https://orcid.org/0009-0006-9226-4450
Figures
Abstract
Previous research suggests that social isolation induces changes in gene expression that encode a starvation-like brain state and reduce sleep. However, the extent to which social isolation alters behaviors via sex-specific brain changes is unclear. Here, we use Drosophila melanogaster to explore sex differences in isolation-induced behavioral and transcriptomic changes. Male and female adult flies were isolated for seven days, and multiple behavioral sex-based differences were identified through tests of activity, motivation, aggression, and sugar preference. RNA-seq analysis also identified several candidate genes that were associated with sex differences in isolation-induced behavioral changes. Takeout (to) expression and sucrose selection were upregulated exclusively in females following social isolation. Following to knockdown in to-expressing cells, sucrose preference decreased in socially isolated females but increased in males. Overall, our results suggest that manipulating to expression influences sucrose choice in opposite directions between females and males following social isolation. It is possible that isolation-induced to overexpression in non-neuronal cells in the female head could contribute to sex differences in this behavior.
Author summary
Chronic stress contributes to detrimental health effects, but our understanding of how stress induces sex differences in brain gene expression and behavior is incomplete. Here, we use a combination of behavioral testing, RNA sequencing, and genetic manipulations in Drosophila melanogaster to explore how social isolation reveals latent sex differences in gene expression and stress-relevant behaviors. We found the most pronounced sex differences in behaviors related to food choice and motivation. RNA profiling revealed isolated female-specific upregulation of over 100 genes, with many of them relating to reproduction and energy metabolism. We manipulated expression of the candidate gene takeout (to) and found that downregulating to in all to-expressing cells decreases sucrose preference in isolated females but increases it in isolated males. Our results suggest that within a chronic stress context, sex-specific effectors in the head may regulate gene expression related to feeding and macronutrient choice to ensure that females prioritize survival over reproduction. Learning more about this system in flies could provide insight into functionally analogous pathways in humans that may be dysregulated in female-biased stress-related disorders.
Citation: Tawa EA, Schneper L, Cho KI, Sim M, Schoeffler E, Notterman DA (2026) Social isolation upregulates takeout expression in female Drosophila melanogaster to promote sucrose preference. PLoS Genet 22(8): e1011985. https://doi.org/10.1371/journal.pgen.1011985
Editor: Giovanni Bosco, Geisel School of Medicine at Dartmouth, UNITED STATES OF AMERICA
Received: December 6, 2025; Accepted: July 21, 2026; Published: August 10, 2026
Copyright: © 2026 Tawa et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The RNA-seq data discussed in this publication have been deposited in NCBI’s Gene Expression Omnibus (Edgar et al., 2002) and are accessible through GEO Series accession number GSE336379 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE336379).
Funding: This work was supported by Princeton University funds (DN and ET), Princeton Neuroscience Institute departmental funds (ET), and the Princeton Center for Health and Wellbeing Program on U.S. Health Policy (ET). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Social isolation has emerged in recent decades as a worldwide public health issue. The World Health Organization estimates that one in four older adults and five to fifteen percent of adolescents experience social isolation and loneliness globally [1], while the Centers for Disease Control and Prevention estimates that one in three adults experience these problems in the United States [2]. In humans, isolation increases the risk of early death and is associated with several diseases and psychiatric conditions, including cardiovascular disease, stroke, type 2 diabetes, depression, anxiety, suicidality, and dementia [1, 2]. Social isolation has also been found to enhance aggression, depression, and anxiety in humans undergoing solitary confinement [3]. Although there is suggestive evidence of sex differences in biological processes leading to differences in health outcomes between men and women experiencing social isolation [4, 5], the exact molecular and neurobiological mechanisms remain largely unknown. In Drosophila melanogaster, chronic social isolation has been shown to reduce sleep, increase feeding, and contribute to a transcriptomic profile indicative of a starvation-like brain state [6]. However, the cited study only examined the effects of chronic social isolation on activity and feeding in males.
Our goal was to develop a battery of behavioral tests and phenotypes that could be used to assess sex differences in the behavioral response to social isolation. We identified a set of fly behaviors that reflected sex-specific changes following seven days of social isolation. The sex differences in these phenotypes prompted us to perform RNA sequencing (RNA-seq) on the heads of male and female flies from socially isolated and group-housed conditions for target identification. Manipulations of one of the emerging target genes, takeout (to), suggest mechanistic relationships between isolation, sex, gene expression changes, and sucrose preference in the value-based feeding decision (VBFD) test. Overall, this study supports the use of Drosophila melanogaster as a valid non-mammalian model for elucidating the molecular mechanisms of social isolation-induced behavioral change, as well as sex differences in these mechanisms.
Results
Locomotion increases to a greater extent in males compared to females following social isolation
To characterize social isolation as a valid fly stress model, we performed a series of behavioral tests that could capture changes of potential translational relevance. A walking assay revealed that social isolation increased distance traveled in both males (Fig 1A, p < 0.0001) and females (Fig 1A, p < 0.0001), but that the magnitude of the change was significantly greater in males (Fig 1B, p < 0.0001).
A) Flies that experienced social isolation (n = 99 individuals) traveled greater distances in a 10-minute period than controls (n = 100 individuals) for both males (U = 1729, p < 0.0001) and females (U = 2808, p < 0.0001). B) Taking differences in distance traveled between socially isolated flies and the control average for each sex revealed that the increase in locomotion in isolated males relative to control males was significantly greater (n = 99, U = 2850, p < 0.0001) than that observed between isolated females and control females (n = 99). C) Set up: flies were briefly anesthetized, positioned along the midline of the arena, and recorded during 60 minutes of exploration time. One half of the arena contained four dots of yeast paste or apple cider vinegar, and the other half had no food. Created in BioRender. Tawa, E. (2026) https://BioRender.com/sg414m7 D) A preference index (PI) was calculated at five-minute intervals, where PI = (the number of flies on the food side of the arena) - (the number of flies on the empty side of arena)/ the total number of flies. A value of 0 can be interpreted as at chance, due to equal numbers of flies being on each side. E) Isolated male flies (n = 13 groups, ~ 260 flies) are relatively more attracted to yeast paste than control males (n = 10 groups, ~ 200 flies) (U = 32, p = 0.02), yet isolated females (n = 11 groups, ~ 220 flies) are less attracted to yeast relative to control females (n = 10 groups, ~ 200 flies) (U = 8, p < 0.0001). F) Both isolated males (U = 4, p < 0.0001) and isolated females (U = 6, p < 0.0001) are more attracted to apple cider vinegar relative to their respective controls (n = 8 groups for all, ~ 160 flies). G) Set up: caps filled with standard fly food and topped with yeast paste were placed in a 12-well plate. Same-sex pairs of flies from the same experimental condition were aspirated into each well and recorded for 60 minutes. Male “hits” were defined as lunges and female “hits” were defined as head butts. Created in BioRender. Tawa, E. (2026) https://BioRender.com/aglrq9y. H) Socially isolated males (n = 16 pairs) hit more compared to group housed controls (n = 16 pairs, U = 76, p < 0.05). There was no difference in the number of hits between isolated (n = 15 pairs) and control females (n = 15 pairs, U = 101, p = 0.87). I) The latency between the flies’ first encounter on the food cup and the first hit was lower in socially isolated males relative to controls (U = 71, p < 0.03). Female controls had a significantly shorter latency to first hit compared to male controls (U = 40, p = 0.002) that did not differ from that of isolated females (U = 79, p = 0.57). J) Isolated males had a greater number of social interactions, aggressive or non-aggressive, than control males (U = 49.5, p = 0.002) in the 10 minutes following the first hit. K) Measuring the proportion of all social interactions that were hits, both control (U = 37, p = 0.0006) and isolated males (U = 14, p < 0.0001) had hits as a greater proportion of all social interactions relative to control females. However, social isolation significantly raised this proportion in females (U = 39.5, p = 0.002) to the levels of males. Error bars are ± SEM. *P < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.
Social isolation impacts motivation to seek food (yeast) differentially in males versus females
We next decided to assess whether social isolation differentially affected the goal-seeking behavior of males and females, specifically in the context of food motivation. To determine if social isolation alters motivated behavior in flies, we used two different stimuli, yeast paste and apple cider vinegar, to measure flies’ motivation to explore a physical space containing food. Yeast volatiles, either alone or as byproducts of fermenting fruit, attract Drosophila [7–10]. Apple cider vinegar, through its high acetic acid content, mimics the odor of rotting fruit, which attracts flies as a source of food and mates. We developed a novel behavior test, inspired by an existing sucrose preference test [11], to assess whether social isolation alters the motivation of flies to investigate an attractive food or food odorant, and if so, whether sex differences are evident.
The preference index (PI) was calculated such that a value of 0 indicated no discrimination between the food side versus non-food side. Using a mixed effects analysis or three-way ANOVA (depending on whether the numbers across experimental groups were equal), we examined the effects of sex, social isolation, and the interaction of sex and social isolation on yeast preference over the 60-minute time course. There was a significant interaction between social isolation and sex (F(1, 477) = 50.89, p < 0.0001). Post-hoc tests indicated that social isolation increased males’ preference for yeast (Fig 1E, U = 32, p = 0.02) but decreased it in females (Fig 1E, U = 8, p < 0.0001). In contrast, using apple cider vinegar as an attractive odorant, we observed a significant effect of social isolation (F(1, 168) = 63.01, p < 0.0001) and the interaction of sex and social isolation (F(1, 168) = 9.14, p = 0.003), but not sex alone (F(1, 168) = 1.21, p = 0.27), on preference. These findings suggest that the interaction is largely driven by social isolation, and that for apple cider vinegar, both males (Fig 1F, U = 4, p < 0.0001) and females (Fig 1F, U = 6, p < 0.0001) were more attracted to the odorant side of the arena following social isolation relative to controls.
Social isolation increases aggression in males and females across different measures
After testing flies on both hyperactivity and food motivation, we assessed aggression, a behavior that has been previously associated with social isolation in fly models [12–14]. We used a set up (Fig 1G) similar to one used by other research groups studying Drosophila aggression [15]: microtube caps were filled with standard fly food, topped with yeast paste, and placed in a 12-well plate, such that each well contained a single cap. This set up created an arena, where same-sex flies would encounter each other and begin fighting for territory and food. We defined aggressive “hits” in male flies as the lunge, a behavior in which the aggressor rears up on his hind legs and quickly snaps down on his opponent [15]. Aggressive “hits” in females were defined as the head butt, a movement in which the aggressor produces an abdominal thrusting-type motion and strikes the opponent with her head [16].
We found that social isolation increased the number of hits in males (p < 0.05) but not females (p = 0.87) (Fig 1H). Social isolation also decreased latency to first hit in males (p < 0.03) but not females (0.57), relative to the respective control groups (Fig 1I). Interestingly, females at baseline have a shorter latency to first hit compared to males. This is reflected in the significant difference in latency to first hit between male controls and female controls (p = 0.002). Thus, there may not be a difference between socially isolated females and control females in latency to first hit because the latency in control females is so short to begin with, creating a floor effect. Isolated males also showed a greater number of total social interactions (both aggressive and non-aggressive encounters) compared to controls (p = 0.002) which was not evident in females (Fig 1J). Finally, when examining aggressive hits as a proportion of all social interactions (Fig 1K), there is a difference between isolated females and their controls, such that hits make up a significantly greater proportion of social behaviors in isolated females (p = 0.002). This difference is not seen between isolated males and control males. Both control males (p = 0.0006) and isolated males (p < 0.0001) showed a higher hit proportion compared to control females, but isolated females’ increase in hit proportion rose to the extent that there was no difference between isolated females and isolated males. Thus, the increase in number of hits observed in isolated males may be a product of the fact that isolated males are also interacting more overall; there does not seem to be an increase in the frequency of hitting relative to other social behaviors. Isolated females’ increased proportion of hits out of total social interactions in combination with females’ overall short latency to hit may indicate enhanced female territoriality.
Social isolation enhances sucrose preference in females
After assessing changes in aggression, we employed the value-based feeding decision (VBFD) test as a final measure to evaluate the effects of social isolation on food motivation and choice. We followed an existing protocol [17], in which flies were placed in an arena for two hours to feed freely on either a blue solution of sucrose, a more nutritious sugar, or a red solution of arabinose, a less nutritious sugar (Fig 2A). The observed color (blue, red, purple, or none) of the crop was used to sort the flies into categories of sugar primarily consumed (Fig 2B). We performed additional testing to confirm that the flies did not have a color preference.
A) Set up: flies were briefly anesthetized and allowed to explore an agarose plate with 2 different sugar solutions for 2 hours. Sucrose (150mM) was dyed blue, and arabinose (150mM) was dyed red. Created in BioRender. Tawa, E. (2026) https://BioRender.com/n8xc0ig. B) Flies were identified (clockwise from top left) as either blue, purple, none, or red and counted at the end of each 2-hour period. All flies shown are female, except for none (bottom right). C) Social isolation increased sucrose preference in females (t(15) = 3.14, p = 0.007). D) Control (n = 13 groups, t(22) = 2.85, p = 0.009) and isolated (n = 10 groups, t(19) = 2.67, p = 0.02) males both demonstrate a stronger sucrose preference than control females (n = 11 groups), but there are no significant differences between these groups and isolated females. Error bars are ± SEM. *p < 0.05, **p < 0.01.
Following social isolation, the proportion of flies consuming sucrose increased significantly in females (Fig 2C, t(15) = 3.14, p = 0.007). No differences were observed in feeding choice between control males and isolated males. Focusing on sucrose, the nutritious sugar used in this test, we observed that both control males (Fig 2D, t(22) = 2.85, p = 0.009) and isolated males (Fig 2D, t(19) = 2.67, p = 0.02) had a greater proportion of flies consuming sucrose relative to control females, but that these differences were erased when comparing males to socially isolated females. Thus, isolation is associated with females increasing their sucrose preference to more closely resemble that of males.
Social isolation uncovers sex differences across behaviors
The outcomes of all behavioral tests are summarized and compared by sex in Table 1. Overall, social isolation produced behavioral alterations across tests of activity, motivation, aggression, and sucrose preference. Following isolation, males increased their activity, their motivation to explore a space containing yeast paste or apple cider vinegar odorant, their number of aggressive hits (lunges), and overall social interactions in the aggression assay compared to male controls. Males’ latency to strike an opponent decreased, indicating heightened aggression. Females also increased their activity, motivation to explore the apple cider vinegar odorant, percentage of overall social interactions that were aggressive hits (head butts), and preference for sucrose relative to arabinose and no food. Isolated females decreased their exploration of the yeast paste side of the arena relative to control females.
Social isolation does not impact lifespan
To explore the stress effects of social isolation further, we examined whether isolation affected longevity. We found no differences in survival between socially isolated and control flies for males (S1 Fig, χ2 = 2.71, df = 1, p = 0.10) or females (S1 Fig, χ2 = 0.25, df = 1, p = 0.62). Additional statistical tests, beyond the log-rank test, are detailed in Table 2.
Differential expression analysis indicates female-specific overexpression of fat body-synthesized proteins following social isolation
Upon identifying multiple sex differences in fly aggression and food motivation following social isolation, we decided to perform RNA-seq to uncover any sex- and isolation-based transcriptional differences. RNA samples were extracted in triplicate from whole fly heads for each experimental group. After performing multidimensional scaling (MDS) (S2 Fig), one sample from each of the female groups was removed to improve target discovery accuracy. When comparing the interaction between social isolation and sex, clusters of genes emerged that were upregulated in isolated females compared to control females and males (Fig 3A). Upon analyzing these significant genes with Metascape [18], changes were identified in groups of genes related to cell division (e.g., mitotic cell cycle process, asymmetric cell division) and metabolism (e.g., cellular catabolic process, fatty acid metabolic process) (Fig 3B).
A) Heat map depicting the scaled (z-score transformed) log2 counts per million of genes differentially expressed (absolute value of log2-fold change > 2 and adjusted p-value < 0.05) when comparing the difference between SI female and control female with the difference between SI males and control males. B) Gene ontology (Metascape) of the significantly differentially expressed genes captured in the heat map reflects changes in broad categories of cell division and metabolism. C) The top differentially expressed genes (absolute value of log2-fold change > 1 and adjusted p-value < 0.05), when comparing the interaction between social isolation and sex, reflect upregulation in isolated females. The red points represent the genes upregulated in socially isolated females (relative to female controls) compared with socially isolated males (relative to male controls). In contrast, blue points represent the genes (CG11211, VhaM9.7-a, and AttB) upregulated in socially isolated males compared to the socially isolated females, relative to their respective controls. D) Gene expression, quantified in transcripts per million (TPM) from RNA-seq, indicates elevation in takeout, Lsp2, Fbp1, and Lsp1α in isolated females relative to control females that is not seen in males. At baseline, control females had lower TPMs compared to control males (t(4) = 5.07, p = 0.007) or isolated males (t(4) = 5.79, p = 0.004). However, socially isolated females had significantly higher TPMs compared to control females (t(4) = 8.56, p = 0.001) which are comparable to those of males. TPMs were also elevated following social isolation in females for Lsp2 (t(4) = 7.06, p = 0.002), Fbp1 (t(2.06) = 8.64, p = 0.01 with Welch’s correction), and Lsp1α (t(4) = 4.14, p = 0.01). Closed circles represent values derived from individual RNA samples, which were extracted from 200 fly whole heads. Error bars are ± SEM. *p < 0.05, **p < 0.01.
Additionally, entering the significantly differentially expressed genes (comparing the interaction of social isolation and sex) into the STRING database [19] returned a network map of potential functional protein-protein interactions among them (S3 Fig). Applying the Markov Cluster (MCL) Algorithm [20] to our data then parsed our genes into 21 potential functional clusters, many of which coincided with the Metascape gene ontology results. Clusters with the strongest functional enrichment included genes related to mitotic cell cycle (S3 Fig), female reproduction (S3 Fig), and hemocyanins/hexamerins and secreted storage proteins (S3 Fig). STRING functional enrichment analysis also reflected enrichment in mitotic cell cycle processes and cell division, egg activation, piRNA metabolic process, and hemocyanins/hexamerins (S4 Fig). Finally, the significant genes were implicated in Drosophila phenotypes of abnormal meiosis and female sterility (S4 Fig).
Finally, we used a significance cut-off of padj < 0.05 to visualize differentially expressed genes in a volcano plot (Fig 3C). Only three significant genes were downregulated when comparing the interaction of social isolation and sex, while 105 were upregulated. The most significantly differentially expressed genes, by adjusted p-value, are listed in Table 3, along with their log fold change values. These genes were all upregulated in isolated females compared to isolated males, when taking into account the differences between the isolated groups and their respective controls.
An analysis of expression quantity data (transcripts per million, TPM) of some of the top genes identified by our model confirmed isolated female-specific overexpression of RNA transcripts for larval serum protein-2 (Lsp-2), larval serum protein-1ɑ (Lsp-1ɑ), fat body protein 1 (Fbp1), and takeout (to). Lsp-2, Lsp-1ɑ, Fbp1, and to are all synthesized in fat body cells [21–24], with the first three genes functioning as amino acid storage proteins and to being implicated in feeding behavior. To expression is higher in control (Fig 3D, t(4) = 5.07, p = 0.007) and isolated males (Fig 3D, t(4) = 5.79, p = 0.004) relative to control females, but this male-female difference is abolished when females have undergone social isolation. To expression is significantly higher in isolated females compared to control females (Fig 3D, t(4) = 8.56, p = 0.001). Similarly, social isolation increases Lsp-2 (Fig 3D, t(4) = 7.06, p = 0.002), Lsp-1ɑ (Fig 3D, t(4) = 4.14, p = 0.01), and Fbp1 (Fig 3D, t(2.06) = 8.64, p = 0.01) expression relative to control levels in females but not in males.
To confirm the RNA-seq results, we used qPCR as a secondary method to validate the expression of to in male and female fly heads. Using remaining RNA from the same samples that were submitted for sequencing, we analyzed expression of to in isolated males and isolated females relative to their respective controls. The qPCR data corroborated the RNA-seq findings. Overexpression of to following social isolation was specific to females and differed significantly from the gene expression change between isolated and control males (S5 Fig, t(4) = 6.14, p = 0.004).
qPCR of ORE-R brains suggests that social isolation increases expression of takeout in females but not males
Following RNA-seq profiling and validation, we identified to as a gene demonstrating female-specific overexpression in the head following social isolation. However, this finding raised the question of whether to expression was changing in the brain or elsewhere in the head. For instance, the sex difference in isolation-induced to expression could stem from transcriptional changes in the pericerebral fat body, where to has been established to be transcriptionally regulated in a sexually dimorphic manner, or in the antennae [23] or foregut [25]. To gain a better understanding of whether to expression is changing within or outside of the fly brain, the brains of additional male and female flies that had experienced either social isolation or group housing (control conditions) were isolated, and RNA was extracted. The subsequent qPCR experiment again reflected female-specific overexpression of to following isolation (S6 Fig, t(10) = 2.28, p < 0.05).
Adult Drosophila express to primarily in sensory neurons, epithelial cells of the gastrointestinal tract, and antennal nerve glia
In addition to validating our bulk RNA-seq results with qPCR, we also compared our findings with existing data from Drosophila melanogaster RNA-seq atlases. RNA-seq data from the modENCODE project [26] indicate that adult male Drosophila have higher whole organism-wide expression of to compared to adult females, which is consistent with our data comparing males to control female heads and brains.
Based on data from the Fly Cell Atlas project [27], to was most highly expressed in sensory neurons, sensory organ cells, glial cells, and epithelial cells out of all adult cell types in the Drosophila body (S7 Fig). The RNA-seq atlas data also showed to to be expressed in multiple epithelial cell types throughout the digestive system (S7 Fig), including the cells of the crop, midgut, esophagus, hindgut, and enteroendocrine cells, which secrete hormones to regulate digestion, metabolic homeostasis, and feeding behavior [28]. Finally, in glial cells, to showed the most widespread expression in adult antennal nerve glia (S7 Fig), which ensheath olfactory receptor neurons and may modulate the neuronal response to olfactory stimuli [29].
Based on these expression patterns and the behaviors to which to has been linked in the literature, we decided to first manipulate to expression systemically. This allowed us to assess whether there was a relationship between overall to expression and sucrose preference before focusing on a specific cell or tissue type. Next, because we were able to replicate our whole head RNA-seq finding in isolated brains, and glial cells only constitute approximately 10% of cells in the Drosophila nervous system [30], we decided to manipulate to expression in postmitotic neurons.
Knocking down to in to-expressing cells decreases sucrose preference in socially isolated females and increases it in socially isolated males
A complete summary of our manipulations and behavioral findings is provided in Table 4. Data on the number of groups tested and average group size across all genotypes and experimental groups can be found in Supporting Information (S1 Table). In order to manipulate the expression of to in to-expressing cells, we mated virgin female flies with GAL4 expressed under the control of a to-specific promoter (to-GAL4) to males with a UAS enhancer 1) inducing the generation of RNA interference (RNAi) specifically targeted to to mRNA (UAS-toRNAi), or 2) driving expression of to (UAS-to). We performed qPCR tests to validate that to expression was downregulated in to-GAL4 x UAS-toRNAi flies relative to ORE-Rs and upregulated in to-GAL4 x UAS-to flies relative to ORE-Rs across experimental groups (S13 Fig). The to-GAL4 x UAS-toRNAi offspring were tested in the VBFD test to assess the effect of to knockdown in to-expressing cells on the proportion of flies that chose to feed on sucrose, a more nutritious sugar, over arabinose, a less nutritious sugar. The “proportion of all flies” dependent measure represents the proportion of all flies in a test group that fed exclusively or almost exclusively on sucrose, compared to all other combinations of feeding choice (a sucrose-arabinose mixture, arabinose only, or neither).
Control males, isolated males, control females, and isolated females of to-GAL4 x UAS-toRNAi and each parent genotype were tested. We observed an increase in sucrose preference in isolated males compared to control males (S8 Fig, t(14) = 2.31, p = 0.04), but a decrease in sucrose preference in isolated females relative to female controls (S8 Fig, t(15) = 5.35, p < 0.0001). This male isolation-induced increase in sucrose choice is specific to the knockdown manipulation and is not evident across either of the parent genotypes or the wild type. Moreover, knocking down to (to-GAL4 x UAS-toRNAi) reverses the trend observed in isolated females relative to control females that is also observed in the UAS-toRNAi and ORE-R flies.
Upon examining the effect of genotype on sucrose consumption, within a specific experimental group, we found that to-GAL4 x UAS-toRNAi isolated females had a significantly lower proportion of sucrose-choosing flies compared to all other genotypes: ORE-R (Fig 4B, t(12) = 5.24, p = 0.0002), to-GAL4 (Fig 4B, t(13) = 5.53, p < 0.0001), UAS-toRNAi (Fig 4B, t(14) = 3.15, p = 0.007), and to-GAL4 x UAS-mCherryRNAi (Fig 4B, t(12) = 3.84, p = 0.002). In isolated males, the to-GAL4 x UAS-toRNAi flies increased their proportion of sucrose preference relative to both parent genotypes (to-GAL4: Fig 4D, t(15) = 2.39, p = 0.03; UAS-toRNAi: Fig 4D, t(14) = 3.84, p = 0.002) and the ORE-Rs (Fig 4D, t(16) = 2.14, p < 0.05).
The proportion of all flies that primarily chose sucrose decreases relative to other genotypes in isolated females upon overexpression of to in to-expressing cells. A) Sucrose preference is elevated in control females of the to-GAL4 x UAS-toRNAi cross relative to ORE-R (t(18) = 3.41, p = 0.003) and UAS-toRNAi (t(22) = 6.07, p < 0.0001) control females but does not differ from to-GAL4 or to-GAL4 x UAS-mCherryRNAi control females. B) Sucrose preference is reduced in isolated females of the to-GAL4 x UAS-toRNAi cross relative to isolated females of all other genotypes: ORE-R (t(12) = 5.24, p = 0.0002), to-GAL4 (t(13) = 5.53, p < 0.0001), UAS-toRNAi (t(14) = 3.15, p = 0.007), and to-GAL4 x UAS-mCherryRNAi (t(12) = 3.84, p = 0.002). C) Sucrose preference is reduced in control males of the to-GAL4 x UAS-toRNAi cross relative to ORE-R control males (t(20) = 2.31, p = 0.03) and to-GAL4 x UAS-mCherryRNAi control males (t(14) = 3.89, p = 0.002) but does not differ from parent genotype control males. D) Sucrose preference increases in isolated males of the to-GAL4 x UAS-toRNAi cross relative to wild type and parent genotype isolated males: ORE-R (t(16) = 2.14, p < 0.05), to-GAL4 (t(15) = 2.39, p = 0.03), and UAS-toRNAi (t(14) = 3.84, p = 0.002). E) Sucrose preference is elevated in control females of the to-GAL4 x UAS-to cross relative to UAS-to control females (t(16) = 4.50, p = 0.0004) but does not differ from ORE-R or to-GAL4 control females. F) Sucrose preference is reduced in isolated females of the to-GAL4 x UAS-to cross relative to isolated females of the wild type and parent genotypes: ORE-R (t(11) = 5.54, p = 0.0002), to-GAL4 (t(12) = 5.65, p = 0.0001), and UAS-to (t(13) = 2.74, p = 0.02). G) Sucrose preference is enhanced in control males of the to-GAL4 x UAS-to cross relative to UAS-to control males (t(18) = 3.91, p = 0.001) but does not differ from wild type or to-GAL4 control males. H) Sucrose preference is elevated in isolated males of the to-GAL4 x UAS-to cross relative to UAS-to isolated males (t(14) = 3.31, p = 0.005) but does not differ from wild type or to-GAL4 control males. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.
Overexpressing to in to-expressing cells decreases sucrose preference in isolated females, but not males, relative to the ORE-R and parent genotypes
Upon testing the to-GAL4 x UAS-to flies on the VBFD test, we found that isolation did not seem to affect sucrose choice in either sex (males: S9 Fig, t(14) = 0.30, p = 0.77; females: S9 Fig, t(15) = 1.59, p = 0.13). However, in isolated females, sucrose preference was depressed in to-GAL4 x UAS-to flies compared to ORE-Rs (Fig 4F, t(11) = 5.54, p = 0.0002) and both parent genotypes (to-GAL4: Fig 4F, t(12) = 5.65, p = 0.0001; UAS-to: Fig 4F, t(13) = 2.74, p = 0.02). In both control (Fig 4G, t(18) = 3.91, p = 0.001) and isolated males (Fig 4H, t(14) = 3.31, p = 0.005), sucrose preference was increased in to-GAL4 x UAS-to flies compared to UAS-to males but did not differ from to-GAL4 or ORE-R flies of the same experimental condition.
Driving to expression in neurons increases sucrose preference in males
After observing sex differences in sucrose preference upon to manipulation in all to-expressing cells (using the to-GAL4 driver), we proceeded to down- and upregulate expression of to in neurons alone. We began by mating elav-GAL4 virgin female flies with UAS-toRNAi males to generate flies expressing RNAi targeted to to in postmitotic neurons. The elav-GAL4 line is commonly used in Drosophila research to drive the expression of GAL4 under control of the elav gene, which is active in postmitotic neurons throughout development and is expressed throughout most cells of the fly nervous system [31, 32]. Via qPCR, we validated that to expression was downregulated in elav-GAL4 x UAS-toRNAi flies relative to ORE-Rs across experimental groups (S13 Fig).
We did not observe any differences in sucrose preference following social isolation in male or female offspring of the elav-GAL4 x UAS-toRNAi cross (S10 Fig). An examination of within-experimental group effects of genotype on sucrose choice also failed to show an effect of to knock down in neurons on behavior. Although elav-GAL4 x UAS-toRNAi control females increased sucrose preference compared to ORE-R (Fig 5A, t(16) = 2.25, p = 0.04) and UAS-toRNAi (Fig 5A, t(20) = 4.72, p = 0.0001) control females, there was no difference relative to the other parent genotype, elav-GAL4. Across isolated females (Fig 5B), control males (Fig 5C), and isolated males (Fig 5D), the elav-GAL4 x UAS-toRNAi flies did not differ in their sucrose preference from any other experimental control group (the wild type and parent genotypes) in a statistically significant manner.
The proportion of all flies that primarily chose sucrose increases relative to parent genotypes in control males and isolated males upon overexpression of to in neurons. A) Sucrose preference is elevated in control females of the elav-GAL4 x UAS-toRNAi cross relative to ORE-R (t(16) = 2.25, p = 0.04) and UAS-toRNAi (t(20) = 4.72, p = 0.0001) control females but does not differ from elav-GAL4 control females. B) Sucrose preference does not differ between elav-GAL4 x UAS-toRNAi, wild type, and parent genotype isolated females. C) Sucrose preference does not differ between elav-GAL4 x UAS-toRNAi, wild type, and parent genotype control males. D) Sucrose preference does not differ between elav-GAL4 x UAS-toRNAi, wild type, and parent genotype isolated males. E) Sucrose preference is elevated in control females of the elav-GAL4 x UAS-to cross relative to UAS-to control females (t(12) = 7.15, p < 0.0001). F) Sucrose preference is reduced in isolated females of the elav-GAL4 x UAS-to cross relative to ORE-R isolated females (t(9) = 2.30, p < 0.05). G) Sucrose preference is enhanced in control males of the elav-GAL4 x UAS-to cross relative to UAS-to control males (t(15) = 3.63, p = 0.003) and elav-GAL4 control males (t(13) = 3.33, p = 0.005). H) Sucrose preference is also elevated in isolated males of the elav-GAL4 x UAS-to cross relative to UAS-to isolated males (t(12) = 3.30, p = 0.006) and elav-GAL4 isolated males (t(12) = 2.71, p = 0.02). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.
Next, we set up elav-GAL4 x UAS-to crosses to assess the behavioral effect of to overexpression in neurons.When we tested the elav-GAL4 x UAS-to flies, there was not an effect of isolation on sucrose choice in females (S11 Fig, t(9) = 0.87, p = 0.41). For within-experimental group effects of genotype on sucrose choice, we noted that in males, both control flies (Fig 5G) and socially isolated flies (Fig 5H) increased sucrose preference relative to both parent genotypes. Thus, overexpressing to in postmitotic neurons seems to increase sucrose preference in males, but not females, irrespective of social housing or isolation.
Masculinizing female neurons via RNAi targeted to transformer (tra) does not alter sucrose preference
Finally, we wanted to test whether altering the sexual identity of neurons to make female neurons more male-like affected sucrose choice. To do this, we crossed elav-GAL4 virgin females with UAS-transformer (tra)RNAi males. This genetic manipulation has been used previously as a tool to masculinize neurons in female flies [33]. The UAS-traRNAi line uses double-stranded RNA to knock down tra, a gene in the Drosophila sex determination pathway that is crucial for female somatic sexual differentiation [34]. Using qPCR, we validated that tra expression was downregulated in elav-GAL4 x UAS-traRNAi flies relative to ORE-Rs across experimental groups (S13 Fig).
Overall, we did not observe changes in sucrose preference in isolated males and isolated females relative to their respective controls in the elav-GAL4 x UAS-traRNAi offspring (S12 Fig, males: t(17) = 1.42, p = 0.17; females: t(17) = 0.91, p = 0.38). Within-condition comparisons across genotypes also did not reflect any clear differences in sucrose preference upon tra knock down in neurons. We did not see any differences across all genotypes in control females (Fig 6A) or isolated males (Fig 6D). Interestingly, qPCR validation revealed that knockdown of tra resulted in decreased expression of to across all experimental groups relative to ORE-R levels (S13 Fig). Since knocking down to expression in neurons through the elav-GAL4 x UAS-toRNAi manipulation did not yield any changes in sucrose preference, it is consistent that decreasing to expression through neuron masculinization similarly did not produce clear changes in this behavior.
A) No significant differences were detected in sucrose preference between elav-GAL4 x UAS-traRNAi control females and control females of the wild type and parent genotypes. B) Sucrose preference is reduced in isolated females of the elav-GAL4 x UAS-traRNAi cross relative to ORE-R (t(10) = 2.54, p = 0.03) and elav-GAL4 isolated females (t(14) = 2.19, p < 0.05). C) Sucrose preference is elevated in elav-GAL4 x UAS-traRNAi control males relative to elav-GAL4 control males (t(13) = 2.61, p = 0.02). D) No significant differences were detected in sucrose preference between elav-GAL4 x UAS-traRNAi isolated males and isolated males of the wild type and parent genotypes. *p < 0.05.
Discussion
We first aimed to establish social isolation as a valid stressor in flies, using behavioral and survival data to capture stress-induced phenotypic changes. Overall, we found changes in males and females across behavioral tests of locomotion, food motivation, and aggression following seven days of social isolation. There were no significant effects of social isolation on lifespan in either males or females.
Our findings suggest a potential female shift in macronutrient preference from protein to sugar following social isolation. This is implicated by isolated females’ decreased exploration of the “yeast side” of the arena and increased preference for sucrose over arabinose. Isolated males were less discerning, showing increased exploration of both yeast and apple cider vinegar and no change in sucrose choice compared to controls. It could be evolutionarily advantageous for females to prioritize sugar over protein consumption while under the environmental stress of isolation, to favor survival over reproduction. Mated females have been shown to increase yeast consumption for the benefit of their offspring to the detriment of their own stress resistance and lifespan, suggesting a fitness-fecundity trade off [35].
We next focused on the sex-specific effects of social isolation on gene expression, and the potential stress-signaling pathways in which these genes participate. Using RNA-seq, we performed differential expression analysis and found effects of sex and social isolation on the expression of several genes related to metabolic processes and storage proteins. Specifically, we identified 105 genes that were significantly upregulated in isolated females compared with isolated males, which included genes encoding fat body-synthesized storage proteins and the related protein takeout (to). Ultimately, we focused on to as a target gene due to its previously characterized expression patterns and behavioral regulatory functions in the literature.
To was originally characterized as a gene encoding a secretory protein that is mainly expressed in the fly head [36] and mediates circadian feeding behavior [37]. It is also expressed in organs related to feeding and olfaction and is upregulated in these structures in response to starvation, with to mutants dying shortly following the onset of starvation [36]. Additionally, it has been found that to expression is regulated by sex-specific factors. In the pericerebral fat body surrounding the brain, the male-specific forms of Doublesex (DsxM) and Fruitless (FruM) proteins activate to expression, but the female-specific form of Doublesex (DsxF) protein suppresses it [23]. This is consistent with our to TPM data from the male and female controls (Fig 3D). Moreover, to has been associated with male courtship behavior, which significantly decreases upon feminization of fat body cells via expression of the female Transformer protein (TraF) in males [23, 38]. To the best of our knowledge, the effect of social isolation on to expression has not been studied.
Upon manipulating to expression with various GAL4-UAS combinations, we found that either up- or downregulating to in to-expressing cells decreases sucrose preference in socially isolated females. However, these same manipulations produce either increases or no changes in sucrose preference in socially isolated males. Taken together, these results suggest that takeout may function in isolated females like a neuromodulator; when either downregulated or upregulated in to-expressing cells, in the context of social isolation, sucrose preference is suppressed. The opposite behavioral response was observed in males.
When to expression was suppressed in neurons, we observed no effects. Based on these results, the sex-specific effects of to on sucrose preference following social isolation may be localized to non-neuronal cells within or in the periphery of the brain. It is possible that the isolated-female specific overexpression of to in brains from our qPCR analyses could be attributed to gene expression changes in glia or adipose cells that were not adequately removed from the brain tissue. In the future, it would be worthwhile to test whether modifying to expression in glial or fat body cells would replicate the results observed when to was manipulated in all to+ cells. Furthermore, it would be useful to repeat the to manipulation in neurons using alternative pan-neuronal GAL4 drivers to elav-GAL4, such as nSyb-GAL4 or ChAT-GAL4 [39], to strengthen our conclusion of a non-neuronal mechanism.
One possibility of how to could influence behavior is through the binding of juvenile hormone, a fly hormone that canonically regulates developmental transitions [40–42]. The protein encoded by to is highly structurally similar to juvenile hormone binding protein (JHBP), and it contains a putative binding domain for juvenile hormone [36, 37]. The secretion of takeout into the hemolymph from the pericerebral fat body and its binding of juvenile hormone could either hinder juvenile hormone signaling, by reducing levels of the hormone actively circulating, or facilitate juvenile hormone’s action on its targets via carrier protein function.
Limitations and future directions
Although the present study has promise for elucidating the molecular mechanisms of sex-based responses to social stress, it is important to note its experimental and interpretational limitations. First, regarding the VBFD test, we quantified sucrose preference by categorizing the flies by crop color as primarily consuming sucrose, arabinose, both sugars (in roughly equal quantity), or nothing, and then calculating a proportion of sucrose-consuming flies out of all flies. The VBFD test does not provide absolute quantification of sucrose consumed, only preference for sucrose over arabinose. We also did not closely examine how flies were choosing. For example, for flies that consumed both sugars, we did not distinguish between flies that began eating arabinose and then switched to sucrose midway through the feeding period, or flies that were indiscriminately feeding between sucrose and arabinose during the entirety of the experiment. To gain a better understanding of how isolation might be influencing female feeding motivation and choice, it would be important to conduct a choice assay between protein and sugar. Future experiments incorporating macronutrient choice and using more rigorous approaches, such as the CApillary FEeder (CAFE) assay [43] to quantify food intake, are necessary to support the interpretation of an isolation-induced dietary switch in females.
Additionally, the present study did not examine sexual behavior or take sexual experience into account when considering the effects of social isolation on other types of behaviors. Tests of courtship and sexual receptivity could be added to examine how social isolation impacts male and female mating. These findings could inform whether isolation might influence female feeding preferences to shift resources away from reproduction. It should also be noted that reproductive state (e.g., virgin versus mated, or mated egg laid versus mated egg not laid) can strongly influence behavior, metabolism, and gene expression in Drosophila [44–47]. The added variability from reproductive state may account for why studies of social isolation have historically been conducted in male flies. In this study, we only tested mated females, but both virgin and mated females should be tested in the future to strengthen claims about the sex-specific effects of isolation on behavior.
Future experiments should also be conducted at more precise zeitgeber times (ZTs) to assess isolation phenotypes. Feeding, locomotion, and to expression are all strongly circadian-regulated [37]. However, the detection of behavioral differences despite testing across different circadian time points could indicate robustness of the observed phenotypic changes.
Regarding the transcriptional analyses, we validated our RNA-seq results by qPCR for selected genes. We used bulk RNA-seq, but performing single-cell RNA sequencing would be useful for determining if female-specific to overexpression is driven by a specific cell type.
Finally, our manipulations of to expression were limited in that they were broad and only focused on spatial control of gene expression. In our to–GAL4 x UAS-toRNAi flies, we knocked down to in all to-expressing cells of the body, not just the ones in the head that were detected in our RNA profiling, and not with respect to developmental timing. For future experiments, it would be helpful to perform immunohistochemistry or fluorescence in situ hybridization to determine where to is being overexpressed in the isolated female head or brain and then target manipulations to a more specific cell or tissue type.
Conclusion
Overall, chronic social isolation alters internal states and behaviors across animal species, and it may do so in a manner that differs by sex. We have found that social isolation in Drosophila impacts motivated behavior and sucrose choice differentially between males and females, and that these behavioral changes may depend on sex differences in the expression of the gene to in the fly’s head. One potential explanation for the increased expression of storage proteins and takeout following social isolation is that these changes allow female flies to tune their behavior to achieve an energetic balance appropriate for a specific environmental or social context. Despite the lack of an exact genetic ortholog for to in humans, it is also possible that takeout could serve as a functional analog to other mammalian proteins. For example, in mammals, lipoproteins in the brain and their metabolites have been found to regulate feeding behavior and energy balance [48]. Collectively, our findings support the use of Drosophila melanogaster as a model system for exploring female-specific molecular mechanisms related to stress-induced changes in food choice, energy metabolism, and possibly fertility.
Materials and methods
Fly stocks and rearing
All fly stocks were raised on standard medium (cornmeal/yeast/molasses/agar) at 25°C and 40% relative humidity on a fixed 12-h light/dark cycle. Male and female Oregon-R flies (wild type) and lines from the Bloomington Drosophila Stock Center (BDSC) were used for the majority of these experiments. Behavioral experiments were conducted within a ZT4-ZT8 time window. The following fly lines were obtained from the BDSC: to-GAL4 (#80938), elav-GAL4 (#8765), UAS-to (#81000), and UAS-toRNAi (#55982). Oregon-R and UAS-mCherryRNAi were gifts from Dr. Jodi Schottenfeld-Roames, Princeton University. The UAS-traRNAi (Vienna Drosophila Resource Center #2560, also known as traIR) stock was kindly provided by Dr. Rachel Monyak, Stonehill College.
Experimental setups and design
Social isolation.
For the social isolation stress model, we used an experimental procedure for chronic social isolation already characterized in the literature [6]. Newly-eclosed male and female flies were housed together for three to five days to acquire social experience. Flies were then housed either singly or in same-sex groups of 25 for seven days. Within 24 hours of the final day of isolation or group housing, flies were either frozen on dry ice for RNA extraction, tested for behavior, or moved to/kept in same-sex group housing for survival analysis.
Survival.
To measure survival, flies (n = 50 per experimental group) underwent social isolation or group-housed control conditions for seven days, following the previously described protocol. At the end of the stress period, all flies were moved into group housing in standard food vials. The flies were counted once per day, and any deaths were recorded, along with the date of death, until all flies were deceased.
Behavioral tests
Locomotion assay.
Flies were transferred to the testing chamber via aspiration or brief anesthesia with CO2. Upon analysis, we found that the use of CO2 did not alter activity when compared to aspiration. Each chamber consisted of a 55mm diameter petri dish filled with 1% agarose gel, to provide the flies with enough room to walk but not fly. The flies were put into same-sex groups of 10, recorded for 10 minutes, and frozen on dry ice. Videos were analyzed using EthoVision XT video tracking software (Noldus Information Technology, VA).
Motivated behavior tests.
Flies were tested in the arena previously described in “locomotion assay” to evaluate anhedonia-like behavior or discrimination of a food source. Flies were briefly anesthetized using CO2 and lined up along the midline of a filter paper evenly divided into two halves: one with yeast paste and one with nothing. Approximately 20 same-sex flies were tested in each group. Behavior was recorded for 60 minutes, and a preference index (PI) was calculated at five minute intervals as PI = [(# of flies on the yeast side) – (# of flies on the empty side)]/ (total). This test was repeated using apple cider vinegar in place of yeast paste, as an additional appealing food stimulus.
Aggression assay.
To test aggression, we modified an experimental protocol previously described by the Kravitz lab [15]. Testing occurred within 24 hours following the end of social isolation or crowding stress. Twelve-well plates were used as behavior chambers for testing same-sex, same-experimental condition pairs of flies. For group housed controls, flies in a fighting pair were taken from different vials to account for social novelty. To create an arena/territory, white plastic microtube caps were filled with standard fly food, topped with a dot of yeast paste, and placed in the wells. A transparent plastic covering with drilled holes was placed on top of the well plate, so that flies could be gently aspirated into the wells at the same time. The holes were then plugged with short screws. Fights were recorded for one hour using a Sony Digital Handycam camcorder (Sony Group Corporation, Tokyo, Japan).
Only behavior that took place in the arena was scored. Flies that did not encounter each other on the food cup within 45 minutes of recording were excluded from analysis. A goal sample size of n = 15 fighting pairs was decided prior to the start of experiments. For the fighting pairs, the times of the first encounter on the food cup and the first hit were recorded to calculate latency to initiate fighting. Fights were then scored for 10 minutes following the first hit (lunge for males or head butt for females). The male lunge was defined as a fly rearing up on its hind legs and snapping downwards onto its opponent. The female head butt was defined as a fly thrusting forward horizontally and then recoiling from the opponent. Number of hits, latency, number of overall encounters, and hits as a percentage of all encounters were used as measures to quantify and interpret aggressive behavior.
Value-based feeding decision (VBFD) test.
The value-based feeding decision (VBFD) test [17] was used as a two-choice assay to examine feeding decision making. Flies were briefly anesthetized with CO2 and placed in the center of the arena used for the locomotion and motivated behavior tests, a 55 mm diameter petri dish filled with 1% agarose gel and topped with a layer of parafilm. Sugars were dissolved in plain water to make two solutions: a 150 mM sucrose solution and a 150 mM arabinose solution. Each solution was labeled with a blue dye (0.01% erioglaucine disodium salt, Acros Organics, Cat# 229730250) or red dye (0.1% Food Red No. 106, TCI, Cat# F0143). Alternating droplets (20 μL volume) of each liquid food were placed evenly around the perimeter of the dish, such that there were eight droplets total, four of each solution. Flies were allowed to explore the arena and feed in the dark for two hours. At the end of this period, flies were collected and counted under a microscope. They were scored by color: blue (ingested primarily blue food), red (ingested primarily red food), purple (ingested both foods), or non-eater (no color accumulation in the abdomen).
RNA extraction
For each sample, flies were decapitated on dry ice. RNA was extracted and purified from homogenized whole heads using a phenol chloroform organic extraction. Samples were also treated using either the RQ1 RNase-Free DNase kit (Promega Corporation) or the TURBO DNA-free Kit (Invitrogen) to address genomic DNA contamination. Total RNA samples were sequenced by the Princeton Genomics Core or used for downstream applications. The samples submitted for RNA-seq were extracted from 200 whole heads. To obtain brains, 20–40 Drosophila heads were dissected one at a time in cold Schneider’s medium. Brains were transferred to a microtube on ice containing 200 μL of Schneider’s medium, briefly spun down on a benchtop centrifuge, and then frozen at -80°C. Prior to homogenization in TRIzol (Invitrogen), samples were spun down, and the Schneider’s medium covering the brains was removed.
Quantitative Polymerase Chain Reaction (qPCR)
RNA was reverse transcribed to complementary DNA (cDNA) using the SuperScript III First-Strand Synthesis System (Invitrogen). Random primers and a 200 ng total RNA template were used in a 20 μL reaction, following the instructions.
For real-time qPCRs, samples used SYBR Green PCR Master Mix (Qiagen or Applied Biosystems, Thermo Fisher Scientific). The PCR was performed in a 10 μL reaction including 3 μL of cDNA and 7 μL of master mix. The master mix for each gene consisted of 0.5 μL of each primer (10 μM), 5 μL SYBR Green, and 1 μL RNase-free water. Forward and reverse primer sequences for all genes can be found in Supporting Information (S2 Table). We used the following PCR conditions: initial incubation at 95°C for 15 min, followed by 40 cycles of 95°C for 15 s and 60°C for 1 min. Relative quantification was performed using the 2−△△CT method [49].
Transcriptome analysis by RNA Sequencing (RNA-Seq)
Sequencing.
RNA-seq was performed by the Princeton University Genomics Core Facility. The integrity of total RNA samples was assessed on the 2100 Bioanalyzer system using the RNA 6000 Pico Chip (Agilent Technologies, CA). The poly-A containing RNA transcripts were enriched from one microgram of total RNA for each sample using oligo-dT beads, further fragmented, and converted to cDNA and Illumina sequencing libraries using the PrepX RNA-seq library preparation protocol on the Apollo 324 NGS Library Prep System (Takara Bio, CA). Different DNA barcodes were attached to each library. The RNA-seq libraries were examined on Agilent Bioanalyzer DNA High Sensitivity chips for size distribution, quantified by Qubit fluorometer (Invitrogen, CA), and pooled at equal molar amounts. Each library pool was denatured and sequenced on one Illumina NovaSeq 6000 S Prime lane using the 100 cycle v1.5 kit. Raw sequencing reads were filtered by Illumina NovaSeq Control software, and only the Pass-Filter (PF) reads were used for further analysis.
Analysis pipeline.
Sequences were aligned to the Drosophila melanogaster genome (BDGP.46) [50,51] using STAR (v. 2.7.10b) [52]. Duplicates were removed using the Picard Toolkit (v. 2.27.5) [53] MarkDuplicates function. The number of reads mapping to each gene was determined using htseq-count (HTSeq v. 2.0.5) [54]. Differential gene expression analysis was performed in R (v. 4.4.0). All read counts from rRNA and pseudo-rRNA genes were removed. Normalization was performed using edgeR (v. 4.4.1) [55]. Multidimensional scaling (MDS) plots were used to identify outlier samples, which were then removed from the analysis. Data was prepared for linear modeling with limma (v. 3.60.6) [56] using voom [57]. The design matrix was constructed to model the relationship between gene expression and condition (Female-SI, Female-Control, Male-SI, Male-Control) adjusted for batch. Contrasts were used to compare gene expression differences between the following groups: 1. SI and control females, 2. SI and control males, 3. SI vs controls ((SI Female + SI Male)/(Control Female + Control Male). A fourth contrast was included to compare the difference between SI Female and Control Female with the difference between SI Male and Control Male. Differentially expressed genes were filtered based on an FDR-adjusted p-value of < 0.05 to determine significant genes with positive or negative values of log2FoldChange when comparing socially isolated males to control males, socially isolated females to control females, and the interaction between social isolation and sex. Heatmaps and additional plots were created using the R packages ComplexHeatmap (v. 2.21.2) [58] and ggplot2 (v. 3.5.1) [59].
Statistical analysis
Statistical analyses were performed using GraphPad Prism 10.4.0 software (GraphPad Software, Inc., La Jolla, CA). The Shapiro-Wilk test of normality (passed at p > 0.05) was used to determine whether data followed a parametric distribution. Unpaired t-tests were used for parametric data assuming equal variances, and Welch’s t-tests were used if the samples had unequal variances based on the result of the F-test. If the data did not pass the test for normality distribution, nonparametric testing was used. All nonparametric data were analyzed with the Mann–Whitney U test for between-group comparisons. Effects were deemed statistically significant at p < 0.05. The Grubbs’ test was applied to identify and exclude outliers in the data. Log-rank tests were used to determine differences in survival curves, with p < 0.05 deemed statistically significant. A mixed-effects model (restricted maximum likelihood) or 3-way ANOVA were used to compare differences between stressed and control flies’ preferences for yeast or apple cider vinegar, respectively, between males and females over time. All statistics for two-way (S3 Table) and one-way (S4 Table) ANOVAs are provided in Supporting Information.
Supporting information
S1 Fig. Social isolation does not affect lifespan in male or female flies.
Log-rank tests showed no significant differences in survival between isolated males (n = 50) and group-housed control males (n = 50, χ² = 2.71, df = 1, p = 0.10) or isolated females (n = 50) and group-housed control females (n = 50, χ² = 0.25, df = 1, p = 0.62).
https://doi.org/10.1371/journal.pgen.1011985.s001
(TIFF)
S2 Fig. Multidimensional Scaling reflects similarity and dissimilarity within and between experimental groups.
One sample (that was most dissimilar from the other two) was removed from each of the female groups to improve target discovery. Key: open = control, solid = social isolation; blue = male, orange = female; square = batch 1, circle = batch 2.
https://doi.org/10.1371/journal.pgen.1011985.s002
(TIFF)
S3 Fig. Significant differentially expressed genes, when comparing the interaction between social isolation and sex, are connected in broader networks of protein interactions.
A) All connected interactions are depicted, with non-connected genes removed. We specified a setting of “medium interaction strength” in STRING. The thickness of the connecting line between genes roughly corresponds to the strength of data support for gene interaction. B) One of our clusters that showed the strongest functional enrichment included genes implicated in the mitotic cell cycle process (strength = 0.82, signal = 0.78, FDR: p = 0.0003). C) Another top cluster consisted of genes connected to piwi, with functional enrichments in cellular process involved in reproduction in multicellular organism (strength = 0.49, signal = 0.45, FDR: p = 0.006), piRNA metabolic process (strength = 1.31, signal = 1.72, FDR: p = 6.97e-08), and egg activation and regulation of oskar mRNA translation (strength = 1.60, signal = 1.62, FDR: p = 9.30e-07). D) The third cluster of interest contains three genes encoding secreted (strength = 0.76, signal = 0.83, FDR: p = 8.01e-05) storage proteins (strength = 1.89, signal = 0.70, FDR: p = 0.004): Lsp2, Lsp1alpha, and Obp99b. All four genes are categorized as hemocyanin/hexamerin (strength = 1.72, signal = 0.92, FDR: p = 0.0006).
https://doi.org/10.1371/journal.pgen.1011985.s003
(TIFF)
S4 Fig. Significant differentially expressed genes, when comparing the interaction between social isolation and sex, are enriched across biological process, local network clustering, and Drosophila phenotype terms relating to cell division, reproduction, and hemocyanin/hexamerin proteins.
A) Top biological process terms include mitotic cell cycle process and nuclear cell division. B) Top local network cluster terms include egg activation, piRNA metabolic process, and regulation of mitotic cell cycle phase transition. C) Top Drosophila phenotypes include abnormal meiotic cell cycle and female sterility.
https://doi.org/10.1371/journal.pgen.1011985.s004
(TIFF)
S5 Fig. Validation of the RNA-seq-derived target genes with qPCR indicates enhanced expression of takeout in socially isolated females relative to isolated males.
A) A comparison of relative expression fold change values reveals that takeout is overexpressed in females but not males following social isolation. The difference in expression between isolated and control flies is significantly greater in females versus males (t(4) = 6.14, p = 0.004). B) Social isolation enhances Lsp-2 expression in males and females (t(4) = 2.56, p = 0.06). C) Social isolation enhances Lsp-1α expression in males and females (t(4) = 2.21, p = 0.09). D) We did not find a statistically significant male-female difference in expression of Fbp1 following social isolation (t(4) = 1.71, p = 0.16). Data points represent individual cDNA samples for a particular condition relative to a control average. Error bars are ± SEM. **p < 0.01.
https://doi.org/10.1371/journal.pgen.1011985.s005
(TIFF)
S6 Fig. Takeout expression is elevated in the brains of females but not males following social isolation.
Using relative expression fold change as a measure of the magnitude of change in takeout (to) expression between the social isolation and control groups, we observed that the values for males are randomly distributed around no change [1], while the female values are largely positive and significantly higher compared to males (t(10) = 2.28, p < 0.05). Data points represent individual cDNA samples for a particular condition relative to a control average. Error bars are ± SEM. *p < 0.05.
https://doi.org/10.1371/journal.pgen.1011985.s006
(TIFF)
S7 Fig. Single cell RNA-seq data from the Fly Cell Atlas identify takeout (to) expression in neuronal, glial, and epithelial cell types.
A) Colors in the ribbon indicate that to expression was identified in that cell type through single nucleus RNA sequencing (snRNA-seq) performed by the Fly Cell Atlas project [27]. A redder color on the blue-red gradient indicates that a greater proportion of cells of that particular cell type express to. White means that there is no available snRNA-seq data available for that cell type. B) For neurons, to was expressed primarily in adult olfactory receptor neurons. C) For epithelial cells, to was expressed across numerous cell types of the gastrointestinal tract (crop, midgut, esophagus, and hindgut). D) For glial cells, to expression was most widespread within the glia of the adult antennal nerve.
https://doi.org/10.1371/journal.pgen.1011985.s007
(TIFF)
S8 Fig. Knock down of to via RNA interference (RNAi) in to-expressing cells significantly decreases sucrose preference in isolated females and increases sucrose preference in isolated males relative to their respective controls.
A) In ORE-R flies, our preliminary results indicated a female-specific increase in sucrose preference following social isolation (t(15) = 3.14, p = 0.007). B) Flies expressing GAL4 under the control of a to-specific promoter do not show differences in sucrose preference following social isolation (t(14) = 0.57, p = 0.58). C) For flies with a UAS enhancer sequence that induces RNAi targeted to to mRNA when activated, isolated females increase sucrose consumption relative to control females (t(21) = 3.41, p = 0.003). D) When RNAi targeted to to mRNA is induced in to-expressing cells, social isolation increases sucrose preference in males (t(14) = 2.31, p = 0.04) yet decreases it in females (t(15) = 5.35, p < 0.0001) relative to respective controls. *p < 0.05, **p < 0.01, ****p < 0.0001.
https://doi.org/10.1371/journal.pgen.1011985.s008
(TIFF)
S9 Fig. Overexpression of to in to-expressing cells results in no differences in sucrose choice between isolated flies and their respective controls.
A) In ORE-R flies, our preliminary results indicated a female-specific increase in sucrose preference following social isolation (t(15) = 3.14, p = 0.007). B) Flies expressing GAL4 under the control of a to-specific promoter do not show differences in sucrose preference following social isolation. C) For flies with a UAS enhancer sequence that induces to expression when activated, isolated females increase sucrose consumption relative to control females (t(14) = 7.86, p < 0.0001). D) When to expression is induced in to-expressing cells, social isolation does not alter sucrose preference in males (t(14) = 0.30, p = 0.77) or in females (t(15) = 1.59, p = 0.13) relative to respective controls. **p < 0.01, ****p < 0.0001.
https://doi.org/10.1371/journal.pgen.1011985.s009
(TIFF)
S10 Fig. Knock down of to in neurons does not alter sucrose preference across experimental groups.
A) In ORE-R flies, females increased their sucrose preference following social isolation (t(15) = 3.14, p = 0.007). B) Flies expressing GAL4 under the control of an elav-specific promoter do not show differences in sucrose preference following social isolation in either sex. C) For flies with a UAS enhancer sequence that induces RNAi targeted to to mRNA when activated, isolated females increase sucrose consumption relative to control females (t(21) = 3.41, p = 0.003). D) When RNAi targeted to to mRNA is induced in neurons, there is no difference in sucrose choice between socially isolated and control flies of either sex. **p < 0.01.
https://doi.org/10.1371/journal.pgen.1011985.s010
(TIFF)
S11 Fig. Overexpression of to in neurons does not yield sucrose preference differences between isolated and control flies of either sex.
A) In ORE-R flies, females increased their sucrose preference following social isolation (t(15) = 3.14, p = 0.007). B) Flies expressing GAL4 under the control of an elav-specific promoter do not show differences in sucrose preference following social isolation in either sex (males: t(16) = 0.30, p = 0.77; females: t(18) = 0.69, p = 0.50). C) For flies with a UAS enhancer sequence that induces to expression when activated, isolated females increase sucrose consumption relative to control females (t(14) = 7.86, p < 0.0001). D) When to expression is driven in neurons, there is no difference in sucrose choice between socially isolated and control flies of either sex (t(9) = 0.87, p = 0.41). **p < 0.01, ****p < 0.0001.
https://doi.org/10.1371/journal.pgen.1011985.s011
(TIFF)
S12 Fig. Isolated males and isolated females show no differences in sucrose preference relative to controls following knock down of tra in neurons.
A) In ORE-R flies, sucrose preference is higher in isolated females compared to control females (t(15) = 3.14, p = 0.007). B) Flies expressing GAL4 under the control of an elav-specific promoter do not show differences in sucrose preference following social isolation in either sex. C) For flies with a UAS enhancer sequence that induces RNAi targeted to tra mRNA when activated, isolated females (t(14) = 2.79, p = 0.01) and isolated males (t(14) = 3.91, p = 0.002) increase sucrose consumption relative to their respective controls. D) When RNAi targeted to tra is induced in neurons, there is no difference in sucrose choice between socially isolated and control flies of either sex (males: t(17) = 1.42, p = 0.17; females: t(17) = 0.91, p = 0.38). *p < 0.05, **p < 0.01.
https://doi.org/10.1371/journal.pgen.1011985.s012
(TIFF)
S13 Fig. Downregulation and overexpression of to and downregulation of tra align with expectations across experimental groups.
A) We performed qPCR to confirm genetic knockdown of to in the to-GAL4 x UAS-toRNAi flies and overexpression of to in to-GAL4 x UAS-to flies. To overexpression was clearer in control males and control females, which could be due to ORE-R isolated males and isolated females already having relatively high expression of to. B) Although the elav-GAL4 x UAS-to flies showed no change or negative expression of to relative to ORE-Rs, the elav-GAL4 x UAS-toRNAi flies showed downregulation of to across experimental groups. C) Genetic knockdown of tra in elav-GAL4 x UAS-traRNAi flies decreased tra expression relative to ORE-Rs. D) Genetic knockdown of tra in elav-GAL4 x UAS-traRNAi flies also decreased expression of to across groups. E) In elav-GAL4 x UAS-to flies, to expression was elevated relative to elav-GAL4 flies in control males and isolated males but not in females. F) In elav-GAL4 x UAS-to flies, to expression was elevated relative to UAS-to flies across all experimental groups. After completing the VBFD test, flies were frozen on dry ice and decapitated. RNA was extracted from whole fly heads and reverse transcribed into cDNA. All groups represent n = 3 samples/group. Error bars are ± SEM.
https://doi.org/10.1371/journal.pgen.1011985.s013
(TIFF)
S1 Table. Average group size and number of groups tested in the VBFD test for all experimental groups across all genotypes.
https://doi.org/10.1371/journal.pgen.1011985.s014
(XLSX)
S2 Table. List of PCR primers used for experiments.
https://doi.org/10.1371/journal.pgen.1011985.s015
(XLSX)
S1 Data. Underlying numerical data for survival analysis.
https://doi.org/10.1371/journal.pgen.1011985.s018
(XLSX)
S2 Data. Underlying numerical data for all figures included in the main text.
https://doi.org/10.1371/journal.pgen.1011985.s019
(XLSX)
S3 Data. Underlying numerical data for all figures included in Supporting Information.
https://doi.org/10.1371/journal.pgen.1011985.s020
(XLSX)
Acknowledgments
We would like to acknowledge Gillian Hilscher, Sarah Brown, and Ella DePaolo for their contributions to the development and execution of the behavioral tests. We thank Dr. Caroline Palavicino-Maggio (McLean Hospital/Harvard Medical School) for guidance regarding the aggression testing.
References
- 1.
WHO Commission on Social Connection. WHO Commission on Social Connection. https://www.who.int/groups/commission-on-social-connection 2025.
- 2.
CDC. Health effects of social isolation and loneliness. https://www.cdc.gov/social-connectedness/risk-factors/index.html 2025. 2025 June 6.
- 3. Ferguson CJ, Averill PM, Rhoades H, Rocha D, Gruber NP, Gummattira P. Social Isolation, Impulsivity and Depression as Predictors of Aggression in a Psychiatric Inpatient Population. Psychiatric Quarterly. 2005;76(2).
- 4. Yang YC, McClintock MK, Kozloski M, Li T. Social isolation and adult mortality: the role of chronic inflammation and sex differences. J Health Soc Behav. 2013;54(2):183–203. pmid:23653312
- 5. Vallée A. Sex differences in the impact of social isolation and loneliness on mortality. Public Health. 2025;246:105831. pmid:40561736
- 6. Li W, Wang Z, Syed S, Lyu C, Lincoln S, O’Neil J, et al. Chronic social isolation signals starvation and reduces sleep in Drosophila. Nature. 2021;597(7875).
- 7. Zhu J, Park K-C, Baker TC. Identification of odors from overripe mango that attract vinegar flies, Drosophila melanogaster. J Chem Ecol. 2003;29(4):899–909. pmid:12775150
- 8. Becher PG, Flick G, Rozpędowska E, Schmidt A, Hagman A, Lebreton S, et al. Yeast, not fruit volatiles mediate Drosophila melanogaster attraction, oviposition and development. Functional Ecology. 2012;26(4):822–8.
- 9. Becher PG, Hagman A, Verschut V, Chakraborty A, Rozpędowska E, Lebreton S, et al. Chemical signaling and insect attraction is a conserved trait in yeasts. Ecol Evol. 2018;8(5):2962–74. pmid:29531709
- 10. Giglioli I. Insects and yeasts. Nature. 1897;56.
- 11. Ramos-Hryb AB, Ramirez MF, Lino-de-Oliveira C, Pagani MR. Stress-mediated hyperactivity and anhedonia resistant to diazepam and fluoxetine in drosophila. Stress. 2021;24(1):96–106. pmid:32319840
- 12. Wang L, Dankert H, Perona P, Anderson DJ. A common genetic target for environmental and heritable influences on aggressiveness in Drosophila. Proc Natl Acad Sci USA. 2008;105(15):5657–63.
- 13. Agrawal P, Kao D, Chung P, Looger LL. The neuropeptide Drosulfakinin regulates social isolation-induced aggression in. J Exp Biol. 2020;223(2).
- 14. Hoffmann AA. A laboratory study of male territoriality in the sibling species Drosophila melanogaster and D. simulans. Animal Behaviour. 1987;35(3):807–18.
- 15. Chen S, Lee AY, Bowens NM, Huber R, Kravitz EA. Fighting fruit flies: a model system for the study of aggression. Proc Natl Acad Sci U S A. 2002;99(8):5664–8. pmid:11960020
- 16. Nilsen SP, Chan Y-B, Huber R, Kravitz EA. Gender-selective patterns of aggressive behavior in Drosophila melanogaster. Proc Natl Acad Sci USA. 2004;101(33):12342–7.
- 17. Yu C-C, Chang F-C, Hong Y-H, Li J-C, Chen P-L, Chen C-H, et al. Assessing the cognitive status of Drosophila by the value-based feeding decision. NPJ Aging Mech Dis. 2021;7(1):24. pmid:34526491
- 18. Zhou Y, Zhou B, Pache L, Chang M, Khodabakhshi AH, Tanaseichuk O, et al. Metascape provides a biologist-oriented resource for the analysis of systems-level datasets. Nat Commun. 2019;10(1):1523. pmid:30944313
- 19. Szklarczyk D, Kirsch R, Koutrouli M, Nastou K, Mehryary F, Hachilif R, et al. The STRING database in 2023: protein-protein association networks and functional enrichment analyses for any sequenced genome of interest. Nucleic Acids Res. 2023;51(D1):D638–46. pmid:36370105
- 20. Van Dongen S. Graph clustering via a discrete uncoupling process. SIAM Journal on Matrix Analysis and Applications. 2008.
- 21. Sato JD, Powell DJ, Roberts DB. Purification of the mRNAs encoding the subunits of larval serum proteins 1 and 2 of Drosophila melanogaster. Eur J Biochem. 1982;128(1):199–207. pmid:6184227
- 22. Burmester T, Antoniewski C, Lepesant JA. Ecdysone-regulation of synthesis and processing of fat body protein 1, the larval serum protein receptor of Drosophila melanogaster. Eur J Biochem. 1999;262(1):49–55. pmid:10231363
- 23. Dauwalder B, Tsujimoto S, Moss J, Mattox W. The Drosophila takeout gene is regulated by the somatic sex-determination pathway and affects male courtship behavior. Genes Dev. 2002;16(22):2879–92. pmid:12435630
- 24. Benes H, Edmondson RG, Fink P, Kejzlarová-Lepesant J, Lepesant JA, Miles JP. Adult expression of the Drosophila Lsp-2 gene. Developmental Biology. 1990;142(1).
- 25. Zhu H, Ludington WB, Spradling AC. Cellular and molecular organization of the Drosophila foregut. Proc Natl Acad Sci U S A. 2024;121(11):e2318760121. pmid:38442150
- 26. Brown JB, Boley N, Eisman R, May GE, Stoiber MH, Duff MO, et al. Diversity and dynamics of the Drosophila transcriptome. Nature. 2014;512(7515):393–9. pmid:24670639
- 27. Li H, Janssens J, De Waegeneer M, Kolluru SS, Davie K, Gardeux V, et al. Fly Cell Atlas: A single-nucleus transcriptomic atlas of the adult fruit fly. Science. 2022;375(6584).
- 28. Guo X, Lv J, Xi R. The specification and function of enteroendocrine cells in Drosophila and mammals: a comparative review. FEBS J. 2022;289(16):4773–96. pmid:34115929
- 29. Castañeda-Sampedro A, Alcorta E, Gomez-Diaz C. Cell-specific genetic expression profile of antennal glia in Drosophila reveals candidate genes in neuron-glia interactions. Sci Rep. 2025;15(1):5493. pmid:39953089
- 30. Kremer MC, Jung C, Batelli S, Rubin GM, Gaul U. The glia of the adult Drosophila nervous system. Glia. 2017;65(4):606–38. pmid:28133822
- 31. Ogienko AA, Andreyeva EN, Omelina ES, Oshchepkova AL, Pindyurin AV. Molecular and cytological analysis of widely-used Gal4 driver lines for Drosophila neurobiology. BMC Genet. 2020;21(Suppl 1):96. pmid:33092520
- 32. Berger C, Renner S, Lüer K, Technau GM. The commonly used marker ELAV is transiently expressed in neuroblasts and glial cells in the Drosophila embryonic CNS. Developmental Dynamics. 2007;236(12).
- 33. Monyak RE, Golbari NM, Chan Y-B, Pranevicius A, Tang G, Fernández MP, et al. Masculinized Drosophila females adapt their fighting strategies to their opponent. J Exp Biol. 2021;224(Pt 6):jeb238006. pmid:33568440
- 34. Rideout EJ, Narsaiya MS, Grewal SS. The sex determination gene transformer regulates male-female differences in Drosophila body size. PLoS Genetics. 2015;11(12).
- 35. Liu C, Tian N, Chang P, Zhang W. Mating reconciles fitness and fecundity by switching diet preference in flies. Nat Commun. 2024;15(1):9912. pmid:39548088
- 36. Sarov-Blat L, So WV, Liu L, Rosbash M. The Drosophila takeout gene is a novel molecular link between circadian rhythms and feeding behavior. Cell. 2000;101(6).
- 37. So WV, Sarov-Blat L, Kotarski CK, McDonald MJ, Allada R, Rosbash M. takeout, a novel Drosophila gene under circadian clock transcriptional regulation. Mol Cell Biol. 2000;20(18):6935–44. pmid:10958689
- 38. Lazareva AA, Roman G, Mattox W, Hardin PE, Dauwalder B. A role for the adult fat body in Drosophila male courtship behavior. PLoS Genet. 2007;3(1):e16. pmid:17257054
- 39. Delandre C, McMullen JPD, Marshall OJ. Dynamic changes in neuronal and glial GAL4 driver expression during Drosophila aging. Genetics. 2025;229(3):iyaf014. pmid:39950543
- 40. Truman JW, Riddiford LM. The morphostatic actions of juvenile hormone. Insect Biochem Mol Biol. 2007;37(8):761–70. pmid:17628276
- 41. Riddiford LM, Ashburner M. Effects of juvenile hormone mimics on larval development and metamorphosis of Drosophila melanogaster. Gen Comp Endocrinol. 1991;82(2):172–83. pmid:1906823
- 42. Zhang X, Li S, Liu S. Juvenile Hormone Studies in Drosophila melanogaster. Front Physiol. 2022;12:785320. pmid:35222061
- 43. Diegelmann S, Jansen A, Jois S, Kastenholz K, Velo Escarcena L, Strudthoff N, et al. The CApillary FEeder Assay Measures Food Intake in Drosophila melanogaster. J Vis Exp. 2017;(121):55024. pmid:28362419
- 44. Nanfack-Minkeu F, Sirot LK. Effects of Mating on Gene Expression in Female Insects: Unifying the Field. Insects. 2022;13(1):69. pmid:35055912
- 45. Kohlmeier P, Zhang Y, Gorter JA, Su C-Y, Billeter J-C. Mating increases Drosophila melanogaster females’ choosiness by reducing olfactory sensitivity to a male pheromone. Nat Ecol Evol. 2021;5(8):1165–73. pmid:34155384
- 46. McGraw LA, Clark AG, Wolfner MF. Post-mating gene expression profiles of female Drosophila melanogaster in response to time and to four male accessory gland proteins. Genetics. 2008;179(3):1395–408. pmid:18562649
- 47. Akpoghiran O, Strich AK, Koh K. Effects of sex, mating status, and genetic background on circadian behavior in Drosophila. Front Neurosci. 2025;18:1532868. pmid:39844849
- 48. Wang H, Eckel RH. What are lipoproteins doing in the brain?. Trends Endocrinol Metab. 2014;25(1):8–14. pmid:24189266
- 49. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) method. Methods. 2001;25(4).
- 50. Adams MD, Celniker SE, Holt RA, Evans CA, Gocayne JD, Amanatides PG, et al. The genome sequence of Drosophila melanogaster. Science. 2000;287(5461):2185–95. pmid:10731132
- 51. Clark AG, Eisen MB, Smith DR, Bergman CM, Oliver B, Markow TA, et al. Evolution of genes and genomes on the Drosophila phylogeny. Nature. 2007;450(7167).
- 52. Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha S, et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics. 2012;29(1):15–21.
- 53.
Broad Institute. Picard Toolkit. https://broadinstitute.github.io/picard/ 2019.
- 54. Anders S, Pyl PT, Huber W. HTSeq—a Python framework to work with high-throughput sequencing data. Bioinformatics. 2014;31(2):166–9.
- 55. Robinson MD, McCarthy DJ, Smyth GK. edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics. 2009;26(1):139–40.
- 56. Law CW, Alhamdoosh M, Su S, Dong X, Tian L, Smyth GK. RNA-seq analysis is easy as 1-2-3 with limma, Glimma and edgeR. F1000Research. 2018;5:ISCB Comm J-1408.
- 57. Law CW, Chen Y, Shi W, Smyth GK. voom: Precision weights unlock linear model analysis tools for RNA-seq read counts. Genome Biol. 2014;15(2):R29. pmid:24485249
- 58. Gu Z, Eils R, Schlesner M. Complex heatmaps reveal patterns and correlations in multidimensional genomic data. Bioinformatics. 2016;32(18).
- 59.
Wickham H. ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag. 2016.