Although sleep is widely conserved across the animal kingdom, different species can exhibit dramatically different amounts of sleep. Koalas, sloths, and brown bats, for example, can sleep for roughly 20 h/day, while other mammals, like horses and elephants, only sleep for 3-4 h each day. 1 The basic functions and regulatory mechanisms that drive such wide differences in sleep between species are poorly understood. Previous studies and meta-analyses examined traits that correlate with interspecies variations in sleep, identifying trends in diet, body size, or life history that are associated with total sleep measurements in vertebrate species. [1][2][3][4][5][6][7] Although these correlations shed light on selective pressures influencing sleep evolution, the feasibility of systematic comparisons and mechanistic studies across many related vertebrate species is limited due to various practical constraints. By contrast, the Drosophila genus provides a diverse range of at least 1,600 species, including the genetic model species D. melanogaster, many of which can be cultured and behaviorally monitored in standard laboratory settings. 8 Different Drosophila species thrive in a wide variety of environmental conditions across the planet, providing a set of natural experiments to explore the physiological adaptations that might be associated with variations in sleep. Exploiting this natural diversity to identify species with strongly elevated or reduced needs for sleep may provide new avenues to examine the fundamental functions fulfilled by sleep, neural signaling mechanisms that regulate sleep, and physiological tradeoffs that might be associated with different sleep strategies. Here, we examined differences in sleep between the genetic model species D. melanogaster and desert-adapted species, including D. mojavensis, a desert-dwelling species that shows heightened resilience to heat, starvation, and desiccation stresses. These features may contribute to their ability to thrive in harsh desert conditions, [9][10][11] but behavioral adaptations that accompany stress resilience in D. mojavensis remain unexplored.
We find that D. mojavensis exhibits increased sleep time across the day and night compared with D. melanogaster and that desert-adapted D. mojavensis flies respond to sleep loss with a homeostatic rebound. We observe several changes in sleep-or wake-related neuromodulator distribution: longsleeping D. mojavensis flies exhibit high levels of serotonin (5-HT), decreased abundance of wake-promoting octopamine (OA), and reduced numbers of cells expressing the circadian (legend continued on next page) output peptide pigment dispersing factor (PDF). Finally, we examine contributions of elevated sleep to stress resilience in D. mojavensis by measuring starvation and dehydration responses. Long-sleeping D. mojavensis flies exhibit extended survival during food or food and water deprivation compared with D. melanogaster, and individual sleep time of D. mojavensis correlates positively with survival time while flies are starved and dehydrated. Together, these results indicate that D. mojavensis exhibits an increased internal sleep quota relative to D. melanogaster and that elevated sleep may contribute to increased stress resilience in desert-adapted flies.
Characterizing high sleep time in Drosophila mojavensis D. melanogaster has become a popular genetic model system to study sleep and circadian rhythms. [12][13][14] Although focus on this model species permits the rapid development and proliferation of genetic tools and mechanistic frameworks, few studies have examined related species that are adapted to thrive in a variety of environmental conditions. Increased sleep is a behavioral adaptation that is hypothesized to support resistance to nutrient scarcity, 15 and artificial selection for starvation resistance in D. melanogaster can result in increased sleep time. 16 To test whether similar changes in behavior might correlate with interspecific changes in stress resistance, we compared sleep and starvation/dehydration responses in D. melanogaster and D. mojavensis. D. mojavensis are found in desert regions of Mexico and the southwestern USA and includes four geographically segregated subspecies: D. moj. mojavensis, D. moj. baja, D. moj. sonorensis, and D. moj. wrigleyi from the Mojave Desert, Baja California, Sonoran Desert, and Santa Catalina Island, respectively. [17][18][19] We measured sleep in all four D. mojavensis subspecies and in two wild-type stocks of D. melanogaster (Cs 20 and Pcf 21 ) using multibeam Drosophila activity monitors. Each D. mojavensis subspecies exhibits significantly elevated sleep throughout the day and night compared with D. melanogaster (Figures 1A and 1B). To test whether elevated sleep in D. mojavensis can be attributed to an elevated pressure to maintain and/or to initiate sleep episodes, we quantified the likelihood that a sleeping fly would awaken (P(wake); Figure 1C) or that a waking fly would fall asleep (P(doze); Figure 1D). 22 Each of the four D. mojavensis subspecies exhibits reduced P(wake) and elevated P(doze) compared with D. melanogaster, consistent with both strengthened sleep maintenance and an elevated pressure to fall asleep. Along with increased sleep time, D. mojavensis also exhibits reduced waking locomotor activity (Figure 1E), consistent with previous reports. 23 We detected similar differences between D. melanogaster and D. mojavensis in male flies: sleep time is elevated in male D. mojavensis compared with D. melanogaster during the day and night, although daytime differences are dampened because male D. melanogaster sleep more during the day than females (Figure S1).
We analyzed the cumulative distribution of bout lengths during the day to better detail sleep architecture in D. mojavensis (Figure 1F) and night (Figure 1G). These analyses found that D. mojavensis flies exhibit an elevated frequency of longer sleep episodes than those observed in either wild-type line of D. melanogaster. Since D. mojavensis sleep consists of longer bouts, D. moj. moj. and D. moj. baja continue to exhibit elevated sleep amounts when we increase the minimum period of quiescence scored for sleep from 5 min, as most commonly used for D. melanogaster, 12,13 to at least 60 min (Figure 1H). Together, these results indicate that elevated sleep in D. mojavensis consists of increased drive to fall asleep and prolonged sleep episodes.
To test for variations in sleep across days, we measured locomotion in D. moj. mojavensis flies across a 7-day period and found that daily sleep varies between individuals but remains relatively stable over time for single flies (Figures 1I and 1J). Because our baseline data reveal nearly identical sleep amounts and architecture between the four D. mojavensis subspecies, we have narrowed our focus for many of our additional behavioral studies on D. moj. mojavensis and D. moj. baja.
Because the D. mojavensis stocks that we describe above were derived from wild populations more recently than either of our lab-reared wild-type D. melanogaster stocks, we tested whether fly lines isolated from the wild sleep more than those reared in lab conditions for longer periods of time. To test whether recently derived stocks show increased sleep, we examined a D. melanogaster stock that originated from flies collected in the Westwood area of Los Angeles in 2021. D. melanogaster descended from recently wild Westwood flies showed comparable sleep amounts to Cs and Pcf laboratory (B) Day and night sleep totals for D. melanogaster (Cs, dark red; Pcf, light red) and D. mojavensis (blues). Two-way repeated measures ANOVA finds a significant genotype-by-time interaction (F (5,577) = 24.981, p < 0.0001). (C and D) P(wake) (C) and P(doze) (D) during the day and night for D. melanogaster (reds) and D. mojavensis (blues) stocks. Two-way repeated measures ANOVA detects a significant genotype-by-time interaction for P(wake) (F (5,579) = 75.43, p < 0.0001) and for P(doze) (F (5,553) = 5.628, p < 0.0001). (E) Waking activity (position movements/waking minute) is decreased in D. mojavensis subspecies (blues) relative to D. melanogaster (reds). Two-way repeated measures ANOVA finds a significant main effect of genotype (F (5,576) = 139.4, p < 0.0001). For (A)-(E), n = 101 Cs, 82 Pcf, 100 D. moj. moj., 100 D. moj. baja, 93 D. moj. sonorensis, 106 D. moj. wrigleyi. (F-H) Cumulative distributions for the duration of sleep bouts of D. melanogaster (reds) and D. mojavensis (blues) during the day (F) and night (G). (H) Depicts total sleep over 24 h for individuals from (F) and (G) using increasing periods for the minimum sleep bout threshold. Kruskal-Wallis tests find significant effects of genotype for day bout lengths (F; Kruskal-Wallis statistic = 467.7, n = 737-1,204 sleep bouts/group from 60 to 64 flies/group, p < 0.0001) and night bout lengths (G; Kruskal-Wallis statistic = 499.7, n = 398-1,380 sleep bouts/group from 58 to 64 flies/group, p < 0.0001). Two-way repeated measures ANOVA finds a significant genotype-by-threshold interaction (F (30,2244) = 6.142, p < 0.0001, n = 62-64 flies/group). (I and J) Sleep time course heatmap (I) and daily sleep totals (J) for D. moj. moj. female flies across a 7-day experiment (Friedman test statistic = 18.47, p = 0.0051, n = 28 flies). * indicates p < 0.05 for sleep on day 2 vs. day 5 and day 5 vs. day 7 by Dunn's pairwise test for (J). See also Figure S1. Group averages and error bars represent means and SEM for all panels. B C D E F Figure 2. Elevated sleep time in D. mojavensis across conditions (A) Sleep time course (left) and total sleep over 24 h (right) for Canton-S (dark red), Pcf (light red), and flies descended from D. melanogaster caught in Westwood, Los Angeles (open purple circles). ANOVAs detect a significant genotype-by-time interaction in sleep time course (F (94,3995) = 2.385, p < 0.0001) and main effect of genotype for total daily sleep (F (2,85) = 5.793, p = 0.0044; n = 38 Canton-S, 28 Pcf, and 22 wild-caught flies). (B) Sleep time course (left) and total sleep over 24 h (right) for D. melanogaster stocks (red) and two D. mojavensis subspecies (blue) from flies reared on Banana-Opuntia media. ANOVAs detect significant genotype-by-time interaction for the sleep time course (F (141,4841) = 8.838, p < 0.0001) and a significant effect of genotype for total daily sleep (F (3,103) = 91.08, p < 0.0001; n = 24 Canton-S, 27 Pcf, 28 D. moj. moj., and 28 D. moj. baja). (C) Sleep time course (left) and total sleep over 24 h (right) for D. melanogaster (red), D. arizonae (green), D. buzzatii (light green), D. mulleri (olive), and D. mojavensis (blues). ANOVA tests find significant genotype-by-time interaction for sleep time course (F (235,17531) = 18.07, p < 0.0001) and a significant effect of genotype for total sleep (F (5,373) = 215.4, p < 0.0001; n = 63 Canton-S, 78 D. arizonae, 62 D. buzzatii, 57 D. mulleri, 69 D. moj. moj, and 50 D. strains (Figure 2A). To test the impact of diet on D. mojavensis, we housed adult flies on media that included extract of opuntia cactus, a natural host for desert-adapted D. mojavensis. Their offspring developed in this media then continued to be fed the same diet as adults. Sleep in D. mojavensis remained elevated relative to D. melanogaster when both species were fed a banana-cactus diet (Figure 2B). In addition to D. mojavensis, several other related fly species, including D. arizonae, D. buzzatii, and D. mulleri, also live in deserts [24][25][26] (phylogeny schematic in Figure S2). As shown in Figure 2C, these three additional desert-adapted species sleep as much, or more, than D. mojavensis, suggesting that elevated sleep is not exclusive to D. mojavensis and could be conserved across the Repleta species that localize to desert regions. 27 D. mojavensis can show a preference for warm temperatures, 28,29 so we also observed flies while they were housed at either 27 C, 29 C, or 31 C for 3 days after a baseline day at 25 C. As shown in Figures 2D and S3, average daily sleep at each of the three warmer temperatures remained strongly elevated in D. mojavensis compared with D. melanogaster. In their desert habitats, D. mojavensis are exposed to environmental stressors that include temperature variations and periods of sparse food and/or water availability. To measure sleep during desert-like temperature fluctuations, we exposed both D. melanogaster and D. mojavensis flies to daytime temperature ramps. Flies were held at 25 C overnight, then began to progressively increase the temperature across the first 6 h of daytime to a peak of 35 C before reducing back to 25 C by lights off at ZT12. Although D. mojavensis maintained higher amounts of sleep than D. melanogaster across most of the day during these conditions (Figure 2E), both species showed a brief period of arousal when temperature peaked at 35 C at mid-day (Figure 2F). As the temperature decreased afterward, D. melanogaster briefly increased their sleep to comparable levels as the desert-adapted D. mojavensis subspecies. These results indicate that sleep in both species can be altered by variations in temperature but that D. mojavensis retain elevated levels of daily sleep under naturalistic daytime temperature conditions.
Elevated sleep in desert-adapted flies could indicate that this species has adapted an elevated need for basic functions that are fulfilled by sleep. To test whether desert-adapted D. mojavensis maintain an elevated sleep quota, we tested whether they respond to mechanical sleep deprivation with a homeostatic rebound. Vortex stimuli delivered for 3 s each minute were sufficient to strongly suppress sleep in D. moj. moj. (Figure 3A) and in D. moj. baja (Figure 3B). Following overnight deprivation, both D. mojavensis subspecies showed a recovery period of significantly increased sleep compared with baseline and regained approximately 20%-40% of their lost sleep after 24 h (Figure 3C). In the 24 h following deprivation, P(wake) is decreased during daytime on the first recovery day after deprivation, an indication of increased sleep depth (Figures S4A and S4B). Additionally, there was no decrease in locomotor activity per time awake (Figures S4C and S4D), indicating that waking locomotor activity is unimpaired by mechanical sleep deprivation. Following the first 24 h of recovery, D. mojavensis flies reduced their sleep nearly to baseline levels on the second recovery day (Figures S4E and S4F). Although D. melanogaster and D. moj. baja showed comparable sleep rebound after overnight deprivation, D. moj. moj. recovered a reduced amount of sleep relative to D. melanogaster (Figure 3C).
To test whether D. moj. moj. exhibit markers of increased sleep depth during recovery, we next probed arousability in recently deprived D. moj. moj. Flies were either left undisturbed, sleep-deprived for 12 h overnight (SD), or sleep-deprived and permitted 24 h of recovery (SD + 24 h) before they were exposed hourly to 60 s pulses of blue light. Light pulses were less likely to awaken sleep-deprived flies than rested controls; arousability returned to control levels in SD + 24 h flies (Figure 3D). After each light pulse, D. moj. mojavensis flies in the SD group had a reduced latency to fall back asleep compared with both the control and SD + 24 h groups (Figure 3E). These results indicate that long-sleeping D. mojavensis responds to mechanical sleep loss with homeostatic increases both in sleep time and intensity, consistent with the hypothesis that D. mojavensis have adapted an increased pressure for sleep.
To further probe responses of D. mojavensis to acute sleep loss, we also exposed D. moj. moj. and D. moj. baja flies to arousing blue light for 12 h overnight (ZT12-0). Overnight blue light disrupted sleep in both desert subspecies and was followed by prolonged rebound during the first recovery day (Figures 3F-3I). During light stimulation, D. moj. moj. lost 83.90% ± 3.50% (mean ± SEM, n = 35) of their sleep, while D. moj. baja reduced their sleep by 42.89% ± 3.74% (mean ± SEM, n = 53) (Figure 3H). Given that overnight light exposure significantly disrupted sleep, we next tested whether acute visual input bidirectionally influences sleep by housing D. mojavensis in 2 days of constant darkness. Both D. moj. moj. (Figure 3J) and D. moj. baja (Figure 3K) significantly increased their sleep when transferred to constant darkness after entrainment in a 12 h:12 h light-dark schedule. We found that in the absence of day-night light signals, the immediate increase in sleep during the subjective daytime persists across at least 2 days (Figure 3L). Previous observations of D. melanogaster have found either reduced or unchanged sleep when flies were housed in constant darkness, [30][31][32][33] indicating that light-dependent modulation of sleep differs between fly species.
Research over the past 20 years identified several neuromodulators and neuropeptides that influence sleep/wake regulation in D. melanogaster, [33][34][35][36][37][38] but interspecies variation of these signals across fly species is not well studied. In particular, we B C D E F G K L M H I J Figure 4. Interspecies variation of sleep-and wake-regulatory modulators between D. melanogaster and D. mojavensis (A-D) Relative LC-MS/MS quantification of 5-HT (A), octopamine (B), dopamine (C), and histamine (D) in heads of D. melanogaster wild-type stocks (reds) and D. mojavensis subspecies (blues). Data represent two independent experiments, each with three biological replicates per group (n = $100 heads/biological (legend continued on next page)
hypothesized that elevated sleep time in D. mojavensis may be correlated with an upregulation of sleep-promoting signals and a decrease in arousal pathways. To identify relevant neuromodulators, we conducted liquid chromatography-mass spectrometry (LC-MS) assays of fly heads from both D. melanogaster and D. mojavensis. We found that long-sleeping D. mojavensis flies from all four subspecies contain a significant increase in 5-HT and decrease of OA (Figures 4A and 4B), indicating a correlation in the abundance of these two neuromodulators with sleep time.
No uniform change in dopamine (DA) or histamine (HA) was measured between species (Figures 4C and 4D). 5-HT signaling promotes sleep in D. melanogaster 35,[39][40][41] and in vertebrates, [42][43][44] while OA, a paralog of norepinephrine, 45 drives arousal. 36,46 Changes in the abundance of 5-HT and OA between species may either indicate: (1) that altered numbers of neurons produce these modulators or (2) that conserved populations of cells have changed their rates of 5-HT and OA synthesis and/or release.
We observed the distribution of 5-HTergic cells by staining for the serotonin transporter (SERT) in D. melanogaster (Figure 4E), D. moj. baja (Figure 4F), and D. moj. moj. (Figure 4G). Images of the anterior and posterior cell bodies indicate that both species show similar overall patterns of 5-HTergic neurons, but it is possible that projection targets or cell numbers within specific clusters may vary. Similarly, we stained D. melanogaster, D. moj. baja, and D. moj. moj. brains for the OA synthesis enzyme Tdc2 (Figures 4H-4J) to observe the number and organization of OAergic cells. Our images reveal weak Tdc2-immunostaining in the anterior superior medial protocerebrum (ASM) neurons of the anterior protocerebrum of D. mojavensis flies (Figures 4H and S5A-S5C), a population of cells that underlies the wake-promoting role of OA. 46 Together, these results indicate that the distribution of 5-HTergic neurons is similar between species but that D. mojavensis may contain either weak signal or only a subset of the OAergic cells that are observed in D. melanogaster.
We next sought to test whether arousal circuitry might retain sensitivity to OA in long-sleeping species by microinjecting D. moj. baja females with 18.4 nL of either 20 mM OA or vehicle control. During the first 24 h after OA injections, we found that D. moj. baja females showed reduced sleep and increased locomotor activity (Figures 4K-4M) compared with vehicle-treated siblings. Although OA abundance is decreased in D. mojavensis, the wake-promoting effect of OA injection suggests that OA-sensitive arousal circuitry is likely conserved in desert-adapted flies.
To examine whether the distribution of other wake-promoting signals might differ between these two fly species, we performed immunostaining for the arousing circadian output peptide PDF. 49 Although D. melanogaster brains contain eight PDF-positive ventrolateral neurons (LNvs) in each hemisphere, four small LNvs (s-LNvs) and four large LNvs (l-LNvs) (Figure 5A), 50 careful analysis reveals inconsistent PDF-expression patterns between D. melanogaster and D. mojavensis. Specifically, D. mojavensis retained three to four PDF-positive l-LNvs but showed no s-LNv cell bodies or dorsal protocerebrum projections that were labeled with anti-PDF (Figures 5B and 5C). A loss of PDF immunostaining in s-LNvs has also been reported in other Drosophila species, indicating that selective pressures may drive reconfiguration of clock circuits as species adapt to different environments. [51][52][53][54] Together, these results indicate that elevated sleep of desertadapted D. mojavensis correlates with both an increase in sleep-promoting 5-HT and reductions of arousing OA and PDF.
Sleep in Drosophila mojavensis supports resilience to nutrient deprivation D. mojavensis sleeps more than D. melanogaster and responds to prolonged waking with a homeostatic rebound, indicating that this species may have an increased drive for sleep relative to D. melanogaster. To further test the functional relevance of heightened sleep pressure in desert-adapted flies, we also measured sleep and survival while flies were deprived of food alone or both food and water. Both Baja and Mojavensis subspecies of D. mojavensis survive longer than wild-type D. melanogaster when housed in glass tubes with non-nutritive agar media (Figure 6A) or in empty, dry glass tubes (Figure 6B), as described previously. 10 Although wild-type D. melanogaster suppress their sleep during food deprivation, 55,56 D. mojavensis instead show subspecies-specific changes. D. moj. baja exhibit moderate increases in sleep time during several days of food deprivation and awaken when both food and water are unavailable (Figure 6C). By contrast, D. moj. moj. show no significant sleep changes when food deprived and only a transient increase in sleep on the first day of food and water deprivation (Figure 6D). These trends are consistent with the hypothesis that elevated sleep in D. mojavensis is associated with prolonged survival during nutrient deprivation. We tested this relationship by depriving D. mojavensis females of food alone or both food and water, then housing them either in 12 h:12 h LD light or in constant blue light (LL) to disrupt sleep. Although constant light did not increase mortality in fed flies (Figure S6A), food-deprived D. moj. baja that were housed in constant blue light die from food deprivation more rapidly than siblings housed in LD (Figures 6E and S6B). replicate; squares represent data from experiment #1, triangles are from experiment #2). One-way ANOVAs find a significant effect of genotype for 5-HT (F (5,30) = 10.26, p < 0.0001), octopamine (F (5,30) = 9.488, p < 0.0001), and histamine (F (5,30) = 5.950, p = 0.0006) but no significant effect of genotype for dopamine (F (5,30) 6J) and subdividing the D. moj. moj. flies that were housed in LL during food deprivation revealed that the half of that group with the lowest sleep time during starvation died earlier than the half with the weakest sleep disruption (Figure 6K).
To more broadly examine whether sleep loss might render D. mojavensis flies more sensitive to starvation, we tested the correlation between average daily sleep and survival time in D. moj. moj. (Figure 6L) and, as with D. moj. baja, found highly significant positive correlations between daily sleep and starvation survival time. When D. moj. moj. were denied both food and water, LL exposure alone had no significant effect on survival time (Figure 6M). When the LL group of D. moj. moj. flies were sorted by daily sleep, we found that the half with the lowest amount of daily sleep during desiccation showed reduced survival time (Figures 6N-6O). Further, plotting individual daily sleep against desiccation survival time for D. moj. moj. that were housed in LD (filled dots) or in constant light (open dots) revealed a significant positive correlation across both experimental groups (Figure 6P). These data indicate that high amounts of sleep may confer desert-adapted flies with resistance to periods of insufficient food or water. We also detected significant negative correlations between survival and daily activity counts, indicating that the effect of sleep could, in part, be linked with decreased energy consumption during locomotion (Figure S6). Due to the significant correlation between waking activity intensity (counts/waking minute) and survival only for food-deprived D. moj. baja (Figure S6E) and not for food-and waterdeprived D. moj. baja or either D. moj. moj. condition (Figures S6H, S6K and S6N), it is likely that the influence of activity on survival can be linked with sleep amount and not necessarily changes in intensity of waking activity.
Periods of adaptive sleep loss have been reported in several vertebrate species, especially in birds 57 and marine mammals. 58,59 During these periods, it is thought that animals can acutely defer or offset the costs that accumulate from sleep loss. Here, we find that D. mojavensis exhibits an opposing behavioral strategy: they chronically show an elevated intrinsic sleep quota, even during periods of insufficient food. This adaptive strategy confers a survival advantage in conditions of hunger or thirst, supporting a functional role for sleep in maintaining efficient energy usage. [60][61][62] Similarly, recent studies found that flies show reduced metabolic rate while asleep 63 and that sleep is elevated in D. melanogaster artificially selected for starvation resistance. 16 Along these lines, it is possible that increasing daily sleep quotas, including during starvation, could protect D. mojavensis by slowing the usage of energy stores when food is not available. Alternatively, high amounts of sleep may allow desert-adapted D. mojavensis to allocate energy reserves specifically to necessary functions most efficiently fulfilled during sleep, 62,64,65 such as clearing metabolic waste, 66,67 managing oxidative stress, [68][69][70] or scaling synaptic connectivity. [71][72][73][74] The high sleep quota in D. mojavensis could indicate that increased metabolic investment in these sleep-restricted functions may be required to offset the costs of physiological adaptations made by desert-adapted flies that allow them to thrive in the desert environment. 9,11,18,[75][76][77][78] In either case, consistently investing large amounts of time to sleep suggests that D. mojavensis likely trade behavioral flexibility for starvation resilience. Because of their reliably high daily sleep quotas, D. mojavensis may provide new opportunities to examine beneficial functions of sleep at times of insufficient food or other (legend continued on next page) physiological stressors. Interestingly, another recent study found that other Drosophila species exhibit a range of homeostatic responses to sleep loss, 79 indicating that broad studies of Drosophila evolution could uncover interspecific adaptations in sleep need or function. Our characterization of increased sleep time in stress-resilient D. mojavensis provides a novel model species to examine the adaptive advantage(s) of elevated sleep and to investigate the evolution of sleep regulatory mechanisms across related species. Recent efforts to sequence the genomes of many Drosophila species have enabled the analysis of genetic correlates to environmental adaptations, 18,[80][81][82][83][84] but the contributions of altered behavioral strategies as populations adapt to environmental niches remain to be explored. We anticipate that combining genomic approaches with behavioral phenotyping across many species could identify common mechanisms that drive changes in sleep regulation and in the underlying functions of sleep across the Drosophila genus. In this case, examining flies that have evolved to withstand high desert temperatures and periods of nutrient deprivation could inform our understanding of the tolls of changing global climates on physiology and identify possible behavioral approaches for animals to cope with a warming world. [85][86][87] Our neurochemical and anatomical studies reveal correlations between sleep time and the abundance of two sleep/wakerelated neuromodulators, 5HT and OA, in D. melanogaster and D. mojavensis. Similarly, we find a restricted distribution of the wake-promoting circadian output peptide PDF in long-sleeping D. mojavensis flies. The similar distributions of serotonergic and octopaminergic neurons within D. melanogaster and D. mojavensis suggest that sleep circuit organization may be conserved between the species but that mechanisms governing neural activity or signaling dynamics could be differentially tuned as populations evolve. The availability of sequenced genomes for many Drosophila species, including D. mojavensis, 47,[88][89][90] may enable future studies to dissect neuromodulator signaling components with precise genetic tools similar to those already applied in D. melanogaster. These studies will be required to clearly test whether the altered 5HT and OA abundance directly alter sleep between species. Although the global organization of 5HT and Tdc2-expressing neurons appears largely similar between D. melanogaster and D. mojavensis, distribution of PDF expression in core circadian circuits differs. Similar to reports in several other Drosophila species, [52][53][54] we did not detect immunostaining for PDF in soma or axonal projections from s-LNvs, indicating that circadian circuit organization may commonly differ between fly species. Future studies will be required to examine the precise contributions of changes in each neuromodulator system to behavioral variations between species, and precisely examining each of these components may provide more insight into the functional importance of high sleep drive in D. mojavensis.
This study did not generate new unique reagents. Data and code availability d Source data have been deposited at Dryad and are publicly available as of the date of publication. DOI is listed in the key resources table. d This paper does not report original code. d Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request. EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS Fly rearing and stocks Fly stocks were cultured on standard cornmeal molasses media (per 1L H 2 O: 12 g agar, 29 g Red Star yeast, 71 g cornmeal, 92 g molasses, 16mL methyl paraben 10% in EtOH, 10mL propionic acid 50% in H 2 O) at 25 C with 60% relative humidity and entrained to a daily 12h light, 12h dark schedule. Experiments with Banana-Opuntia media used a recipe from the National Drosophila Species Stock Center (NDSSC; Cornell University): per 1L H 2 O: 14.16g agar, 27.5 g yeast, 2.23g methyl paraben, 137.5g blended bananas, 95g Karo Syrup, 30g Liquid Malt Extract, 22.33g 100% EtOH, 2.125g powdered opuntia cactus. 48 Locomotor activity was measured as the number of movements between beams per one-minute bins. Periods of sleep were defined by at least 5 minutes with no change in position within the multibeam activity monitors. Sleep time courses display 30-min time bins and X-axis time labels denote zeitgeber time (ZT) in hours after lights-on
Sleep deprivations were performed mechanically by mounting DAM5M activity monitors onto platform vortexers (VWR 58816-115). Individual tubes were plugged with food at one end and 3D-printed PLA plastic caps at the other. Monitors were vortexed at an intensity of 2.5g for 3-second pulses every minute through the duration of the 12-hour dark period. Arousability was tested in a darkened incubator with 60 seconds of blue light (luminance 0.048 Lv) every hour for 24 hours following sleep deprivation.
All flies were put in DAM5H activity monitors on standard food for baseline recording. After 2-3 days, control flies were transferred to tubes containing fresh food, food-deprived flies to tubes containing a 1% agar gel, and food-and-water-deprived flies to empty tubes plugged with foam at both ends. Flies immobile for at least 24 hours were defined as dead and data subsequent to their last full day alive was removed from sleep analysis.
Pharmacological microinjections 4-8 day old female flies were loaded into behavior tubes and monitored in DAM5M Activity Monitors to obtain baseline sleep and locomotor activity under 12h light: 12h dark (25 C). After 1-2 days of baseline in DAM5M monitors, flies housed in borosilicate tubes were placed on ice for anesthetization prior to injection using Drummond Nanoject II. For injection of exogenous neuromodulators, the anteriormost ocelli of D. mojavensis baja were injected with 18.4nl of 20mg/mL of Octopamine (Sigma-Aldrich, Catalog # O0250). For each round of injections, new OA is solubilized using Schneider's Drosophila Medium with L-Glutamine (Genesee Scientific, Catalog # 25-515). Following each individual injection, flies are returned back into individual borosilicate tubes, and placed in respective DAM5M Activity Monitors to continue sleep and activity surveillance for >48h.
ANOVA detects a significant time-by-strain interaction (F (376,17484) = 11.30, p < 0.0001). * represents time points at which Holm-Sidak pairwise comparisons find p < 0.05 between D. melanogaster and each of the four D. mojavensis subspecies. See also FiguresS2 and S3. Group averages and error bars represent mean and SEM for all panels.
Current Biology 34, 2487-2501.e1-e3, June 3, 2024
Current Biology 34, 2487-2501.e1-e3, June 3, 2024 e2