Author response: Single-cell transcriptome profiles of Drosophila fruitless-expressing neurons from both sexes
Year: 2022
Type: peer-review
Abstract: Article Figures and data Abstract Editor's evaluation Introduction Results and discussion Materials and methods Data availability References Decision letter Author response Article and author information Metrics Abstract Drosophila melanogaster reproductive behaviors are orchestrated by fruitless neurons. We performed single-cell RNA-sequencing on pupal neurons that produce sex-specifically spliced fru transcripts, the fru P1-expressing neurons. Uniform Manifold Approximation and Projection (UMAP) with clustering generates an atlas containing 113 clusters. While the male and female neurons overlap in UMAP space, more than half the clusters have sex differences in neuron number, and nearly all clusters display sex-differential expression. Based on an examination of enriched marker genes, we annotate clusters as circadian clock neurons, mushroom body Kenyon cell neurons, neurotransmitter- and/or neuropeptide-producing, and those that express doublesex. Marker gene analyses also show that genes that encode members of the immunoglobulin superfamily of cell adhesion molecules, transcription factors, neuropeptides, neuropeptide receptors, and Wnts have unique patterns of enriched expression across the clusters. In vivo spatial gene expression links to the clusters are examined. A functional analysis of fru P1 circadian neurons shows they have dimorphic roles in activity and period length. Given that most clusters are comprised of male and female neurons indicates that the sexes have fru P1 neurons with common gene expression programs. Sex-specific expression is overlaid on this program, to build the potential for vastly different sex-specific behaviors. Editor's evaluation This study presents a valuable single-cell sequencing dataset of fruitless-expressing neurons in the male and female Drosophila nervous system. The quality data and convincing analyses allowed the authors to conclude that most neuronal types are present in both Drosophila sexes, suggesting that sex-specific versions of the transcription factor Fruitless can modify neural function in a sex-specific way without completely altering core neural identity. This work will be of interest to developmental biologists and neuroscientists with a focus on sex-specific differences. https://doi.org/10.7554/eLife.78511.sa0 Decision letter Reviews on Sciety eLife's review process Introduction A current goal of neuroscience research is to understand molecular differences at the single cell/neuron level to better understand the diverse cellular components of the nervous system, with the ultimate goal of understanding how diverse cells work together to generate behavior (Ngai, 2022). Drosophila is an excellent model for this approach, given there are defined and experimentally tractable sets of neurons that generate the potential for complex behaviors. Indeed, a Drosophila single-cell RNA-seq (scRNA-seq) atlas has been generated to understand neurons underlying adult circadian biology, with the characterization of the adult core circadian clock neurons (Ma et al., 2021). In addition, there are atlases that have been generated to understand the cellular components of the brain and ventral nerve cord (VNC) during development and adult stages, using single-cell, genome-wide approaches (Allen et al., 2020; Croset et al., 2018; Davie et al., 2018; Konstantinides et al., 2018; Kurmangaliyev et al., 2020; Li et al., 2017; Li et al., 2022; McLaughlin et al., 2021; Özel et al., 2021; Simon and Konstantinides, 2021; Xie et al., 2021). Here, we present a scRNA-seq study to understand how the potential for sexually dimorphic adult reproductive behaviors are specified in the developing nervous system in males and females. We gain insight into how neurons that arise from a sex-shared developmental trajectory can underlie vastly different behaviors in males and females (Ren et al., 2016). In Drosophila, the neuronal substrates that direct reproductive behaviors are specified by the sex-specific transcription factors (TFs) encoded by fruitless (fru; fru P1 transcript isoforms) and doublesex (dsx) (Figure 1A), produced as an outcome of alternative pre-mRNA splicing by the sex determination hierarchy (reviewed in Andrew et al., 2019; Cline and Meyer, 1996). Sex-specific splicing at the 5′ end of fru transcripts produced under control of the most distal promoter (P1 promoter; fru P1 transcripts) results in the production of male-specific (FruM) isoforms. FruM isoforms have an additional 101 amino acid region on the amino terminus that is not present in common Fru isoforms. In females, fru P1 transcripts encode for a short peptide that is predicted to be nonfunctional. fru transcript isoforms are also alternatively spliced at the 3′ end, resulting in products with different DNA-binding domains (Gramates et al., 2022; Ito et al., 1996; Ryner et al., 1996). The identification of master regulatory TFs has provided an unprecedented molecular inroad into a behavioral question, allowing for high-resolution genomic and genetic interrogation, microscopic visualization, and the physiological manipulation of neurons directing behavior. dsx and fru P1-expressing neurons (fru P1 neurons hereafter) are present in both sexes and each set arises from sex-shared developmental lineages (Ito et al., 1996; Lee et al., 2002; Manoli et al., 2005; Ren et al., 2016; Robinett et al., 2010; Ryner et al., 1996; Sanders and Arbeitman, 2008; Stockinger et al., 2005). However, these neurons direct dramatically different innate behaviors in the sexes due to differences in morphology, connectivity, physiology, and number. Males display an intricate courtship display that includes chasing the female, tapping her with his leg, and singing a song by wing vibration. The female will either accept or reject the male’s courtship advances. Once the female has mated, she shows a broad range of post-mating changes including changes in her receptivity to subsequent courtship displays (reviewed in Anholt et al., 2020; Auer and Benton, 2016; Dauwalder, 2011; Greenspan and Ferveur, 2000; Manoli et al., 2006; Peng et al., 2021; Villella and Hall, 2008; Yamamoto et al., 2014). Figure 1 with 8 supplements see all Download asset Open asset scRNA-sequencing of fru P1 neurons from males and females at 48-hr after puparium formation (APF). (A) The Drosophila sex determination hierarchy is an alternative pre-mRNA splicing cascade that generates sex-specific transcription factors that regulate somatic sexual differentiation. The pre-mRNA splicing regulators are encoded by sxl, tra, and tra-2. Early production of Sxl in females results in continued production of functional Sxl and Tra in females. Tra and Tra-2 regulate the sex-specific splicing of doublesex (dsx) and fruitless transcripts from the P1 promoter (fru P1). In females, female-specific Dsx (DsxF) is produced. In males, male-specific (DsxM) and male-specific Fru (FruM) are produced due to default pre-mRNA splicing. Sxl also regulates dosage compensation (DC) (reviewed in Andrew et al., 2019; Cline and Meyer, 1996). (B) Confocal maximum projections of 48-hr APF male (top) and female (bottom) brains and ventral nerve cords (VNCs) expressing membrane-bound GFP (mcd8::GFP) in fru P1-expressing neurons (green). Male tissues are immunostained for male-specific FruM (magenta). Magnification of the boxed regions in male brain and VNC are below the male tissues (×40). Scale bars = 50 μm. (C) Uniform Manifold Approximation and Projection (UMAP) plot of 7988 fru P1 cells, from male central nervous system (CNS) at 48-hr APF, grouped into 77 clusters (male data analysis). (D) UMAP plot of 17,530 fru P1 cells, from female CNS at 48-hr APF, grouped into 88 clusters (female data analysis). (E) UMAP plot of 25,518 total fru P1 neurons, from both sexes (full data set), grouped into 113 clusters. For all UMAP plots, the annotations shown were determined using marker gene expression (Source data 3). Clusters with numbers have additional annotations listed in Source data 3. Here, we focus on fru P1 neurons that are found in both males and females. Earlier studies of fru P1 neurons showed that there are not large sex differences in their number or cell body positions (Manoli et al., 2005; Stockinger et al., 2005). fru P1 was initially thought to be important for male courtship based on analyses of different fru P1 mutant allele combinations (Anand et al., 2001; Gailey and Hall, 1988; Ryner et al., 1996; Villella et al., 1997). Additional studies showed that fru P1 neurons are both necessary and largely sufficient for male courtship behaviors, based on experiments where fru P1 neurons were either activated or silenced and behavior was examined, and by studies where FruM was produced in females in fru P1 neurons (Clyne and Miesenböck, 2008; Demir and Dickson, 2005; Manoli et al., 2005; Stockinger et al., 2005). Female receptivity was later shown to be directed by fru P1 neurons using neuronal silencing approaches (Kvitsiani and Dickson, 2006). fru P1 neurons make up ~2–5% of all neurons, in both sexes, with expression found in the periphery, brain, and VNC (Lee et al., 2000; Manoli et al., 2005; Stockinger et al., 2005). fru P1 neurons in peripheral structures are important for detecting con-specific mates. In the central nervous system (CNS), fru P1 is in regions important for higher-order processing of experience/sensation and regions involved in motor outputs for reproductive behaviors (reviewed in Auer and Benton, 2016). The spatial positions of fru P1 neurons in the brain and VNC have been assigned to named spatial clusters, with the idea that neurons in close proximity may share functions (Billeter and Goodwin, 2004; Lee et al., 2000; Manoli et al., 2005; Stockinger et al., 2005). However, it is not clear if all neurons within a spatial cluster are functionally and/or molecularly similar, which we address here. Furthermore, while the position and number of neurons are similar between the sexes, there is dimorphism in cell number within several spatial clusters, male-specific clusters, sex differences in neuronal projections, and differences in their physiology (reviewed in Auer and Benton, 2016; Billeter and Goodwin, 2004; Cachero et al., 2010; Clowney et al., 2015; Datta et al., 2008; Kallman et al., 2015; Kimura, 2008; Kimura et al., 2005; Ruta et al., 2010; Yu et al., 2010). Sex differences in fru P1 neurons begin to be established during metamorphosis, coinciding with when FruM is at its peak expression in males, during the mid-pupal phase (Arthur et al., 1998; Belote and Baker, 1987; Chen et al., 2021; Kimura et al., 2005; Lee et al., 2000; Stockinger et al., 2005). Therefore, a major remaining question is how does fru establish these sex differences and ultimately the potential for sex-specific behavior during development. This scRNA-seq study addresses what are the different molecular classes of fru P1 neurons that arise during development in each sex. Work from our laboratory and others have examined gene expression in fru P1 neurons, FruM regulated expression, FruM binding, chromatin modifications in fru P1 neurons, and open chromatin regions in fru P1 neurons (Brovkina et al., 2021; Dalton et al., 2013; Goldman and Arbeitman, 2007; Neville et al., 2014; Newell et al., 2016; Palmateer et al., 2021; Vernes, 2014). Among these studies, only two have evaluated the mid-pupal stage (Neville et al., 2014; Palmateer et al., 2021). However, these studies used bulk sequencing approaches, thereby lacking the resolution to examine differences in gene expression across individual neurons. To determine the gene expression signatures for individual fru P1 cells, we performed single-cell RNA-seq during a critical developmental stage for establishing the potential for sex-specific behavior. This produced three cell atlases containing: male fru P1 neurons, female fru P1 neurons, and a full data set of male and female fru P1 neurons co-analyzed. We focus our analyses on the full data set of 25,518 neurons from both males and females that formed 113 molecular clusters after dimensionality reduction using Uniform Manifold Approximation and Projection (UMAP) and clustering. We examine sex differences in gene expression and neuron number within these clusters. We also manually annotated several clusters using the enriched gene expression signatures. We annotate clusters based on gene expression consistent with producing fast-acting neurotransmitters, neuropeptides, receptors, TFs, and several previously characterized fru P1 neuronal populations. We compare our cluster annotations to those in recent Drosophila neuronal scRNA-seq studies, by examining marker gene overlap. Additionally, we examine the in vivo spatial patterns of overlapping expression with fru P1 and several marker genes to provide anatomical links to our analyses. Our identification of a population of neurons that express genes that regulate circadian behavior led to our discovery of a set of fru P1 neurons that regulate activity and period length in a sex-specific manner. Altogether, our data sets show the molecular and functional heterogeneity of fru P1 neurons, revealing the diversity of neurons required to build reproductive behaviors in males and females. While the data reveal sex differences in gene expression within molecularly defined clusters, the observation that nearly all clusters are comprised of male and female neurons indicates that male and female fru P1 neurons also share common gene expression repertoires, with sex-specific information overlaid on these core patterns. Results and discussion scRNA-seq of fru P1-expressing cells in the pupal CNS reveals cell type diversity To gain insight into the molecular and functional differences across fru P1-expressing neurons (fru P1 neurons) in males and females, we performed scRNA-seq, using the 10× Genomics platform. This approach allows us to examine the transcriptomes of individual fru P1 neurons. To enrich for fru P1 neurons, fluorescence-activated cell sorting (FACS) was performed on dissociated neurons from pupae CNS tissue, at 48 hr after puparium formation (APF), that expressed membrane-bound GFP in fru P1 neurons (fru P1>mCD8::GFP) (Manoli et al., 2005; Figure 1B). This is the pupal stage where FruM is at peak expression in males, based on immunostaining, and a critical period for when fru P1 establishes the potential for behavior (Arthur et al., 1998; Belote and Baker, 1987; Chen et al., 2021; Lee et al., 2000). After 10× library construction, Illumina sequencing, read mapping, and processing with the 10× Genomics CellRanger pipeline, we obtained gene-barcode matrices for further analyses. These matrices were filtered to retain high-quality cells: cells were removed that had >5% mitochondrial transcripts (dying and/or stressed cells),<200 genes detected (empty droplets), those expressing more than 4000 genes (nfeatures), and/or possessing more than 20,000 unique molecular identifiers (UMIs; a metric for transcript counts) (potential multiplets) (Figure 1—figure supplement 1). After filtering, there were 7988 male cells and 17,530 female cells in the pooled replicates (Figure 1—figure supplement 1). The pooled replicates from males have 1653 genes and 4712 UMIs per cell (median values). The pooled replicates from females have 1729 genes and 5186 UMIs per cell (median values) (Source data 1). The combined filtered data from both sexes contain 25,518 fru P1-expressing cells (hereafter called the full data set). In the full data set, the cells have a median of 1706 genes and 5046 UMIs detected per cell, which is on par with or better coverage than other single-cell atlas studies that examine the Drosophila CNS (Allen et al., 2020; Avalos et al., 2019; Croset et al., 2018; Davie et al., 2018; Li et al., 2021; Source data 1). As expected, given that fru P1 is expressed in neurons, 99% of the cells in the full data set are classified as neurons (25,509 cells). This is based on detecting expression of at least one of three genes with known neuronal expression (embryonic lethal abnormal vision, elav; neuronal Synaptobrevin, nSyb; long non-coding RNA, noe; Source data 2; DiAntonio et al., 1993; Kim et al., 1998; Robinow et al., 1988). Given this, the cells in the scRNA-seq data set will be called neurons, hereafter. There is minimal cell-cycle impacts on gene expression in neurons in this study, given that during the pupal stage we examined, the majority of neurons are post-mitotic, with only four neuroblasts continuing to divide into the late pupal stages (Ito and Hotta, 1992). Furthermore, we observe that a large percentage of neurons express the post-mitotic markers genes elav (77%) and nSyb (93%) (Source data 2). We also searched the data set for expression of string (stg) that encodes a tyrosine protein phosphatase required for cell-cycle progression, with elevated expression in mitotic neuroblasts. stg expression was only found in 0.3% of neurons in our data set (Source data 2), which also indicates that nearly all the neurons examined are post-mitotic. We next scaled and normalized the expression data from the full data set, and also the male and female data sets separately, using a Seurat workflow (Stuart et al., 2019). To generate UMAP plots for visualization and analysis, we selected the number of significant principle components from the expression data, based on Jackstraw analyses (Stuart et al., 2019) (p < 0.05 for PCs; Figure 1—figure supplement 2). The male and female replicates show large overlap in UMAP space, indicating that there are minimal batch effects between replicates within sex (Figure 1—figure supplement 1G–I). To identify neurons that exhibit similarities in their gene expression profiles, we performed graph-based clustering analysis on each data set. To examine gene expression differences between the clusters in each data set, we identified marker genes, using the Seurat ‘FindAllMarkers’ function and the Wilcox rank sum test (min.pct = 0.25, logfc.threshold = 0.25; which resulted in all called marker genes having p < 0.05). Marker genes are those that have significantly higher expression in one cluster compared to all other clusters within each analysis (Source data 3). Here, we present an in-depth examination of the UMAP and clustering from the full data set and provide links to the sex-specific data sets, in some sections. The male data set has 77 clusters (Figure 1C, Figure 1—figure supplement 3A), the female data set has 88 clusters (Figure 1D, Figure 1—figure supplement 3C), and the full data set has 113 clusters (Figure 1E, Figure 1—figure supplement 4). We observed a high correlation in the number of neurons in each cluster between the replicates from each sex (male replicates r = 0.93; female replicates r = 0.88) (Source data 2), indicating small technical variation in the number of neurons per cluster. Given that fru P1 is expressed in ~2000 neurons (Lee et al., 2000), it is possible we are examining the full repertoire of fru P1 neurons (~3.9× coverage for male cells and ~8.8× coverage for female cells), assuming there are not biases in the ability to capture different fru P1 neurons in the experimental workflow. In addition to the full data set analysis on the merged male and female data, we also performed an analysis on integrated male and female data, using the Seurat integration workflow. These results were highly concordant in terms of overlap (Figure 1—figure supplement 5C, D, Source data 2), providing evidence that the overall conclusions are of this in the data processing methods for additional We also the integrated analysis for (Figure 1—figure supplement Source data and 3). we the molecular differences across the clusters by examining marker There is a range in the number of marker genes for each in the male data set, in the female data set, and in the full data set (Source data 3). In we marker genes in the male data set, in the female data set, and in the full data set (Source data 3). expression show the marker genes, based on for each cluster (Figure 1—figure supplement The expression differences of the marker genes across clusters further that the clusters are different at the molecular level and that the resolution for the clusters results in To understand the functions of neurons in the different clusters, we provide of the clusters, with a focus on the full data set clusters, Figure Source data 3). To gain insight into which genes to focus on for we determined which gene functional are enriched in the marker genes by gene analyses. examination of the terms for three and (Figure 1—figure supplement Source data shows that they are across the three data set analyses (Figure 1—figure supplement and are of developing neurons. In addition, we also performed protein and analyses for these marker gene and that the enriched terms are also across the three data sets (Source data 4). We annotate each cluster by indicating the of the marker genes from some of the enriched protein and and from a of genes that are (Source data 3). Additional of the clusters is based on expression data, of marker genes from each and overlap with other single-cell studies in the CNS (Source data et al., 2020; Avalos et al., 2019; Croset et al., 2018; Davie et al., 2018; et al., 2021). We provide a for clusters and information at the level of marker gene expression for the remaining clusters. of fru P1 scRNA-seq clusters with sex differences in neuron number examination of the full data set UMAP plot shows of male and female neurons within each with some clusters having more neurons from one sex (Figure Figure 1—figure supplement This of the male and female cells is not due to the highly expressed male-specific long non-coding on the 1 and/or and based on a full Seurat analysis performed when these genes are removed (Figure 1—figure supplement We that some of the may be due to developmental differences between males and females, given that females have a pupal phase than males and The results that male and female neurons have gene expression, with some expression differences that the in UMAP This is consistent with our studies that showed male and female fru P1 neurons have gene expression profiles, using a RNA-sequencing approach called et al., 2016; Palmateer et al., 2021; et al., we examine the overlap in marker genes from the three scRNA-seq data sets with gene from our study that examined gene expression enriched in fru P1 neurons at 48-hr APF, we significant overlap between the gene (Source data Palmateer et al., providing further of the scRNA-seq The idea that fru P1 neurons from males and females share gene expression also with the observation that fru P1 neurons in males and females arise from a developmental trajectory (Ren et al., with cell that in similar anatomical positions (Manoli et al., 2005; Stockinger et al., 2005). Additionally, studies using genetic to examine of fru P1 neurons neurons in both sexes, across a large range of molecular expression see et al., 2021; Cachero et al., 2010; Palmateer et al., 2021; Yu et al., 2010). Figure with supplements see all Download asset Open asset Sex differences in single-cell clustering. Uniform Manifold Approximation and Projection (UMAP) plots of the 25,518 fru P1 neurons from males and females (full data with all cells by sex (A) and by sex in cell number (B) (Source data (B) is based on normalized number of female cells on Source data numbers and female-specific clusters and the male-specific clusters and male have more neurons compared to the other sex. and have more neurons compared to the other sex. (C) plot of marker genes in each sex-specific cluster based on in expression (Source data 3). (D) and ventral nerve cord (VNC) maximum projections from to for fru P1-expressing neurons, with expression shown in for males and females Scale bars = 50 μm. analyses of neurons from male-specific and clusters that express dsx are by and The are shown with sex by with male cells in and female cells in Female cells are present only in and 3. plot of marker genes in each dsx based on in expression (Source data 3). plot of genes with known VNC expression in dsx for both For all plot the indicates the percentage of neurons expressing each gene per cluster normalized expression level is shown by We next evaluated the clusters that have differences in neuron number between the sexes in our full data set. Given there are more female neurons in the full data set we several to determine the of this on clustering. After the number of neurons in each female cluster by by we that of the clusters show a sex in their neuron show a show sex and of the clusters are sex (Figure Source data to clusters in the full data set analysis are on normalized data (Source data The workflow to fru P1 neurons may have biases in the neurons our data replicates the is highly given the correlation of cell numbers per cluster between the two replicates (Source data 2). We next determined if this is when numbers of neurons between the sexes are present in the by of female neurons to the number of male neurons, which was performed three We that clusters their after across some of the cell analyses (Figure supplement Source data we the full Seurat workflow by female cells at and UMAP plots and clustering analysis for three with numbers of male and female To examine the of gene expression between the clusters, we in et al., We of the clusters in the analyses to those found in the full data set, and of those there is high of sex-specific and clusters (Figure supplement Source data However, female-specific clusters are present in the analyses and three additional male-specific clusters are (Figure supplement Source data differences are in the analyses when female neurons are removed to the number of male cells, most clusters are we with the analysis of the clusters in the full data set, as the additional female cells provide Sex-specific clusters In the full data set there are four sex-specific clusters female-specific and one with two after the analysis cluster and male-specific cluster (Figure of the marker genes for the female-specific cluster is which encodes a neuropeptide that (Ren et al., neurons may be from the given that cluster marker genes genes and and VNC and et al., 2013; et al., To determine if there are female-specific neurons that also express fru P1 fru in the we used a genetic approach to fru P1-expressing neurons. The approach on the system that expression of a membrane-bound protein et al., of the of a The expression of is directed by fru P1 regulatory
Source: eLife
Institution: Florida State University
Language: English
Subfield: Cellular and Molecular Neuroscience
Field: Neuroscience
Domain: Life Sciences
Open Access status: gold
Estimated APC paid: $2,500
OpenAlex ID: W4318820325