis reveals sexually dimorphic lineage compositions in the Drosophila brain . Current Biology 26 , 2583 – 93 . OpenUrl CrossRef PubMed ↵ Rezával C. , Nojima T. , Neville Megan C. , Lin Andrew C. & Goodwin Stephen F . ( 2014 ) Sexually dimorphic octopaminergic neurons modulate female postmating behaviors in Drosophila . Current Biology 24 , 725 – 30 . OpenUrl CrossRef PubMed ↵ Rideout E.J. , Billeter J.-C. & Goodwin S.F . ( 2007 ) The sex-determination genes fruitless and doublesex specify a neural substrate required for courtship song . Current Biology 17 , 1473 – 8 . OpenUrl CrossRef PubMed Web of Science ↵ Rideout E.J. , Dornan A.J. , Neville M.C. , Eadie S. & Goodwin S.F . ( 2010 ) Control of sexual differentiation and behavior by the doublesex gene in Drosophila melanogaster . Nature Neuroscience 13 , 458 – 66 . OpenUrl CrossRef PubMed Web of Science ↵ Roberts R.J.V. , Pop S. & Prieto-Godino L.L . ( 2022 ) Evolution of central neural circuits: state of the art and perspectives . Nature Reviews Neuroscience 23 , 725 – 43 . OpenUrl CrossRef PubMed ↵ Robinett C.C. , Vaughan A.G. , Knapp J.-M. & Baker B.S . ( 2010 ) Sex and the single cell . II. There Is a time and place for sex. PLOS Biology 8 , e1000365 . OpenUrl CrossRef PubMed ↵ Röder L. , Vola C. & Kerridge S. ( 1992 ) The role of the teashirt gene in trunk segmental identity in Drosophila . Development 115 , 1017 – 33 . OpenUrl Abstract ↵ Sanders L.E. & Arbeitman M.N . ( 2008 ) Doublesex establishes sexual dimorphism in the Drosophila central nervous system in an isoform-dependent manner by directing cell number . Developmental Biology 320 , 378 – 90 . OpenUrl CrossRef PubMed Web of Science ↵ Schmid A. , Chiba A. & Doe C.Q . ( 1999 ) Clonal analysis of Drosophila embryonic neuroblasts: neural cell types, axon projections and muscle targets . Development 126 , 4653 – 89 . OpenUrl Abstract ↵ Schoofs L. , De Loof A. & Van Hiel M.B. ( 2017 ) Neuropeptides as regulators of behavior in insects . Annual Review of Entomology 62 , 35 – 52 . OpenUrl CrossRef PubMed ↵ Schretter C.E. , Aso Y. , Robie A.A. , Dreher M. , Dolan M.-J. , Chen N. , Ito M. , Yang T. , Parekh R. , Branson K.M. & Rubin G.M . ( 2020 ) Cell types and neuronal circuitry underlying female aggression in Drosophila . Elife 9 , e58942 . OpenUrl CrossRef PubMed ↵ Seeholzer L.F. , Seppo M. , Stern D.L. & Ruta V . ( 2018 ) Evolution of a central neural circuit underlies Drosophila mate preferences . Nature 559 , 564 – 9 . OpenUrl CrossRef PubMed ↵ Sellami A. & Veenstra J.A . ( 2015 ) SIFamide acts on fruitless neurons to modulate sexual behavior in Drosophila melanogaster . Peptides 74 , 50 – 6 . OpenUrl CrossRef PubMed Shafer M.E.R. , Sawh A.N. & Schier A.F . ( 2022 ) Gene family evolution underlies cell-type diversification in the hypothalamus of teleosts . Nature Ecology & Evolution 6 , 63 – 76 . OpenUrl PubMed ↵ Shami A.N. , Zheng X. , Munyoki S.K. , Ma Q. , Manske G.L. , Green C.D. , Sukhwani M. , Orwig K.E. , Li J.Z. & Hammoud S.S . ( 2020 ) Single-cell RNA sequencing of human, macaque, and mouse testes uncovers conserved and divergent features of mammalian spermatogenesis . Developmental Cell 54 , 529 – 47 .e12. OpenUrl CrossRef PubMed ↵ Shankar S. , Chua J.Y. , Tan K.J. , Calvert M.E.K. , Weng R. , Ng W.C. , Mori K. & Yew J.Y . ( 2015 ) The neuropeptide tachykinin is essential for pheromone detection in a gustatory neural circuit . Elife 4 , e06914 . OpenUrl CrossRef PubMed ↵ Shen P. , Wan X. , Wu F. , Shi K. , Li J. , Gao H. , Zhao L. & Zhou C . ( 2023 ) Neural circuit mechanisms linking courtship and reward in Drosophila males . Current Biology 33 , 2034 – 50 .e8. OpenUrl CrossRef PubMed ↵ Shih M.-F.M. , Davis F.P. , Henry G.L. & Dubnau J . ( 2019 ) Nuclear transcriptomes of the seven neuronal cell types that constitute the Drosophila mushroom bodies . G3 Genes|Genomes|Genetics 9 , 81 – 94 . OpenUrl CrossRef Shiozaki H.M. , Wang K. , Lillvis J.L. , Xu M. , Dickson B.J. & Stern D.L . ( 2024 ) Activity of nested neural circuits drives different courtship songs in Drosophila . Nature Neuroscience 27 , 1954 – 65 . OpenUrl CrossRef PubMed Shirangi Troy R. , Wong Allan M. , Truman James W. & Stern David L. ( 2016 ) Doublesex regulates the connectivity of a neural circuit controlling Drosophila male courtship song . Developmental Cell 37 , 533 – 44 . OpenUrl CrossRef PubMed ↵ Skinnider M.A. , Squair J.W. & Foster L.J . ( 2019 ) Evaluating measures of association for single-cell transcriptomics . Nature Methods 16 , 381 – 6 . OpenUrl CrossRef PubMed Soffers J.H. , Beck E. , Sytkowski D.J. , Maughan M.E. , Devasri D. , Zhu Y. , Wilson B. , Chen Y.-C.D. , Erclik T. , Truman J.W. , Skeath J.B. & Lacin H . ( 2025 ) A library of lineage-specific driver lines connects developing neuronal circuits to behavior in the Drosophila ventral nerve cord . eLife , 14 : RP106042 . OpenUrl Spieth H.T . ( 1974 ) Courtship behavior in Drosophila . Annual Review of Entomology 19 , 385 – 405 . OpenUrl CrossRef PubMed Web of Science ↵ Stern D.L. , Crocker J. , Ding Y. , Frankel N. , Kappes G. , Kim E. , Kuzmickas R. , Lemire A. , Mast J.D. & Picard S. ( 2017 ) Genetic and transgenic reagents for Drosophila simulans, D. mauritiana , D. yakuba , D. santomea , and D. virilis . G3 (Bethesda) 7 , 1339 – 47 . OpenUrl Abstract / FREE Full Text ↵ Stoop R . ( 2012 ) Neuromodulation by oxytocin and vasopressin . Neuron 76 , 142 – 59 . OpenUrl CrossRef PubMed Web of Science ↵ Taghert P.H. & Nitabach M.N . ( 2012 ) Peptide neuromodulation in invertebrate model systems . Neuron 76 , 82 – 97 . OpenUrl CrossRef PubMed Web of Science ↵ Takagi S. , Sancer G. , Abuin L. , Stupski S.D. , Roman Arguello J. , Prieto-Godino L.L. , Stern D.L. , Cruchet S. , Álvarez-Ocaña R. , Wienecke C.F.R. , van Breugel F. , Jeanne J.M. , Auer T.O. & Benton R. ( 2024 ) Olfactory sensory neuron population expansions influence projection neuron adaptation and enhance odour tracking . Nature Communications 15 , 7041 . OpenUrl CrossRef PubMed ↵ Tamura K. , Subramanian S. & Kumar S . ( 2004 ) Temporal patterns of fruit fly ( Drosophila ) evolution revealed by mutation clocks . Molecular Biology and Evolution 21 , 36 – 44 . OpenUrl CrossRef PubMed Web of Science ↵ Tanaka K. , Barmina O. , Sanders L.E. , Arbeitman M.N. & Kopp A . ( 2011 ) Evolution of sex -specific traits through changes in HOX-dependent doublesex expression . PLOS Biology 9 , e1001131 . OpenUrl CrossRef PubMed ↵ Tang M. , Kaymaz Y. , Logeman B.L. , Eichhorn S. , Liang Z.S. , Dulac C. & Sackton T.B . ( 2020 ) Evaluating single-cell cluster stability using the Jaccard similarity index . Bioinformatics 37 , 2212 – 4 . OpenUrl ↵ Tirian L. & Dickson B.J. ( 2017 ) The VT GAL4, LexA, and split-GAL4 driver line collections for targeted expression in the Drosophila nervous system . bioRxiv , 198648. ↵ Toker I.A. , Ripoll-Sánchez L. , Geiger L.T. , Saini K.S. , Beets I. , Vértes P.E. , Schafer W.R. , Ben-David E. & Hobert O . ( 2024 ) Molecular patterns of evolutionary changes throughout the whole nervous system of multiple nematode species . bioRxiv , 2024.11.23.624988. ↵ Tregenza T. , Wedell N. & Chapman T . ( 2006 ) Introduction. Sexual conflict: a new paradigm? Philosophical Transactions of the Royal Society B: Biological Sciences 361 , 229 – 34 . OpenUrl CrossRef PubMed ↵ Van Den Pol A.N. ( 2012 ) Neuropeptide transmission in brain circuits . Neuron 76 , 98 – 115 . OpenUrl CrossRef PubMed Web of Science ↵ Villella A. & Hall J.C . ( 1996 ) Courtship anomalies caused by doublesex mutations in Drosophila melanogaster . Genetics 143 , 331 – 44 . OpenUrl Abstract / FREE Full Text ↵ von Philipsborn A.C. , Liu T. , Yu J.Y. , Masser C. , Bidaye S.S. & Dickson B.J. ( 2011 ) Neuronal control of Drosophila courtship song . Neuron 69 , 509 – 22 . OpenUrl CrossRef PubMed Web of Science ↵ Wang H.-Y. , Chen J.-Y. , Li Y. , Zhang X. , Liu X. , Lu Y. , He H. , Li Y. , Chen H. , Liu Q. , Huang Y. , Jia Z. , Li S. , Zhang Y. , Han S. , Jiang S. , Yang M. , Zhang Y. , Zhou L. , Tan F. , Ji Q. , Meng L. , Wang R. , Liu Y. , Liu K. , Wang Q. , Seim I. , Zou J. , Fan G. , Liu S. & Shao C . ( 2024 ) Single-cell RNA sequencing illuminates the ontogeny, conservation and diversification of cartilaginous and bony fish lymphocytes . Nature Communications 15 , 7627 . OpenUrl CrossRef PubMed ↵ Wei J.-R. , Hao Z.-Z. , Xu C. , Huang M. , Tang L. , Xu N. , Liu R. , Shen Y. , Teichmann S.A. , Miao Z. & Liu S . ( 2022 ) Identification of visual cortex cell types and species differences using single-cell RNA sequencing . Nature Communications 13 , 6902 . OpenUrl CrossRef PubMed ↵ Williams T.M. , Selegue J.E. , Werner T. , Gompel N. , Kopp A. & Carroll S.B . ( 2008 ) The regulation and evolution of a genetic switch controlling sexually dimorphic traits in Drosophila . Cell 134 , 610 – 23 . OpenUrl CrossRef PubMed Web of Science ↵ Wittkopp P.J. & Kalay G . ( 2012 ) Cis -regulatory elements: molecular mechanisms and evolutionary processes underlying divergence . Nature Reviews Genetics 13 , 59 – 69 . OpenUrl CrossRef PubMed ↵ Wu S. , Guo C. , Zhao H. , Sun M. , Chen J. , Han C. , Peng Q. , Qiao H. , Peng P. , Liu Y. , Luo S.D. & Pan Y . ( 2019 ) Drosulfakinin signaling in fruitless circuitry antagonizes P1 neurons to regulate sexual arousal in Drosophila . Nature Communications 10 , 4770 . OpenUrl CrossRef PubMed ↵ Yamamoto D. & Koganezawa M . ( 2013 ) Genes and circuits of courtship behaviour in Drosophila males . Nature Reviews Neuroscience 14 , 681 – 92 . OpenUrl CrossRef PubMed ↵ Ye D. , Walsh J.T. , Junker I.P. & Ding Y . ( 2024 ) Changes in the cellular makeup of motor patterning circuits drive courtship song evolution in Drosophila . Current Biology 34 , 2319 – 29 .e6. OpenUrl CrossRef PubMed ↵ Yilmazer Y.B. , Koganezawa M. , Sato K. , Xu J. & Yamamoto D . ( 2016 ) Serotonergic neuronal death and concomitant serotonin deficiency curb copulation ability of Drosophila platonic mutants . Nature Communications 7 , 13792 . OpenUrl CrossRef PubMed ↵ Zeng H . ( 2022 ) What is a cell type and how to define it? Cell 185 , 2739 – 55 . OpenUrl CrossRef PubMed ↵ Zhang S.X. , Rogulja D. & Crickmore M.A . ( 2016 ) Dopaminergic circuitry underlying mating drive . Neuron 91 , 168 – 81 . OpenUrl CrossRef PubMed ↵ Zhao Z. , Boyle T.J. , Bao Z. , Murray J.I. , Mericle B. & Waterston R.H . ( 2008 ) Comparative analysis of embryonic cell lineage between Caenorhabditis briggsae and Caenorhabditis elegans . Developmental Biology 314 , 93 – 9 . OpenUrl CrossRef PubMed Web of Science ↵ Zhou C. , Pan Y. , Robinett Carmen C. , Meissner Geoffrey W. & Baker Bruce S . ( 2014 ) Central brain neurons expressing doublesex regulate female receptivity in Drosophila . Neuron 83 , 149 – 63 . OpenUrl CrossRef PubMed Web of Science View the discussion thread. Back to top Previous Next Posted May 10, 2025. Download PDF Supplementary Material Email Thank you for your interest in spreading the word about bioRxiv. NOTE: Your email address is requested solely to identify you as the sender of this article. Your Email * Your Name * Send To * Enter multiple addresses on separate lines or separate them with commas. You are going to email the following High-resolution comparative single-cell transcriptomics of doublesex-expressing neurons reveals evolutionary conservation and diversity of sexual circuits in Drosophila Message Subject (Your Name) has forwarded a page to you from bioRxiv Message Body (Your Name) thought you would like to see this page from the bioRxiv website. Your Personal Message CAPTCHA This question is for testing whether or not you are a human visitor and to prevent automated spam submissions. Share High-resolution comparative single-cell transcriptomics of doublesex -expressing neurons reveals evolutionary conservation and diversity of sexual circuits in Drosophila Justin Walsh , Ian P. Junker , Yu-Chieh David Chen , Yen-Chung Chen , Helena Gifford , Yun Ding bioRxiv 2025.05.06.652433; doi: https://doi.org/10.1101/2025.05.06.652433 Share This Article: Copy Citation Tools High-resolution comparative single-cell transcriptomics of doublesex -expressing neurons reveals evolutionary conservation and diversity of sexual circuits in Drosophila Justin Walsh , Ian P. Junker , Yu-Chieh David Chen , Yen-Chung Chen , Helena Gifford , Yun Ding bioRxiv 2025.05.06.652433; doi: https://doi.org/10.1101/2025.05.06.652433 Citation Manager Formats BibTeX Bookends EasyBib EndNote (tagged) EndNote 8 (xml) Medlars Mendeley Papers RefWorks Tagged Ref Manager RIS Zotero Tweet Widget Facebook Like Google Plus One Subject Area Evolutionary Biology Subject Areas All Articles Animal Behavior and Cognition (7624) Biochemistry (17651) Bioengineering (13873) Bioinformatics (41887) Biophysics (21424) Cancer Biology (18566) Cell Biology (25465) Clinical Trials (138) Developmental Biology (13365) Ecology (19871) Epidemiology (2067) Evolutionary Biology (24293) Genetics (15591) Genomics (22478) Immunology (17715) Microbiology (40331) Molecular Biology (17150) Neuroscience (88492) Paleontology (666) Pathology (2828) Pharmacology and Toxicology (4817) Physiology (7635) Plant Biology (15114) Scientific Communication and Education (2044) Synthetic Biology (4286) Systems Biology (9817) Zoology (2268)