The following Drosophila strains were used in the study: repo-Gal4 (pan-glial cell marker, BDRC no. 7415), elav-Gal4 (pan-neuronal cell marker), tubGal80ts; repo-Gal4, tubGal80ts; elav-Gal4, UAS-ho (Cui et al., 2008), UAS-hoRNAi (Cui et al., 2008), UAS-cnc, UAS-cncRNAi, UAS-GFP.Valium10 (BDRC no. 35786), and wild type (CS). The strains UAS-ho and UAS-hoRNAi were kindly provided by Dr. Taketani (Kyoto Institute of Technology, Japan). Constructs are inserted on the 3rd chromosome in w1118 background.

To study the effect of overexpression/silencing of ho or cnc at different developmental stages, flies crossed with strains containing tubGal80ts were kept at different temperature conditions: (i) larva-specific, at 29°C from embryo to pupariation and then at 18°C; (ii) pupa-specific, at 18°C from embryo to pupariation followed by 29°C during the pupal stage and at 18°C after eclosion; (iii) adult-specific, at 18°C from embryo until eclosion and then at 29°C. For chronic overexpression/silencing, flies without tubGal80ts were kept constantly at 25°C. Control flies, both parental strains backcrossed to wild-type flies, were kept in the same temperature conditions as the experimental flies.

After eclosion, flies were separated according to sex and kept in groups of 30 individuals in specific temperature conditions. Every day the number of dead flies was counted, and every 3 days, flies were transferred to the new vials with fresh food. The experiment continued until the death of the last fly in the vial.

Males aged 7, 14, and 30 days-old were transferred into an empty vial without using CO₂ to avoid effects of this type of anaesthesia on locomotion and climbing ability. After a short recovery, flies were gently tapped to the bottom of a vial, and those that climbed vertically beyond the 5 cm marked line in 15 s were counted. The experiment was carried out in a dark room with red light only, which is invisible to flies. Every trial with 30 flies was repeated three times. Every experimental group had three repetitions.

The locomotor activity was recorded using a Drosophila Activity Monitoring System (DAMS, Trikinetics). Males, 2 days-old, were anesthetized with CO₂, placed on ice and then transferred into glass tubes (one fly per tube) sealed at both ends by food (agar, sugar, yeasts) and by a foam stopper. These tubes were placed in monitors (maximum 32 tubes per monitor) equipped with infrared light-emitting diodes and detectors that were connected to a computer. Whenever the fly passed the emitter/detector, the infrared beam was interrupted, transmitting a signal to the computer. To analyze the circadian period of locomotor activity, flies were maintained for 7 days under LD12:12 and next under constant darkness (DD) for 7 days. The locomotor activity was scored every 1 min. To study sleep, the activity of flies was analyzed on the second day of LD12:12 conditions. Sleep was measured as intervals of at least 5 min of inactivity.

Data were analyzed in Excel by using “Befly!” software (Department of Genetics, Leicester University). Lomb–Scargle normalized periodogram was used to determine rhythmic flies; flies with a period value lower than 10 (confidence level 0.05) were regarded as arrhythmic. Flies that did not survive until the end of each experiment were removed from analyses. Every experiment was repeated three times with at least 60 flies per group.

To compare ho or cnc expression between experimental groups, samples were collected at different developmental stages: the whole body of the 3rd instar larvae or pupae (5 individuals per sample, 3 repetitions), and heads of 5 days-old males and females (20 individuals per sample, 3 repetitions), separately. All samples were collected 1 h after the lights-on. We decided to examine gene expression in the whole body of larvae and pupae, and in heads of adults. In case of larvae we did not want to exclude motoneurons and glia located in the peripheral nervous system. There are about 700 neurons and 77 glial cells in each segment of the larval body (Bittern et al., 2021). Since we were not sure which cells in the central or peripheral nervous system were more affected in the experiments we decided to use whole body samples, although other tissues may have higher levels of both gene mRNA than neurons and glial cells. In case of pupae the nervous system is changed from larval to adult one and it is very difficult to dissect brains in early pupae. The experiments were carried out on wandering L3 larvae and in case of pupae in the pupal early stage when pupae changed color from white to brown (prepupae or P2 according to Bainbridge and Bownes, 1981). Total RNA was isolated using the Trizol method, purity was checked using NanoDrop and confirmed as acceptable with 260/280 and 260/230 ratios higher than 1.8. Total RNA (500 ng) was used for reverse transcription. Gene expression was measured using Kapa Sybr Green (KAPA Biosystems, Cape Town, South Africa) using specific primer sequences (ho: For 5′-ACCATTTGCCCGCCGGGATG-3′, Rev 5′-AGTGCGACGGCCAGCTTCCT-3′, gene accession no. CG14716; and cnc: For 5′-GAGGTGGAAATCGGAGATGA-3′, Rev 5′-CTGCTTGTAGAGCACCTCAGC-3′, gene accession no. CG43286) on StepOnePlus Real-Time PCR System. Rpl32 was used as a reference gene (For 5′-TATGCTAAGCTGTCGCACAAATG-3′, Rev 5′-AGCACGTGTATAAAAAGTGCCA-3′, gene accession no. CG7939). A standard curve was used to calculate gene expression levels. The number of target gene copies was normalized to the reference gene.

All data were examined for distribution normality, and statistical tests were chosen accordingly. The Kaplan–Meier curve was used to present data on the survival percentage of flies plotted against their age (in days). The survival (in %) in every phase of the aging process and the maximum lifespan between genotypes (i.e., experimental group versus control 1 and experimental group versus control 2) were analyzed using the log-rank test performed with the R/R Studio free statistical version 4.2.0.¹ The non-parametric analysis of variance (ANOVA) was used to analyse climbing performance and qPCR results (Dunn’s test). For locomotor activity and sleep analysis, the parametric ANOVA was used (Dunnett’s test). Statistical analyses were performed with the GraphPad Prism 7.05 software. Detailed statistics is provided in Supplementary Tables S1, S5, S7–S9.

The role of HO or its upstream regulator CNC in the brain during development was first examined on the survival pattern and maximum lifespan of adult flies. For developmental studies, the TARGET system is commonly used, however, flies are cultured at high or low temperatures which may affect their longevity. Universally, adult flies go through three phases of aging under an optimal temperature condition: health span (4–30 days after eclosion or DAE), transition phase (31–60 DAE), and senescence phase (61–120 DAE) (Phom et al., 2014). Here, we used 25°C as the optimal temperature for culturing of flies with chronic GAL4 activation, 18°C to inhibit Gal4 system, and 29°C to inactivate Gal80ts and activate cell-specific gene expression. For transgenic groups maintained at 18°C or 29°C, we used the following descriptions to identify the DAE range for each phase: (i) health span for those DAE with no natural death, (ii) transition phase for those DAE with a slight decline in the mortality curve showing 10% death, and (iii) senescence phase for those DAE with a steady decline in mortality curve represented by the window between the end of the transition phase till maximum lifespan of the fly (Ayajuddin et al., 2022). The DAE range for each phase could change after exposure to lower or higher temperature conditions than the optimal one (Mołoń et al., 2020). Effects of disrupting the expression of ho or cnc in the brain during development on the life expectancy of adults were analysed in each phase. In general, the adult longevity of Drosophila changed when ho or cnc mRNA level was changed during development. Interestingly, we observed opposite effects between chronic modification and those targeted at specific developmental stages (Table 1; Figures 1, 2; Supplementary Table S1). Survival was mostly better in adults when ho was either overexpressed or silenced chronically in the brain. Whereas low or unaffected survival was mainly detected when ho overexpression or silencing was explicitly done in the larval, pupal, or adult stage. We also observed differences in the effects of changing ho expression level in neurons or glia during development on adult longevity.

Flies with chronic ho overexpression in neurons (elav > ho) had higher survival during the health span in both sexes, but over time, changes did not vary with the parental controls, except in females during senescence phase (Figure 1A). Similarly, chronic ho overexpression in glia (repo > ho) resulted in a higher percentage of survival during the health span stage for both male and female adults (Figure 1B). Upregulation of ho in neurons during the larval stage had the same effect (Figure 1C), while in glia, it resulted in a significantly lower chance of survival in the health span to the transition phase for both sexes which extended in males until senescence phase (Figure 1D). Pupa-specific ho overexpression in neurons resulted in a lower percentage of survival in the health span and the transition phase but only for females (Figure 1E), while in glia, a lower percentage of survival was observed in similar pattern but with male adults (Figure 1F). Activating ho overexpression in neurons during the adult stage resulted in lower survival during the health span in males only (Figure 1G), while in glia, adults did not have any changes in their survival pattern except for males in senescence phase (Figure 1H).

Chronic ho silencing either in neurons (elav > hoRNAi) or glia (repo > hoRNAi) produced similar survival patterns. Female adults survived better before senescence phase, while male adults had a higher percentage of survival during the senescence phase (Figures 2A,B). Larva-specific ho silencing in neurons did not give any significant differences in survival patterns, except in males with a lower percentage of survival during the transition phase (Figure 2C). On the contrary, larva-specific ho silencing in glia had lower survival during the health span and the senescence phase for both males and females (Figure 2D). The survival pattern of adults was not changed after downregulating ho expression in neurons during the pupal stage (Figure 2E), while in glia, only female adults showed a significantly lower chance of survival during the transition phase (Figure 2F). Silencing ho in neurons during the adult stage resulted in higher survival in the transition phase (only males) (Figure 2G). On the other hand, pan-glial ho silencing during the adult stage exhibited lower chances of survival during the health span and the senescence phase for both male and female adults (Figure 2H).

Similarly, flies with modified cnc expression levels in the brain were observed to have opposite effects between chronic expression and those at specific developmental stages (Table 2). The survival of adults was better when cnc was either overexpressed or silenced continuously in comparison to those modifications activated specifically in the larval, pupal, or adult stage. The survival of chronic elav > cnc flies was significantly higher in male adults during health span while female adults during the senescence phase (Figure 3A). In chronic repo > cnc flies, the higher percentage of survival was almost detectable in all phases except for females in senescence stage (Figure 3B). Upregulation of cnc in glia during the larval stage did not produce any significant differences in survival patterns (Figure 3D), while in neurons, only males had lower percentages of survival during the transition to the senescence phase (Figure 3C). Pupa-specific cnc overexpression in neurons did not produce any significant differences, except in females with a lower percentage of survival during the transition phase (Figure 3E). On the contrary, female adults with pupa-specific cnc overexpression in glia had a lower percentage of survival in all phases (Figure 3F). Activating cnc overexpression in neurons during the adult stage did not produce any significant differences in survival patterns, except males in the senescence phase (Figure 3G), while in glia, a significantly lower survival was observed only in males during the transition phase (Figure 3H).

For chronic elav > cncRNAi flies, a higher percentage of survival was observed in all phases for female adults and only in the health span for males (Figure 4A). In chronic repo > cncRNAi flies, a better percentage of survival was detectable in the transition to the senescence phase for females only (Figure 4B). Larva-specific cnc silencing in neurons obtained a lower percentage of survival in the senescence phase (only females) (Figure 4C), whereas in glia, male adults had a lower percentage of survival in the health span stage and females in the senescence phase as well (Figure 4D). After downregulating cnc expression in the glia during the pupal stage, only males showed significantly lower survival in the transition to the senescence phase (Figure 4F). In contrast, cnc silencing in neurons at the pupal stage did not affect survival at all (Figure 4E). A significantly lower likelihood of survival was found during the transition phase for males and the senescence phase for females after silencing cnc in neurons during the adult stage (Figure 4G). In comparison, adult-specific pan-glial cnc silencing generated a lower percentage of survival only during the senescence phase for both sexes (Figure 4H).

The maximum lifespan of adults in each experimental condition reflected differences when the expression level of ho or cnc was targeted chronically or at a specific developmental stage (Supplementary Table S1). Under 25°C, the maximum lifespan in chronic experiments was 73–97 DAE. For adult-specific experiments, when flies were kept at 29°C during adult life, aging was faster (Mołoń et al., 2020). It may explain why adults survived only 32–52 DAE. For larva- and pupa-specific experiments that were kept at 18°C during their adult life, longevity generally increased (Mołoń et al., 2020). In this case, adults survived maximum 194 DAE. The effects of modifying ho expression level were not adequately strong to cause changes in the maximum lifespan of adult flies, regardless of the timing (chronically or at specific developmental stages) when overexpression/silencing was done. Most changes in the maximum lifespan were detected only in flies with ho silencing in the glia. Silencing of ho in the glia chronically, in pupae or adults did not increase survival compared with parental controls, while larva-specific ho silencing in the glia extended lifespan of adult flies. In the case of the chronic modifications in cnc expression, we found that flies with cnc overexpression in either neurons or glia survived significantly longer than controls. For flies with pan-neuronal overexpression at specific developmental stages, the flies treated during the adult stage showed a significantly shorter maximum lifespan compared to the control groups. It had a similar pattern with adult-specific repo > cnc. Also, both larva- and pupa-specific overexpression of cnc in glia did not survive longer than their parental controls. The downregulation of cnc in neurons or glia mostly led to a shorter maximum lifespan, regardless of which developmental stage was treated, except for chronic elav > cncRNAi which had an extended maximum lifespan. Differences in maximum lifespan between male and female adults were also observed. In most cases, females had longer maximum lifespans compared to males.

Changing the ho or cnc mRNA level, specifically during the larval, pupal, or adult stage, did not cause any significant differences in the climbing performance of adult flies (Supplementary Figures S2–S4 and Supplementary Table S5). The effects of ho or cnc expression modifications on the climbing behavior of flies were visible only after chronic overexpression or silencing of either ho or cnc in the brain. Flies with sustained upregulation of ho expression, elav > ho, did not show any changes in their climbing ability (Figure 5A and Supplementary Table S4), while repo > ho had a significantly lower percentage of climbing at 30 DAE, which were observed only in female adults (Figure 5B). For elav > hoRNAi flies, climbing defects were observed in males at 7 DAE, whereas in females, the effect was stronger as their climbing performance at 7, 14, and 30 DAE was decreased (Figure 5C). In repo > hoRNAi flies, only males had a lower percentage of climbing at 30 DAE (Figure 5D). In flies with chronic cnc overexpression, sex-dependent effects were observed at 30 DAE. Only a few males of elav > cnc were able to climb up the 5 cm demarcation line (Figure 5E), while the same effect was observed in repo > cnc females (Figure 5F). Continuous cnc silencing in both neurons and glia did not bring any significant results regarding climbing performance (Figures 5G,H).

The locomotor activity of flies after changing ho or cnc mRNA level in the brain is summarized in Table 3 (Supplementary Figure S6 and Supplementary Table S7). Significant differences in the total activity were mostly observed in flies with chronic modifications of ho expression in the brain. Only elav > ho flies showed identical activity patterns to their parental controls, while repo > ho flies were generally less active. In the case of elav > hoRNAi flies, they were typically more active. Whereas the total activity of repo > hoRNAi flies was not significantly different from the control groups. Flies with modifications in cnc expression in the brain showed changes in their locomotor activity compared to ho since we found significant differences not only in chronic experiments but also in those at specific developmental stages. In chronically treated elav > cnc flies, no significant difference was observed in total activity, while repo > ho flies exhibited a reduction in their total activity. In flies with chronic cnc silencing either in neurons or glia, the total activity was lower when compared to parental controls. When changes in cnc expression level were introduced at different developmental stages of Drosophila, we found higher total activity in larva-specific cnc overexpression in glia, larva-specific cnc silencing in either neurons or glia and adult-specific cnc silencing in glia.

Sleep was also checked in adults with modifications of ho expression during their development. We found significant differences in sleep when ho expression was upregulated or downregulated in the brain at different developmental stages (Table 4; Supplementary Table S8).

When ho was chronically overexpressed in the brain, changes in sleep patterns were significant compared to their parental controls. In elav > ho flies, daytime sleep was higher while nighttime sleep was unaffected (Figure 6A and Supplementary Table S8). On the other hand, daytime and nighttime sleep increased in repo > ho adults (Figure 6B). Pan-neuronal ho overexpression at specific developmental stages did not produce any significant results in sleep patterns. On the contrary, different responses were observed after overexpressing ho in the glia at specific developmental stages. Nighttime sleep was not affected by expression changes made during the larval, pupal, or adult stages. But then, opposite effects were observed regarding daytime sleep when pan-glial ho overexpression was induced at the larval or adult stage. Daytime sleep was reduced in flies with larva-specific ho overexpression in the glia, while it was increased in flies treated during the adult stage. In pupa-specific ho overexpression in glia, the daytime sleep was not changed at all. The effects of ho silencing in the brain on sleep in adults were rather weak (Figures 6C,D). We only found significant differences in daytime sleep in chronically treated elav > hoRNAi, which was shorter, and in adult-specific ho silencing in glia which had longer daytime sleep.

The chronic overexpression of cnc in neurons or glia brough similar results in daytime and nighttime sleep, which were longer in comparison to their respective parental controls (Figures 6E,F). On the other hand, continuous silencing of cnc in neurons or glia showed similar effects, increasing daytime and nighttime sleep (Figures 6G,H). Activating cnc overexpression at specific developmental stages generated different effects. The sleeping pattern was changed only when pan-neuronal overexpression of cnc was carried out at the pupal stage and daytime sleep was increased in this case. For pan-glial overexpression of cnc, we only found significant differences in sleep after overexpression at the larval stage (with decreased nighttime sleep), and at the adult stage (increased daytime sleep). On the other hand, continuous silencing of cnc in neurons or glia brough similar effects because daytime and nighttime sleep increased in both cases. It was the opposite after larvae-specific silencing of cnc in neurons or glia since both daytime and nighttime sleep was decreased. Pupae-specific downregulation of cnc in the brain reduced nighttime sleep, however, when directed to glia, it also increased daytime sleep. The partial suppression of cnc in either neurons or glia during the adult stage did not bring any significant changes in the sleeping behavior of adult flies.

To have baseline data about the transcriptional regulation of ho and cnc during development in Drosophila, we examined ho and cnc mRNA levels at different life stages of wild-type flies. Both ho and cnc mRNA levels were similarly maintained at higher levels during the larval stage compared to the pupal and adult stages under optimal temperature conditions (Figure 7A and Supplementary Table S8). At 29°C, there were no significant changes in ho and cnc mRNA levels between CS larvae (whole body), pupae (whole body), and adults (head) (Figure 7B and Supplementary Table S9). At 18°C, both ho and cnc mRNA were significantly higher in adults (head) compared to larvae (whole body) and pupae (whole body) (Figure 7C and Supplementary Table S9).

The mRNA level of ho was also examined in larvae (whole body), pupae (whole body), and adults (heads of males and females, separately) after modifying its expression chronically or at specific developmental stages. Surprisingly, we did not find any significant differences in ho mRNA levels in larvae, pupae, and heads of adults after changing ho expression, specifically during the larval, pupal, or adult stage (Supplementary Table S8). Nevertheless, we found a trend of ho mRNA increase in larvae, pupae, or adults in flies with ho overexpression and unaffected or lesser ho mRNA in flies with RNAi. However, the differences were not statistically significant. We only observed significant changes in ho mRNA levels in male and female adults (heads) after chronic expression. The continuous ho overexpression in both neurons and glia induced significantly higher ho mRNA levels in the heads of males and females (Figures 8A,B and Supplementary Table S9). On the contrary, we observed significantly high ho mRNA levels in the heads of female elav > hoRNAi and male repo > hoRNAi flies (Figures 8C,D and Supplementary Table S9). We suspected compensatory effects after partially suppressing ho in neurons or glia. We found that the high expression of ho originated from the retina (Figure 9 and Supplementary Table S9). In the brain, ho mRNA level did not change. It seems that the compensatory effect occurs when the compound eye is fully developed as we did not find any changes in ho mRNA level in larvae or pupae. However, it is worth mentioning that we manipulated ho expression only in specific cell types, neurons or glia, and this change can be difficult to detect in whole head samples.

We also quantified cnc mRNA levels in larvae (whole body), pupae (whole body), and adults (head) after the chronic modification of cnc expression in the brain, however, significant differences between experimental and control groups were only detected in adults. Chronic overexpression of cnc in neurons elevated cnc mRNA levels in both males and females (Figure 10A and Supplementary Table S9). In glia, cnc overexpression was only detected in females; however, we observed a similar trend in males (Figure 10B and Supplementary Table S9). The reduction of cnc transcript level was detected only in males and females after chronic cnc silencing in glia (Figure 10C and Supplementary Table S8). We found no significant differences when cnc was silenced in neurons (Figure 10D and Supplementary Table S9). Nonetheless, cnc mRNA levels were similar to controls.