Flies were cultured at 25°C with 50–60% relative humidity and kept on a diet of yeast, dark corn syrup and agar under a 12-h light:12-h dark cycle. Tim-GAL4 (BL #7126); Clk.856-GAL4(BL #93198); Clk4.1 M-GAL4 (BL #36316); Clk4.5F-GAL4 (BL #37526) and UAS-Pdfr-RNAi (BL #42508) were obtained from the Bloomington stock center. per⁰¹, Pdfr⁵⁰³, C929-GAL4; and R6-GAL4 were a kind gift from Paul Taghert (Washington University in St. Louis).

A 30 cm X 2.5 cm aluminum runway was positioned between a hot plate and ice to generate a gradient ranging between 30°C and 17.9°C (Figures 1A,B). Clear Plexiglas walls and ceiling were coated with Fluon (Fisher Scientific #NC0533515) to prevent flies from avoiding contact with the runway by walking on the walls. The chamber was allowed to reach equilibrium for 10 min and the temperature range was verified using an infrared thermometer (Fisher Scientific, 15–077-968; Figure 1C).Thermal preference was assessed in a room at approximately 25°C, a relative humidity of ~50–65% using stable light illumination. Individual flies were then introduced into the chamber for the specified time and their precise temperature was recorded using the infrared thermometer. The temperature range of the thermal gradient was validated between flies. Unless stated otherwise, all experiments were conducted between 2:30–4:30 pm. All experiments were replicated by at least two independent investigators.

Sleep was assessed as previously described (Shaw et al., 2000). Briefly, flies were placed into individual 65 mm tubes containing the same food as they were reared on. All activity was continuously measured through the Trikinetics Drosophila Activity Monitoring System (¹Waltham, MA). Locomotor activity was measured in 1-min bins and sleep was defined as periods of quiescence lasting at least 5 min. All sleep experiments were replicated a minimum of two times.

Sleep deprivation was performed as previously described (Shaw et al., 2002; Seugnet et al., 2008). Briefly, flies were placed into individual 65 mm tubes and the sleep-nullifying apparatus (SNAP) was used to sleep deprive or sleep restrict flies Sleep deprivation was performed for 12 h during the dark phase (lights out to lights on). For sleep deprivation, the SNAP was activated once every 20 s for the duration of the experiment. Sleep restriction was performed for 48 h. The SNAP was activated for 18 s once every 15 min for 48 h, yielding a total of 192 stimuli lasting ~60 min; this regime both reduced and fragmented sleep as previously described (Klose and Shaw, 2019). To determine whether the stimulus induced by the SNAP was able to alter temperature preference flies were continually exposed to the SNAP for an equal number of exposures (Bang-Control). Sleep homeostasis was calculated for each individual as a ratio of the minutes of sleep gained above baseline during the 48 h of recovery divided by the total min of sleep lost during 12 h of sleep deprivation.

Short-term memory (STM) was assessed by Aversive Phototaxic Suppression (APS) as previously described (Seugnet et al., 2008, 2009). The experimenters were blinded to conditions. In the APS, flies are individually placed in a T-maze and allowed to choose between a lighted and darkened chamber over 16 trials. Flies that do not display phototaxis during the first block of 4 trials are excluded from further analysis (Le Bourg and Buecher, 2002; Seugnet et al., 2009). During 16 trials, flies learn to avoid the lighted chamber that is paired with an aversive stimulus (Filter paper is wetted with 10^-1M quinine hydrochloride). The performance index is calculated as the percentage of times the fly chooses the dark vial during the last 4 trials of the 16-trial test. In the absence of quinine, where no learning is possible, it is common to observe flies choosing the dark vial once during the last 4 trials in Block 4 (Seugnet et al., 2009). In contrast, flies never choose the dark vial 2 or more times during Block 4 in the absence of quinine (Seugnet et al., 2009). Thus, short term memory is defined as two or more photonegative choices in Block 4. For short term memory experiments following a 12 h sleep deprivation, the deprivation continued until evaluation by the APS. Power analysis using G*Power calculates a Cohen’s d of 1.8 and indicates that 8 flies/group are needed to obtain statistical differences (Seugnet et al., 2009).

All comparisons were done using a Student’s T-test or, if appropriate, ANOVA and subsequent planned comparisons using modified Bonferroni test unless otherwise stated. Note that a significant omnibus-F is not a requirement for conducting planned comparisons (Keppel, 1982). All statistically different groups are defined as *p < 0.05.

We evaluated temperature preference in adult Canton-S (Cs) flies using a slightly modified protocol from earlier studies (Sayeed and Benzer, 1996; Hamada et al., 2008). A schematic of the apparatus is shown in Figure 1A. A heat map of the runway floor is shown in Figure 1B. The temperature along the gradient, was measured using an infrared thermometer and is shown in Figure 1C. Importantly, the temperature, was stable, reproducible, and linear, with a slope of 0.6°C/cm, as previously described (Sayeed and Benzer, 1996). Previous studies have evaluated thermal preference in groups of 10–30 flies that are monitored for 20–30 min (Tang et al., 2017; Hague et al., 2020; Ito and Awasaki, 2022). However, we prefer studying individual flies in which (1) the sleep history of that individual has been well characterized, and (2) the individual fly can be retrieved at the end of the assay and their behavior can be further evaluated (Seugnet et al., 2009; Dissel et al., 2015b). Interestingly, when we evaluated individual flies, it appeared as if they settled down and chose a preferred temperature much sooner than 20–30 min. Thus, we examined temperature preference in individual flies in 3 min, 10 min, 15 min and 20 min intervals. As seen in Figure 1D, temperature preference did not change across intervals ranging from 3 to 20 min when tested individually. It is worth noting that the temperature preference of individual flies defined using our protocol closely matches that reported by Sayeed and Benzer (1996).

Evaluating temperature preference in individual flies for 3 min differs from previous studies. Thus, we asked whether we could replicate published findings with our approach. Previous studies indicate that flies display a daily temperature preference rhythm in which they select warmer temperatures across the light period (Kaneko et al., 2012). As seen in Figure 1E, when assessed after 3 min, individual flies select progressively warmer temperatures as the day progresses. Interestingly, previous studies indicate that temperature preference is similar between male and female flies run in groups. As seen in Figure 1F, no sexual dimorphisms were observed in our temperature preference assay when flies were assessed individually. To determine whether the temperature preference of an individual fly was stable, we examined temperature preference in individual flies over two successive day. As seen in Figure 1G, temperature preference is stable in an individual fly across days. A previous report indicates that flies select cooler temperature in the dark compared to siblings maintained in the light (Head et al., 2015). We find similar results using individual flies (Figure 1H). Finally, mutants for the receptor for the Pigment dispersing factor receptor (Pdfr⁵³⁰⁴) select warmer temperatures than controls; we find similar results when temperature preference is evaluated for 3 min in individual flies (Figure 1I). Together these data indicate that monitoring thermal preference in individual flies for 3 min accurately identifies thermal preferences and replicates published studies even when using a slightly different protocol.

To determine the effects of sleep loss on temperature preference, Cs flies were exposed to 12 h of sleep deprivation. As can be seen in Figures 2A,B, sleep deprivation was effective in keeping flies awake and Cs flies showed a typical sleep rebound (Shaw et al., 2002). Importantly, sleep deprived flies selected warmer temperatures in the thermal gradient compared to their untreated controls when tested between in the morning (8:30–10:30 am; Figure 2C). Flies also selected warmer temperatures than controls when sleep deprivation was extended into the afternoon and tested for thermal preference between 2:30–4:30 pm (25.4 ± 0.12°C vs. 23.8 ± 0.08°C, respectively, p < 0.05). We have previously shown that sleep deprivation disrupts short-term memory and that these memory deficits are completely restored after only 2 h of recovery sleep (Seugnet et al., 2008). Thus, we asked how long it would take for fly’s temperature preference to return to baseline. An independent cohort of flies was sleep deprived for 12 h during their primary sleep period. The effectiveness of the sleep deprivation is shown in Figure 2D. Importantly, siblings that were allowed to recover showed sleep rebound (Figure 2E). As seen in Figure 2F, sleep deprivation resulted in flies selecting warmer temperatures replicating the data shown in Figure 2C. Similarly to what we have observed for short-term memory, 2 h of recovery sleep was sufficient to return temperature preference to baseline levels (Figure 2F). Thus, sleep deprivation increases temperature preference in flies as it does in mammals.

We have validated an ethologically relevant sleep fragmentation protocol in flies that can disrupt sleep for extended periods (Klose and Shaw, 2019). The protocol is based upon observations that conditions that support brain plasticity increase daytime sleep-bout duration from ~10 min to ~25 min (Seugnet et al., 2008, 2011; Donlea et al., 2011, 2014; Dissel et al., 2015a). These data suggest that there is a minimum amount of sleep consolidation that is required for sleep to fulfil its functions. Thus, in our sleep fragmentation protocol, flies are kept awake for 1 min every 15 min to limit their ability to obtain restorative sleep. Sleep fragmentation is achieved using the sleep nullifying apparatus as previously described(Klose and Shaw, 2019). As seen in Figures 3A–C, sleep fragmentation modestly disrupts sleep, is effective in preventing long sleep-bouts and does not robustly activate sleep homeostasis. Importantly, sleep fragmentation did not disrupt short-term memory as assessed using Aversive Phototaxic Suppression (APS; Figure 3D). In the APS, flies are individually placed in a T-maze and must learn to avoid a lighted chamber that is paired with an aversive stimulus (quinine/ humidity; Seugnet et al., 2008). The performance index is calculated as the percentage of times the fly chooses the dark vial during the last 4 trials of the 16 trial test and short-term memory is defined as selecting the dark vial on 2 or more occasions during Block 4 (Dissel et al., 2015a,b,c). Despite having a normal short-term memory, flies exposed to 48 h of sleep fragmentation selected warmer temperatures in the thermal gradient compared to untreated siblings when tested in the afternoon (2:30–4:30 pm; Figure 3E). Temperature preference was also increased when flies were exposed to 48 h of sleep fragmentation and tested in the morning (8:30–10:30 am; 25.06 ± 0.06°C vs. 23.4 ± 0.12°C, p < 0.05). To exclude the possibility that the stimulus used to keep the animal awake altered temperature preference independently from sleep fragmentation, we exposed flies to the same number of stimuli that accrued during sleep fragmentation but in a consolidated block of 60 min during the light period and immediately evaluated temperature preference. As seen in Figure 3F, no changes in temperature preference were observed. Finally, we examined temperature preference during 2 h, 6 h and 24 h recovery from 48 h of sleep fragmentation. To facilitate comparisons across groups, the timing of recovery was staggered such that temperature preference was evaluated in all groups between 2:30 pm and 4:30 pm. Thus, sleep fragmentation ended at 12:30 pm and 8:30 am for the 2 h and 6 h recovery groups, respectively. The 2:30–4:30 pm time window is ideal since flies can choose either lower or higher temperatures at this time and thereby minimizes the impact of ceiling and floor effects present during the morning and evening time points (Figure 1E; Head et al., 2015). As seen in Figure 3G, temperature preference remained elevated even after 6 h of recovery. Thus, sleep fragmentation increases temperature preference in a thermal gradient and these effects persist for several hours.

The clock is comprised of 150 neurons that can be divided into two major groups. (1) Lateral neurons (LNd, sLNvs, lLNvs) and (2) Dorsal neurons (DN1, DN2, DN3; Helfrich-Forster, 2003; Taghert and Shafer, 2006; Ma et al., 2021). Given the role that clock neurons play in regulating temperature preference (Head et al., 2015; Goda et al., 2016), we evaluated their role in mediating the effects of sleep fragmentation. To begin we used RNAi to knock down Pdfr using the pan-clock drivers, tim-GAL4 and Clk856-GAL4. As seen in Figure 5A, sleep fragmentation produced modest changes in sleep time in all genotypes. Importantly, sleep fragmentation increased temperature preference in the thermal gradient for tim-GAL4/+, UAS-Pdfr^RNAi/+, Clk856-GAL4/+ parental controls. However, sleep fragmentation did not alter temperature preference in the thermal gradient in either tim-GAL4/+ > UAS-Pdfr^RNAi/+ or Clk856-GAL4/+ > UAS-Pdfr^RNAi/+ experimental lines (Figure 5B). Together these data indicate that the clock circuitry can influence the impact of sleep fragmentation on temperature preference via the Pdfr.

We next evaluated the role of Pdfr using GAL4 lines that express in DN1 neurons. DN1 neurons play a role in sleep regulation and modulate light-dependent temperature preference (Kunst et al., 2014; Guo et al., 2016; Lamaze et al., 2018; Schlichting et al., 2022). We used RNAi to knock down Pdfr using Clk4.1 M-GAL4 and Clk4.5F-GAL4. As seen in Figure 6A, sleep fragmentation produced modest changes in sleep time in all genotypes. Interestingly, sleep fragmentation increased temperature preference in the thermal gradient for Clk4.1 M-GAL4/+, UAS-Pdfr^RNAi/+ and Clk4.5F-GAL4/+ parental controls. However, sleep fragmentation did not alter temperature preference in the thermal gradient in either Clk4.1 M-GAL4/+ > UAS-Pdfr^RNAi/+ or Clk4.5F-GAL4/+ > UAS-Pdfr^RNAi/+ experimental lines (Figure 6B). These data indicate that the DN1s neurons can influence the impact of sleep fragmentation on temperature preference via the Pdfr.

Finally, we evaluated the role of the small (sLNvs) and large ventrolateral neurons (lLNvs) in modulating temperature preference after sleep fragmentation. The sLNvs and lLNvs play a role in regulating sleep and waking (Nitabach et al., 2006; Parisky et al., 2008; Sheeba et al., 2008; Chung et al., 2009; Shang et al., 2013). However, the relationship between the sLNvs and lLNvs and temperature preference is more complicated. Initial studies indicated that neither the sLNvs nor the lLNvs influence daytime temperature preference rhythms or light-dependent temperature preference (Kaneko et al., 2012; Head et al., 2015). However, Tang and colleagues report that the sLNvs play a role in setting preferred temperature before dawn (Tang et al., 2017). Interestingly, while pigment dispersing factor (pdf) is expressed in both the sLNvs and lLNvs; the Pdfr is only expressed in the sLNvs in unperturbed adult flies (Shafer et al., 2008; Im and Taghert, 2010). However, the Pdfr is re-expressed in the lLNvs of adult flies after sleep deprivation, sleep fragmentation and starvation (Klose and Shaw, 2019). Since Pdfr is only expressed in the lLNvs of adult flies after a perturbation, its role in regulating temperature preference remains unexplored. Thus, we used UAS-Pdfr^RNAi to knock down the Pdfr in sLNvs (R6-GAL4) and lLNvs (c929-GAL4). As seen in Figure 7A, sleep fragmentation produced modest changes in sleep time in all genotypes. Interestingly, sleep fragmentation increased temperature preference in the thermal gradient for R6-GAL4/+, UAS-Pdfr^RNAi/+ and c929-GAL4/+ parental controls (Figure 7B). However, sleep fragmentation did not alter temperature preference in the thermal gradient in either R6-GAL4/+ > UAS-Pdfr^RNAi/+ or c929-GAL4/+ > UAS-Pdfr^RNAi/+ experimental lines. These data indicate that both the sLNvs and the lLNvs can influence the impact of sleep fragmentation on temperature preference via the Pdfr.

Social jet lag is a misalignment between an individual’s internal biological clock and their social schedule. Social jet lag can occur when people wake up early for work during the week and then stay up later and sleep in on the weekends to make up for lost sleep (Juda et al., 2013; Pilz et al., 2018; Reis et al., 2020; Fischer et al., 2021). Social Jet lag can be modeled in Drosophila where it has been shown to dampen rhythmicity in most circadian neurons (Nave et al., 2020). Thus, we explored the effects of social jet lag on temperature preference in Cs flies. To induce social jet lag, the timing of lights-off is delayed by 3 h on Friday night and then shifted back 3 h on Sunday night to mimic a weekend schedule (Figure 8A, arrows). As can be seen in Figure 8A, sleep in flies exposed to Social Jet Lag closely resembles that seen in their age-matched untreated controls beginning on Monday after the light schedule has returned to normal to mimic a typical work week schedule. Indeed, quantification of sleep parameters, including total sleep time and sleep consolidation (average sleep bout duration during the day and average sleep bout duration at night) do not differ between flies exposed to social jet lag and their controls after being placed on a work-week schedule (Figure 8B). To further explore how social jet lag alters sleep, we evaluated metrics designed to measure sleep pressure P_Doze and sleep depth P_Wake (Wiggin et al., 2020). As seen in Figure 8C, social jet lag did not alter either metric compared to untreated siblings. Social Jet lag has been reported to disrupt short-term memory (Nave et al., 2020). We replicate that finding here (Figure 8D). In addition, to memory impairment, we show that flies exposed to social jet lag choose warmer temperatures in a thermal gradient when tested in the afternoon (2:30–4:30 pm; Figure 8D). Flies exposed to social jet lag also select warmer temperatures when tested in the morning (8:30–10:30 am; 23.4 ± 0.24°C vs. 24.2 ± 0.08°C, p < 0.05). Together, these data indicate that social jet lag disrupts both learning and memory and temperature preference.

To determine how long the effects of social jet lag might persist, we evaluated sleep, short-term memory and temperature preference on Friday. Recall that the light schedule had been returned to baseline on Sunday evening. As seen in Figures 8E,F measures of sleep time, sleep architecture and sleep depth are not altered by social jet lag compared to untreated controls. Despite normal sleep metrics, social jet lag treated flies shown memory impairments and increases in temperature preference several days after being placed on their typical weekday schedule (Figure 8G). These data indicate that social jet lag results in long-lasting changes to short-term memory as well as temperature preference.