What this is
- This research investigates sex differences in in Drosophila, focusing on the role of specific neuropeptides.
- It emphasizes the resilience of female circadian systems to disruptions in the pigment-dispersing factor () pathway.
- The findings suggest that female Drosophila maintain more stable activity-rest rhythms compared to males under similar genetic manipulations.
Essence
- Female Drosophila exhibit greater resilience in compared to males when disrupted by mutations in the pathway. This resilience may be due to a more distributed network of clock neurons in females.
Key takeaways
- Female are less affected by mutations in the and its receptor compared to males. This suggests that females maintain more stable rhythmicity despite genetic disruptions.
- CRISPR-Cas9-mediated manipulation of in specific neurons resulted in pronounced effects on male behavior, but not on females. This indicates a sex-specific response to genetic alterations in the circadian network.
- Speeding up the molecular clock in M-cells advanced the morning peak of activity in males but had no effect on females. This highlights the dominant role of M-cells in male circadian timing.
Caveats
- The study primarily focuses on Drosophila, which may limit the generalizability of findings to other species, including humans.
- The effects of other neuropeptides on were not fully explored, potentially overlooking additional factors influencing sex differences.
Definitions
- circadian rhythms: Biological processes that display an endogenous, entrainable oscillation of about 24 hours, influencing sleep-wake cycles and other physiological functions.
- PDF (pigment-dispersing factor): A neuropeptide released by specific neurons that plays a crucial role in regulating circadian rhythms in Drosophila.
Simplified
Introduction
Differences in neuronal circadian timekeeping between sexes remain relatively unexplored, despite the expanding body of research highlighting the influence of sex on the mechanisms underlying neuronal control of behavior [1]. In mammals, steroid hormones display daily, clock-driven changes in abundance, and while these sex hormones are not required to maintain rhythms, they differentially influence the amplitude of activity behavior between the sexes [1,2]. Furthermore, structural and functional sex differences have been observed in brain areas that receive direct input from the brain’s circadian timekeeping center [2,3]. Research in humans has also revealed significant sexual dimorphism: men tend to have lower-amplitude endogenous rhythms than women [4], are less resilient to nocturnal sleep disruptions, and spend less time asleep [5].
In mammals, the main circadian pacemaker resides in the suprachiasmatic nuclei (SCN), which in mice consist of a network of ~ 20,000 neurons (reviewed in [6]). The Drosophila circadian clock network has ~ 240 neurons and is the functional equivalent of the mammalian SCN [7–9] (reviewed in [10]). Each circadian clock neuron has an intracellular molecular timekeeping mechanism based on a transcriptional–translational feedback loop: the genes Clock (Clk) and cycle (cyc) promote rhythmic transcription of several key genes, including period (per) and timeless (tim), which build up daily and inhibit their own transcription [11]. Multiple kinases that act on components of these clock proteins and can affect the pace of the molecular clock have been identified. One such kinase is doubletime (DBT), which binds to and phosphorylates PER, regulating its nuclear accumulation [12,13].
The fly clock network consists of lateral neurons (LNs), which include ventrolateral (LNv), dorsolateral (LNd), and lateral posterior neurons (LPNs), as well as three groups of dorsal neurons (DN1, DN2, and DN3), some of which can be further subdivided [8,14–17]. The ventral and dorsal LNs are sufficient to produce the normal endogenous bimodal rhythm of sleep and activity [18,19]. The small ventral lateral neurons (s-LNvs) are usually referred to as morning cells (M-cells) since they control the morning peak of activity under light–dark cycles (LDs). These cells are also essential for maintaining rhythmicity under free-running conditions [20,21]. The evening peak is controlled by the LNds and a Pdf-negative LNv, the 5th LNv (E-cells) [18,22,23]. Some dorsal neurons (DNs) also contribute to the timing and amount of sleep via the modulation of M and E cells [24–27].
The release of the circadian neuropeptide pigment-dispersing factor (PDF) by s-LNvs is essential for endogenous circadian timekeeping. A Pdf null mutation, Pdf1, results in a substantial fraction of arrhythmic flies [20], desynchronization of molecular oscillations [28,29], and phase changes in the electrical activity of some clock clusters, most notably the LNds [30]. Loss of PdfR also leads to loss of behavioral rhythms [31–33]. Interestingly, PDF and PDFR also regulate behaviors that are sex-specific or sexually dimorphic. Rival-induced long mating durations require PDF expression in s-LNvs, PDFR expression in a subset of LNds, and NPF expression in LNds [34]. PDF controls rhythms in the sexually dimorphic pheromone profiles produced by oenocytes [35] and is involved in long-term mating suppression in males [36]. Both PDF and PDFR contribute to geotactic behaviors [31], and the phenotypes of Pdf1 mutants are sexually dimorphic, with males showing a more extreme negative geotaxis phenotype [37].
Sexual dimorphism in Drosophila sleep/wake cycles has been studied mostly under LD cycles. Males exhibit lower levels of activity and more sleep during the light phase [38–40]. This increase in midday sleep is due to the activity of a subset of sleep-promoting DN1s, which are more active in males [38] and receive input from the male-specific P1 neurons that control male courtship [39]. Unlike studies on circadian rhythms, Drosophila sleep research often involves only females. Males also have an earlier and more pronounced morning peak and a larger phase angle between the morning peak and the evening peak [41]. Under conditions of constant darkness and temperature (DD), males of several wild-type strains have a small but significant reduction in the free-running period (FRP) relative to females of the same strain [41]. Moreover, males are more likely to retain a bimodal activity pattern in DD [41]. A recent transcriptomic analysis of fruitless (fru)-expressing neurons revealed clusters that are enriched for circadian clock genes [42]. A previous study reported that DN1s express the male-specific FruM protein [43] and that the number of cells in the DN1ps cluster is sexually dimorphic [44]. In addition, the E3 subset of LNds has been shown to be dimorphic in its expression of the neuropeptide NPF [45,46].
Given the sexually dimorphic roles of neuropeptides, including PDF, in other behaviors [47], we asked whether females were similarly affected by manipulations of the Pdf/PdfR pathway. We found that female circadian rhythms are less affected by null mutations in both Pdf and PdfR and that similar effects are observed via CRISPR-Cas9-mediated Pdf mutagenesis, specifically in the LNvs. Moreover, speeding up the molecular clock in the LNvs via expression of DBTs leads to an advance of the morning peak in males but not in females, and the pace of the FRP of activity is significantly shortened only in males. Taken together, our results show that the female circadian system is more resilient to manipulations in the PDF pathway and suggest that Pdf+ neurons play a more dominant role in the male than in the female circadian network.
Results
Mutations in PDF and PDFR lead to sexually dimorphic phenotypes
A null mutation in Pdf results in pronounced behavioral phenotypes in Drosophila males [20]. We assayed the locomotor activity rhythms of Pdf1 females under free-running conditions (DD) and found that a large proportion of the experimental females were still rhythmic (Fig 1A–1C). The rhythmic power of experimental flies was significantly reduced in both sexes (Fig 1D), but the effect was less pronounced in females (Fig 1E), suggesting that Pdf1 females have more consolidated rhythms than Pdf1 males. We employed virgin females, as female rhythm strength has been shown to be significantly reduced after mating [48]. Mutant females that were rhythmic had a slightly, but significantly, shorter FRP than the controls (Table 1). This phenotype was not observed in experimental males (Table 1), consistent with a recent study [49]. Sleep cycles under DD also appeared to be more consolidated in females (Fig 1F). Pdf mutants have increased sleep, and this effect is mediated by PDF acting on the LNvs themselves [50]. We found that both sexes show an increase in total sleep in LD, but the effect was more pronounced in females (Fig 1G–1I). While the increase in sleep in males was most prominent at midday, females exhibited increased sleep throughout most of the light phase (Fig 1G). We are employing a Pdf1 mutant in w1118 background, and neither the advanced evening peak nor the short FRP of the Pdf null mutant are consistently observed. The lack of a short FRP in rhythmic Pdf1 mutant flies in this genetic background is consistent with a previous study [49].
To rule out the presence of remnant PDF expression in Pdf1 females, we stained the brains of control and experimental males and females with an anti-PDF antibody. We did not observe any traces of PDF in experimental flies of either sex, even with increased laser intensity (S1A Fig). PDF accumulates rhythmically in the dorsal termini of the s-LNv projections in a time-of-day-dependent manner both in LD and DD [20,21]. To determine if there were differences in the amplitude of PDF cycling between the sexes in a wild-type background (Canton-S), we dissected control males and females on the third day under DD at 6 time points over a 24-h cycle. Using a COSINOR-based curve fitting method [51], we found that both males and females have clear 24-h rhythms in PDF cycling in their dorsal projections, with no significant sex differences in amplitude (S1B and S1C Fig and Table 2).
Next, we asked whether the effects of a Pdf receptor mutation (PdfR) on activity and sleep were also sexually dimorphic. The expression of PDFR, a GPCR, can be detected in most clock neurons, with the exception of 3 LNds, half DN1ps, and some DN3s [52], which coincides with Cryptochrome expression in clock neurons [52]. The han5304 mutant is a PdfR hypomorph and exhibits Pdf1-like behavioral phenotypes under both LD 12:12 and DD [31–33]. Under DD, both Han males and females showed a significant reduction in rhythmicity compared with the controls, but there was a greater proportion of rhythmic females (~65%) than males (~16%) (Fig 1J). The FRP of the experimental flies was significantly shorter for both sexes (Table 3), as reported previously for males. Rhythmic power was significantly lower than that of the controls for both han5304 males and females (Fig 1K), but the effect was more pronounced in males, suggesting that females have more consolidated rhythms (Fig 1L). Similar to the effect of the Pdf mutation, han5304 flies showed significantly higher levels of LD sleep than controls, and this effect was also more pronounced in females (Fig 1M–1O). Taken together, these results suggest that female circadian rhythms are less affected by the loss of both Pdf and PdfR. However, the LD sleep phenotypes were more pronounced in females.
Mutations inandlead to sexually dimorphic phenotypes. Pdf PdfR Representative double-plotted actograms of(left) and(right) male flies subjected to 6 days of LD followed by seven days of DD.Representative actograms of(left) and(right) female flies subjected to six days of LD followed by seven days of DD.Percentages of rhythmic flies are plotted for control () andmales (= 86 (),= 81 ()) and females (right,= 75 (),= 101 ()). The error bars represent the SEM values plotted across three replicate experiments.Rhythmic power of control () andmales and females were calculated using the Chi-squared Periodogram.The differences in rhythmic power between experimental males and females and their respective controls are plotted.Average sleep plots of flies over seven days in DD are plotted for male and female control () and experimental () flies. The controls are plotted in gray, and the experimental males and females are plotted in blue and magenta, respectively.Average sleep plots under LD 12:12 for the control () and experimental () groups are plotted for males (left) and females (right). The plots are averaged over flies and days for a period of three days under LD 12:12. The controls are plotted in gray, and the experimental males and females are plotted in blue and magenta, respectively.Total sleep values under LD conditions are plotted for male (left) and female (right) control () and experimental () flies.The differences in total LD sleep values between experimental males and females and their respective controls are plotted.Percentage of rhythmic flies are plotted for control () andmales (= 86 (),= 115 ()) and females (= 74 (),= 94 ())Rhythmic power of the control () andmales and females calculated using the Chi-squared periodogram are plotted.The differences in rhythmic power between experimental males and females and their respective controls are plotted.Average sleep plots under LD 12:12 for control () and experimental () flies are plotted for males (left) and females (right). The plots are averaged over flies and days for a period of three days under LD 12:12. The controls are plotted in gray, and the experimental males and females are plotted in blue and magenta, respectively.Total sleep values under LD conditions are plotted for male (left) and female (right) control () and experimental () flies.The differences in total LD sleep values between experimental males and females and their respective controls are plotted. Statistical comparisons were performed between the control and experimental flies of both sexes using a Mann–Whitneytest and percentage of rhythmic flies were compared using Fisher’s exact test. The box plots extend from the 25th to 75th percentile, with whiskers extending from the smallest to the largest value, and each point represents data from a single fly. Combined data from at least three replicate experiments are plotted. *< 0.05, **< 0.01, ***< 0.001. (A) (B) (C) (D) (E) (F) (G) (H) (I) (J) (K) (L) (M) (N) (O) w Pdf w Pdf w Pdf n w n Pdf n w n Pdf w Pdf w Pdf w Pdf w Pdf w han n w n han n w n han w han w han w han U p p p 1118 1 1118 1 1118 1 1118 1 1118 1 1118 1 1118 1 1118 1 1118 1 1118 5304 1118 5304 1118 5304 1118 5304 1118 5304 1118 5304 https://doi.org/10.6084/m9.figshare.28601348.v1
| Pdf1 | |||||
|---|---|---|---|---|---|
| Genotype | n | % Rhythmicity ± SEM | E-peak phase ± SEM | Free-running period ± SEM | Rhythmic power ± SEM |
| w1118 (male) | 86 | 97.3 ± 2.66 | 11.92 ± 0.07 | 23.8 ± 0.04 | 97.8 ± 3.88 |
| Pdf01 (male) | 81 | 30.3 ± 5.05**# | 11.72 ± 0.07 | 23.72 ± 0.13 | 22.1 ± 3.24***# |
| w1118 (female) | 75 | 92.8 ± 3.27 | 11.71 ± 0.07 | 24.1 ± 0.04 | 109.9 ± 5.63 |
| Pdf01 (female) | 101 | 56.9 ± 3.63**# | 11.85 ± 0.08 | 23.9 ± 0.09 001 | 49.63 ± 4***# |
| Cosinor parameters for PDF cycling in s-LNv dorsal termini | |||||
|---|---|---|---|---|---|
| Genotype | p value | PR | Mesor ± s.e. | Amplitude ± s.e. | Phase ± s.e. |
| (male)Canton-S | 2.02e − 06 | 42.1 | 17.15 ± 2.67 | 22.38 ± 3.79 | −4.71 ± 9.64 |
| (female)Canton-S | 7.33e − 07 | 37.554 | 14.59 ± 2.9 | 24.47 ± 4.07 | −19.13 ± 9.65 |
| PdfR1 | |||||
|---|---|---|---|---|---|
| Genotype | n | % Rhythmicity ± SEM | E-peak phase ± SEM | Free-running period ± SEM | Rhythmic power ± SEM |
| w1118 (male) | 86 | 97.3 ± 2.66 | 11.68 ± 0.07 | 23.8 ± 0.03 | 103.1 ± 4.15 |
| PdfR01 (male) | 115 | 16.4 ± 1.64***# | 10.33 ± 0.04 | 21.8 ± 0.12***# | 17.46 ± 2.28***# |
| w1118 (female). | 74 | 92.7 ± 3.25 | 11.45 ± 0.07 | 24.1 ± 0.05 | 89.13 ± 4.75 |
| PdfR01 (female) | 94 | 65.4 ± 5.8***# | 10.36 ± 0.07 | 22.4 ± 0.07***# | 47.3 ± 3.5***# |
CRISPR-Cas9-mediatedmutagenesis has more pronounced effects on male behavior Pdf
In the Pdf null mutant, background effects could contribute to the sexual dimorphism observed in behavioral rhythms. We therefore employed a tissue-specific CRISPR-Cas9-mediated knockout of Pdf in both males and females, as described in a recent study that focused on males [53]. To assess the efficiency of the manipulation, we stained for PDF in flies that constitutively expressed Pdf gRNA and Cas9 in Pdf+ neurons. This experiment was conducted at 28 °C, as this temperature was more effective at mimicking the behavioral phenotypes of the Pdf1 mutant males.
PDF was reduced in the s-LNvs in both sexes (Fig 2A–2C), although in most experimental brains, we noted faint staining in the dorsal projections of at least one s-LNv in at least one brain hemisphere (Fig 2A). We quantified PDF intensity in the cell bodies of the s-LNvs and found that the signal intensity and the number of cells were reduced in both sexes in a similar manner (Fig 2B–2E). PDF expression within the large LNvs was less affected and could be detected in 2–3 l-LNv cell bodies in most brains (Fig 2A). In addition to behavioral phenotypes, Pdf1 mutation leads to pronounced misrouting of s-LNv projections in male flies [54]. We employed a transgene expressing a red fluorescent protein under the Pdf regulatory sequence [55] and observed faint projections occasionally defasciculating from the main bundle in one or both hemispheres in Pdf > Pdf-g; Cas9 flies of both sexes (Fig 2A, middle panels). To determine whether driver strength was similar between males and females, we analyzed nuclear signal levels in the s-LNvs of male and female Pdf > nls-mCherry flies and found no significant sex differences (S1D Fig). However, subtle sex differences in driver strength may be undetectable due to the constitutively high expression levels of the Gal4/UAS system.
We analyzed activity–rest rhythms in Pdf > Pdf-g; Cas9 flies and found that while experimental males were largely arrhythmic, the percentage of rhythmic experimental females was not different from the controls (Fig 3A and 3B). The FRP of the experimental males was not significantly different from that of the controls, but a wide range of periods were observed. Compared with parental controls, Pdf > Pdf-g; Cas9 females had significantly shorter FRPs (Fig 3C). The rhythmic power was significantly lower in experimental flies of both sexes (Fig 3D), but the effect was less pronounced in females (Fig 3E). Sleep under LD was not affected (S2A and S2B Fig), whereas sleep under DD was similarly increased in males and females, both for several days under DD and for DD1 only (S2C–S2F Fig). We also calculated the activity/waking minute for control and experimental flies and found that there are no significant differences in the waking activity for experimental flies of both sexes (S2G Fig).
We next restricted the CRISPR mutagenesis of Pdf to the small LNvs via a specific driver from the split-Gal4 collection generated by the Rubin Laboratory (SS00681-Gal4). Pdf knockdown in the s-LNv resulted in most males being arrhythmic (~30% rhythmicity), whereas the experimental females were ~57% rhythmic (S3A and S3E Fig and S1 Table). The percentage of rhythmic flies was significantly lower than that of both controls for both experimental males and females, but females were significantly more rhythmic (S3A Fig). The FRP of the experimental males and females was shorter than that of their respective control flies (S3B Fig). The rhythmic power of experimental males and females was lower than that of their respective controls (S3C Fig), but the effect was less pronounced in females (S3D Fig). This suggests that PDF from the s-LNv is important for the behavioral and sex-specific differences observed in Pdf1 mutants.
Tissue-specific CRISPR-mediatedmanipulation leads to a reduction in PDF levels and the misrouting of s-LNdorsal termini in both sexes. Pdf v Representative confocal images of control () (top) and experimental () (middle, males; bottom, females) flies stained with RFP and PDF antibodies. Experimental flies show a significant reduction in PDF levels in the s-LNs (white arrows, PDF channel) and misrouting of the s-LNdorsal termini (white arrows, RFP channel).Representative confocal images of the small and large LNs of control () (males, top) and (females, bottom),Representative confocal images of the small and large LNs of experimental () (males, top) and (females, bottom) brains stained with RFP and PDF antibodies.Quantification of PDF levels from s-LNcell bodies in control () and experimental () flies are plotted for males (left) and females (right).> 12 brains for all genotypes.Differences in the PDF intensity values of experimental males and females from their respective parental controls.Number of PDF-positive s-LNs in each brain are plotted for control () and experimental () males and females.> 13 brains for all genotypes. Statistical comparisons were performed between the control and experimental flies of both sexes using the Mann–Whitneytest. The box plots extend from the 25th to 75th percentile, with whiskers extending from the smallest to the largest value, and each point represents data from a single fly. Combined data from at least two replicate experiments are plotted. *< 0.05, **< 0.01, ***< 0.001. Scale bars = 50 μm. (A) (B) (left) (right) (C) (D) (E) Pdf-RFP, Pdf-Gal4; tub-Gal80 Pdf-RFP, Pdf-Gal4; tub-Gal80 > Cas9; Pdf-g Pdf-RFP, Pdf-Gal4; tub-Gal80 Pdf-RFP, Pdf-Gal4; tub-Gal80 > Cas9; Pdf-g Pdf-RFP, Pdf-Gal4; tub-Gal80 Pdf-RFP, Pdf-Gal4; tub-Gal80 > Cas9; Pdf-g n Pdf-RFP, Pdf-Gal4; tub-Gal80 Pdf-RFP, Pdf-Gal4; tub-Gal80 > Cas9; Pdf-g n U p p p ts ts ts ts ts ts ts ts v v v v v v https://doi.org/10.6084/m9.figshare.28598054.v1
CRISPR-Cas9-mediatedmutagenesis has more pronounced effects on male behavior. Pdf Representative actograms of control () andand experimental () males (left) and females (right) are plotted for six days of LD followed by nine days of DD.Percentage of rhythmic flies are plotted for control (and experimental (males (= 85 (),= 89 (),= 106 ()) and females (= 79 (),= 78 (),= 85 (The error bars represent SEM values plotted across three replicate experiments.Free-running periods of control (and experimental (males (left) and females (right) calculated via the chi-squared periodogram are plotted.Rhythmic power of control (and experimental (males (left) and females (right) calculated via the chi-squared periodogram are plotted.The differences in rhythmic power of experimental males and females from their respective controls are plotted. Flies were kept at 28 °C throughout development, and as adults, experiments were conducted at 28 °C. Statistical comparisons were performed between the control and experimental flies of both sexes using the Kruskal–Wallis test followed by Dunn’s multiple comparisons test for panels B–D and Mann–Whitneytest for. Percentage of rhythmic flies were compared using Fisher’s exact test. The box plots extend from the 25th to 75th percentile, with whiskers extending from the smallest to the largest value, and each point represents data from a single fly. Combined data from at least three replicate experiments are plotted. **< 0.01, ***< 0.001. (A) (B) (C) (D) (E) Pdf-RFP, Pdf-Gal4; tub-Gal80 (UAS Cas9; Pdf-g) Pdf-RFP, Pdf-Gal4; tub-Gal80 > Cas9; Pdf-g Pdf-RFP, Pdf-Gal4; tub-Gal80 ) and (UAS Cas9; Pdf-g) Pdf-RFP, Pdf-Gal4; tub-Gal80 > Cas9; Pdf-g) n Pdf-RFP, Pdf-Gal4; tub-Gal80 n UAS Cas9; Pdf-g n Pdf-RFP, Pdf-Gal4; tub-Gal80 > Cas9; Pdf-g n Pdf-RFP, Pdf-Gal4; tub-Gal80 n UAS Cas9; Pdf-g n Pdf-RFP, Pdf-Gal4; tub-Gal80 > Cas9; Pdf-g)). Pdf-RFP, Pdf-Gal4; tub-Gal80 and UAS Cas9; Pdf-g) Pdf-RFP, Pdf-Gal4; tub-Gal80 > Cas9; Pdf-g) Pdf-RFP, Pdf-Gal4; tub-Gal80 and UAS Cas9; Pdf-g) Pdf-RFP, Pdf-Gal4; tub-Gal80 > Cas9; Pdf-g) U p p ts ts ts ts ts ts ts ts ts ts ts ts Fig 3E https://doi.org/10.6084/m9.figshare.28601354.v1
Speeding up the M-cell clock results in a more effective period shortening in males
Next, we sought to determine if the influence of the Pdf-releasing cells themselves was sexually dimorphic. While PDF is released from both large and small LNvs, only s-LNvs (Morning cells) play key roles in regulating free-running rhythm properties [18]. In males, manipulations that change the pace of the clock specifically in the LNvs result in changes in the phase of the morning peak of activity and in FRP [22,56]. We expressed the doubletime ‘short’ (DBTs) allele [57] under the Pdf-Gal4 driver (Fig 4A) and analyzed the effects on behavior in both sexes. We found that Pdf > DBTs males, but not females, have an advanced phase of the morning peak of activity (M-peak, Fig 4B and 4C). This suggests that under LD cycles, the M-oscillator is more effective at setting the phase of male than female behavior.
Under free-running conditions, both Pdf > DBTs male and female flies showed a significantly lower percentage of rhythmic flies than their controls (Fig 4D), but there were fewer rhythmic females (~40%) than males (~65%) (Fig 4D and 4E). The FRP of most Pdf > DBTs males was ~ 18.5 h (Fig 4F), whereas the Pdf > DBTs females showed a large proportion of individuals with a period of ~ 24 h, reflecting the pace of the molecular clock in the rest of the clock network (Figs 4F, 4G, and S4H). Some Pdf > DBTs males (~23%) and a smaller proportion of females (~6%) also presented complex rhythms (Table 4). The average period value of the second-period component (which has a lower power value) was ~ 24.07 ± 0.4 for the experimental males and 19.7 ± 0.4 for the experimental females. Neither male nor female control flies exhibited complex rhythms (Table 4). Among rhythmic flies, both Pdf > DBTs males and females had lower rhythmic power than the controls (Fig 4H), with no difference between the sexes (Fig 4I). As a control, we expressed DBTs via Clk856-Gal4 which is expressed in most clock neurons (S4A Fig). Speeding up the molecular clock in most clock neurons significantly advanced both the morning and evening activity peaks of both males and females (S4B–S4D Fig and S2 Table). There were no significant differences between experimental males and females in the percentage of rhythmic flies, rhythmic power, or shortening of FRPs following DBT expression in all clock neurons (S4E–S4G Fig). We also compared per1 mutant males and females, and our results show that both are nearly completely arrhythmic (S4I Fig).
These results support the notion that M-cells are more dominant in the male than in the female circadian network. Previous studies have shown that blocking synaptic neurotransmission by expressing the tetanus toxin light chain (TeTxLC) in small and large LNvs affects male activity rhythms, likely in a Pdf-independent manner [58–60]. We analyzed male and female Pdf > TeTxLC flies and found that neither sex significantly changed the ability to maintain rhythmicity under free-running conditions (S5A and S5B Fig, and S3 Table). Both sexes significantly lengthened the FRP, but the effect was more pronounced in males (S5C and S5D Fig). Rhythmic power was not affected in experimental flies of either sex (S5E Fig). These results are consistent with previous studies which show that TeTxLC expression in Pdf+ neurons lead to behavioral phenotypes that are different from those of loss of Pdf, and suggest that manipulations of neuronal activity of the Pdf-expressing neurons have a more pronounced effect in male FRP.
We next asked whether changing the pace of the clock in LNds (E-cells) via DBTS expression also had sexually dimorphic effects on behavior. These cells can be subdivided into at least three different clusters on the basis of their anatomy [8], physiology [56], transcriptomic profiles [15], and connectivity patterns [17]. The PDFR-expressing E1 and E2 clusters have been shown to regulate evening activity under LD [61,62] and to be able to maintain free-running activity rhythms in the absence of a functional clock in M cells [63,64], whereas the behavioral role of the E3 cluster remains unknown. We used the MB122-B split-gal4 driver to target the E1 and E2 subsets (Fig 5A) and found that while expressing DBTS in this group of evening cells significantly advanced the phase of the E-peak in experimental flies of both sexes (Fig 5B and 5C), the effect was more pronounced in females (Fig 5D). Speeding up of clocks in the E1 + E2 LNds did not significantly alter the FRP or rhythmic power of experimental flies of either sex (Fig 5E and 5F). M-cells have been shown to be the dominant oscillators in DD and to regulate rhythm properties such as persistence and the FRP of endogenous locomotor rhythms to a large extent [18,22,23,56], although manipulations of other clock cells can affect rhythm properties to some extent [56,65]. Thus, speeding up the clock in the PDFR+ E1 and E2 clusters leads to similar behavioral phenotypes in males and females under free-running conditions. Taken together, these results indicate that the relative influence of the M and E subsets of clock neurons are sexually dimorphic.
Speeding up the clock in M-cells leads to sexually dimorphic phenotypes. Depiction of the adultbrain hemisphere indicating the clock cell subsets (colored) having a faster running molecular clock.Average activity plots of control () and (UAS-) and experimental () flies are plotted for males (left) and females (right). The plots are averaged over flies and days for a period of three days under LD 12:12.The phase of the morning peak of activity under LD 12:12 is plotted for control (), (UAS-) and experimental () males (left,= 116 (),= 120 (UAS-),= 88 ()) and females (right,= 85 (),= 109 (UAS-),= 73 ()).Percentages of rhythmic flies are plotted for controls (and experimental () males (= 90 (),= 108 (),= 93 () and females (= 66 (),= 83 (),= 70 (The error bars represent the SEM values plotted across three replicate experiments.Representative actograms of controls (and UAS-) and experimental () females (top) and males (bottom) plotted for five days of LD followed by 10 days of DD.Free-running periods of control () and experimental () males (left) and females (right) calculated via the chi-squared periodogram are plotted.Percentage of flies having a short (18–21.5 h) and normal (23–25 h) free-running periods are plotted for experimental () males and femalesRhythmic power of control () and experimental () males (left) and females (right) calculated using the Chi-squared periodogram are plotted.The differences in rhythmic power between experimental males and females and their respective controls are plotted. Statistical comparisons were performed between the control and experimental flies of both sexes using the Kruskal–Wallis test followed by Dunn’s multiple comparisons test for panels C, F, and H and Mann–Whitneytest for. Percentage of rhythmic flies were compared using Fisher’s exact test. The box plots extend from the 25th to 75th percentile, with whiskers extending from the smallest to the largest value, and each point represents data from a single fly. Combined data from at least three replicate experiments are plotted. *< 0.05, **< 0.01, ***< 0.001. (A) (B) (C) (D) (E) (F) (G) (H) (I) Drosophila Pdf-Gal4 DBT Pdf > DBT Pdf-Gal4 DBT Pdf > DBT n Pdf-Gal4 n DBT n Pdf > DBT n Pdf-Gal4 n DBT n Pdf > DBT Pdf-Gal4), (UAS-DBT ), Pdf > DBT n Pdf-Gal4 n UAS-DBT n Pdf > DBT n Pdf-Gal4 n UAS-DBT n Pdf > DBT ). Pdf-Gal4 DBT Pdf > DBT Pdf-Gal4), (UAS-DBT Pdf > DBT Pdf > dBTs Pdf-Gal4), (UAS-DBT Pdf > DBT U p p p s s s s s s s s s s s s s s s s s s s s Fig 4I https://doi.org/10.6084/m9.figshare.28601357.v1
Speeding up the clock in an E-cell subset leads to a more advanced evening peak in females. Depiction of the adultbrain hemisphere indicating the clock cell subsets (colored) having a faster running molecular clock.Average activity plots of control () and () and experimental () flies are plotted for males (left) and females (right). The plots are averaged over flies and days for a period of three days under LD 12:12.Phase of the evening peak of activity under LD 12:12 for controls () and () and experimental () males (left,= 92 (),= 93 (),= 93 ()) and females (right,= 85 (),= 87 (),= 89 ()).The differences in the phase of the E peak between experimental males and females and their respective controls are plotted.Free-running period of control (and) and experimental () males (left) and females (right) calculated via the chi-squared periodogram are plotted.Rhythmic power of control () and experimental () males (left) and females (right) calculated via the chi-squared periodogram are plotted. Statistical comparisons were performed between the control and experimental flies of both sexes using the Kruskal–Wallis test followed by Dunn’s multiple comparisons test for panels 5C, 5E, and 5F and Mann–Whitneytest for. The box plots extend from the 25th to 75th percentile, with whiskers extending from the smallest to the largest value, and each point represents data from a single fly. Combined data from at least three replicate experiments are plotted. **< 0.01, ***< 0.001. (A) (B) (C) (D) (E) (F) Drosophila MB122B-Gal4 UAS-DBT MB122B > DBT MB122B-Gal4 UAS-DBT MB122B > DBT n MB122B-Gal4 n UAS-DBT n MB122B > DBT n MB122B-Gal4 n UAS-DBT n MB122B > DBT MB122B-Gal4 UAS-DBT MB122B > DBT MB122B-Gal4 and UAS-DBT MB122B > DBT U p p s s s s s s s s s s s s Fig 5D https://doi.org/10.6084/m9.figshare.28601363.v1
| Pdf > DBTs | |||||
|---|---|---|---|---|---|
| Genotype | n | % Rhythmicity ± SEM | % Complex rhythms ± SEM | Free-running period ± SEM | Rhythmic power ± SEM |
| (male)Pdf-Gal4 | 90 | 95.86 ± 2.73 | 0 | 24.3 ± 0.03 | 111.8 ± 5.58 |
| (male)UAS DBTs | 108 | 96.79 ± 2.21 | 0 | 23.9 ± 0.01 | 134.8 ± 4.44 |
| (male)Pdf > DBTs | 93 | 69.87 ± 6.89***# | 23.73 ± 5.51*** | 18.9 ± 0.14*** | 49.48 ± 3.54*** |
| (female)Pdf-Gal4 | 66 | 93.56 ± 4.02 | 0 | 24.4 ± 0.03 | 122 ± 6.76 |
| (female)UAS DBTs | 83 | 89.45 ± 2.74 | 0 | 24.2 ± 0.03 | 89.7 ± 5.19 |
| (female)Pdf > DBTs | 70 | 38.41 ± 2.74***# | 5.54 ± 3.45 | 22.5 ± 0.58 | 24.8 ± 3.55*** |
Discussion
The critical importance of considering sex as a biological variable has gained increasing recognition in biomedical research [66,67]. Bias toward male subjects is particularly prevalent in neuroscience, with single-sex studies using male animals outnumbering those using female animals at a ratio of 5.5:1 [68]. This disparity extends to chronobiology, resulting in a limited understanding of how sex affects circadian organization in the nervous system. However, work from several laboratories has revealed sexual dimorphism within the SCN and in its input and output pathways [1,69]. Sex differences in SCN morphology have been described in both animal models and humans, and sex differences in SCN electrical activity and steroid hormone receptors have also been reported (reviewed in [2]). Notably, sex differences in the number of SCN neurons that express the neuropeptide vasoactive intestinal polypeptide (VIP) and in Vip mRNA expression have been reported (reviewed in [2]). The roles of the mammalian VIP and the Drosophila PDF in circadian physiology are highly similar, although neither these peptides nor their receptors are sequence orthologs [70].
Several studies have shown the importance of PDF in generating coherent rhythms of ~ 24-h periodicity. Here, we report that females lacking Pdf or its receptor PdfR are more likely to maintain consolidated activity–rest behavior than males. This could be because of sex differences in PDF signaling mechanisms, PDFR expression, or the influence of other clock neurons within the network. In males, other neuropeptides are known to act in concert with PDF to maintain consolidated rhythms in the network, although none of them have as profound an effect as PDF in regulating activity-rest rhythms in DD [49,71]. Single mutants of DH31 and CCHamide1 do not affect activity rhythms by themselves, but the double mutants of these neuropeptides along with Pdf1 (DH311/Pdf1 and Pdf1/CCHaSK8) are almost completely arrhythmic, suggesting that these neuropeptides act hierarchically in the network, with PDF being at the top of that hierarchy [49,71]. The importance of PDF relative to other peptides released by clock neurons may also be sexually dimorphic.
Although CRISPR manipulation was only partially effective at eliminating PDF expression, it nevertheless produced phenotypes reminiscent of those produced by the Pdf1 mutation. We observed faint staining in the dorsal projections of at least one s-LNv in at least one hemisphere in most brains, and PDF staining in a single s-LNv projection reaching the dorsal brain has been shown to be sufficient for behavioral rhythms [72]. Experimental flies in which Pdf was knocked out starting at the onset of promoter expression early in development showed extensive misrouting of their dorsal termini, similar to what has been reported for Pdf1 males [54]. Instances of s-LNv misrouting have also been observed in other core clock mutants, such as per1 and tim1 [73] and cyc1 [74]. No correlation between misrouting and behavioral phenotypes was found by others for Pdf1 males [54]. Importantly, Pdf > Pdfg;Cas9 manipulation recapitulates the sexually dimorphic circadian phenotypes of Pdf1 mutants: a larger fraction of females are rhythmic, and females exhibit greater rhythm power.
In males, changing the speed of the M cell clock leads to phase changes in the morning peak under LD [61]. To determine whether M-cell manipulations also have sexually dimorphic effects on behavior, we sped up the molecular clock by expressing the doubletime short allele (DBTs). Surprisingly, our results revealed that speeding up the clock of M-cells advances the phase of the morning peak in males, whereas the female morning peak phase is not affected. These results support previous studies conducted in males on the role of M-cells in regulating the morning peak of activity [18,23,61] and suggest that M-cells are unable to regulate the phase of morning activity in the same way in females. In DD, males had largely coherent short-term rhythms, and the majority (65%) were rhythmic. In contrast, only 40% of the females were rhythmic, and their period showed a bimodal distribution. These findings further support the notion that M-cells are less dominant than the circadian clock network in females. A possible explanation for this is that other clock neurons are able to “resist” their influence, and the conflict between the fast-paced M cell clock and the ~ 24 clock other clock neurons is what leads to greater arrhythmicity in females.
The expression of a TeTxLC construct in flies blocks neurotransmission by binding and cleaving the synaptic protein Synaptobrevin [75]. Expressing tetanus toxin in LNvs did not result in a reduction in rhythmic power, but it lengthened the male FRP, as reported in previous studies [59,60,76]. The behavioral phenotypes resulting from the blockade of synaptic transmission differ from those resulting from the loss of PDF [20] or the ablation of Pdf-expressing neurons [23], possibly because tetanus toxin affects classical transmission and not the dense core vesicle-mediated release of neuropeptides such as PDF [77]. Abrogating the dorsal termini of the small LNvs, where most of the output synapses are found [17,77], does not result in behavioral phenotypes similar to those of Pdf1 under either LD or DD [24]. Our results show that blocking synaptobrevin-dependent synaptic transmission in M-cells does not affect rhythmicity but rather lengthens the FRP. The period lengthening is more pronounced in males, supporting the notion that M-cells have a greater influence on the circadian network in males.
Cryptochrome and PdfR-expressing clusters of evening cells—the sNPF-expressing E1 cluster and the ITP-expressing E2 cluster [78]—have roles in setting the phase of the E-peak under LD and sustaining behavioral rhythms in the absence of a functional molecular clock in M-cells [63,64]. To test whether these cells have a differential influence on the network in males and females, we expressed DBTS under a driver that is expressed specifically in the E1 and E2 subsets of LNds. Our results showed that speeding up the clocks in the E1 + E2 clusters resulted in a phase advance in the evening peak of activity in both sexes, but the effect was more pronounced in females. A possible reason for the behavioral differences observed between the sexes could be redundancy in females, such that the network is not as dependent on PDF or M-cells for timekeeping. This finding suggests that the female network could have a more distributed mode of timekeeping throughout the circadian clock network.
Across species, sex differences in the circadian timing system are largely related to the regulation of reproduction-related behaviors. In mammals, the SCN determines the timing of the release of reproductive hormones and influences the timing of mating (reviewed in [69]) and aggression [79]. In Drosophila, the circadian clock controls the timing of sex-specific and sexually dimorphic behaviors, such as male courtship [80] and female sexual receptivity [81] and egg laying [82]. This regulation of rhythmic behaviors requires connectivity between clock neurons and downstream sex-specific circuits. For example, the DN1p cluster, which has been shown to be more active in males [38], is functionally connected to the male-specific fru-expressing P1 neurons that regulate male courtship [39]. In females, Allatostatin C-producing DN1ps have been shown to connect to downstream targets to control rhythms in oogenesis [83], and the Janelia female hemibrain connectome revealed that the LNds form connections with the doublesex-expressing PC1 cluster [17]. Our data suggest that the relative hierarchy of circadian oscillators is sexually dimorphic, with a less dominant M oscillator in females. Interestingly, when males are exposed to constant light (LL), the prevalence of the morning oscillator decreases, while the evening oscillator becomes more dominant [84–86]. One proposed mechanism for this shift in the hierarchy of circadian oscillators involves GW182, which influences the circadian neural network’s response to light and modulates the level of PDFR signaling [87]. If the evening oscillator is more dominant in the female circadian network, it is possible that males and females respond differently to constant light, at least under specific light intensities. The existence of sex differences in the hierarchy of circadian oscillators may serve an adaptive purpose, ensuring the precise timing of essential female-specific behaviors crucial for reproductive fitness, such as sexual receptivity and egg laying.
Materials and methods
Fly lines and rearing
All the genotypes were reared on corn syrup soy media (Archon Scientific; Durham, NC) under LD 12:12 cycles at 25 °C unless specified otherwise (see figure legends for details). The fly lines used in this study were Canton-S, w1118, Pdf1, PdfR1, Pdf-RFP, Pdf-Gal4; tub-Gal80ts, UAS-Cas9; UAS-Pdfg, Pdf-Gal4, UAS-DBTs, UAS TeTxLC, Clk856Gal4, s-LNvGal4, and MB122B-Gal4. See the fly lines and reagents table below for more details. All experiments were conducted with virgin females, as mating affects female rhythm strength [48]. We employed a Pdf1 mutant line outcrossed in the w1118 background. See Table 5 for details about fly lines.
| REAGENT or RESOURCE | SOURCE | IDENTIFIER |
|---|---|---|
| Experimental Models: Organisms/Strains | ||
| Canton-S | BloomingtonStock CenterDrosophila | BDSC 64349 |
| w;+;+1118 | BloomingtonStock CenterDrosophila | BDSC 3605 |
| yw;Pdf-RFP,Pdf-Gal4;Tub-gal80ts | Justin. Blau, NYU | |
| w;;Pdf1 | Paul Taghert, Wash U Med. School | |
| w PdfR;;1 | Paul Taghert, Wash U Med. School | |
| w; Pdf-gal4; + | Paul Taghert, Wash U Med. School | |
| ;UAS Cas9; UAS pdfg | Michael Rosbash, Brandeis University | BDSC 99650 (pdfg) and BDSC 58985 (Cas9) |
| w; + ;DBTs | Jeffrey Price, University of Missouri | |
| w; MB122B-Gal4; + | Gerry Rubin, HHMI Janelia Research Campus | |
| w; Clk856-Gal4; + | Orie Shafer, CUNY ASRC | |
| s-LNv Gal4 | Gerry Rubin, Janelia Research Campus | SS00681-Gal4 |
| w; UAS TeTxLC; + | Cahir O’Kane, University of Cambridge | |
| Antibodies | ||
| Rabbi anti-RFP (1:2000) | Rockland | #600-401-379-RTU |
| Mouse anti-PDF (1:3000) | Developmental Studies Hybridoma Bank | |
| Anti-rabbit Alexa-568 (1:3000) | Thermo Fisher | A11036 |
| Anti-mouse Alexa-488 (1:3000) | Thermo Fisher | A11029 |
| Software | ||
| Fiji | http://fiji.sc | RRID: SCR_002285 |
| MATLAB R2022b | MathWorks, Natick | RRID: SCR_001622 |
| GraphPad Prism 9.0 | GraphPad Software | RRID: SCR_002798 |
| DAM FileScan | Trikinetics | |
| ClockLab | Actimetrics | RRID:SCR_014309 |
| Chemicals, Peptides, and Recombinant Proteins | ||
| Vectashield Mounting Medium | Vector Laboratories | #H-1000-10 |
| Premix PBS Buffer (10x) | Sigma–Aldrich | Cat# 11666789001 |
| 2% Paraformaldehyde (PFA) | Sigma–Aldrich | 47608-250ML-F |
| Triton-X-100 | Bio Basic | CAS#9002-93-1 |
| Schneider’sMedium () Drosophila S2 | Thermo Fisher | 21720024 |
| Other | ||
| DAM2Activity MonitorsDrosophila | Trikinetics | |
| DAM Drosophila Environmental Monitors | Trikinetics | |
Activity recording and analysis
Individual male and virgin female flies (3–5 days old) were housed in glass locomotor tubes containing 2% agar–4% sucrose food on one end and yarn on the other end. Locomotor activity was recorded using Drosophila activity monitors (DAM, Trikinetics, Waltham, United States of America). The experiments were conducted in Tritech or Percival incubators under controlled light and temperature conditions. Flies were entrained to 12:12 LD cycles for at least 5 days and then transferred to constant darkness (DD) for at least 7 days at a constant temperature of 25 °C unless otherwise specified (see figure legends for details). The raw data obtained from the DAM system were scanned and binned into activity counts of 15-min intervals via the DAM File scan. The data were analyzed via the CLOCKLAB software (Actimetrics, Wilmette, IL).
The values of period and rhythmic power were calculated for a period of 7 days via a chi-squared periodogram with a cutoff of p = 0.01. The rhythmic power for each designated rhythmic fly was determined by subtracting the chi-squared significance value from the power of the periodogram. Flies that did not exhibit a periodicity peak above the significance threshold were categorized as “arrhythmic,” and their period and rhythmic power were not included in the analysis. The values of the morning and evening peaks were calculated via PHASE software [88]. The total LD sleep values for all the genotypes were calculated for a period of 3 days (LD days 2–4) via the PHASE software. Representative actograms were generated via ClockLab, and activity plots were generated via PHASE. The period, rhythmic power, total sleep, and phase values of all the flies for a particular experimental genotype were compared against the background or parental controls via either the Mann–Whitney test or the Kruskal–Wallis ANOVA followed by the Dunn’s multiple comparisons test. The details of the statistical comparisons and the number of flies used in a given experiment are indicated in their respective figure legends. The number of rhythmic flies of the experimental genotype was compared against their respective background or parental controls via Fisher’s exact test. All the statistical analyses were performed via GraphPad Prism 9.0.
If both the experimental males and females were significantly different from their respective control flies of the same sex, the extent of sex differences were calculated by subtracting the average values of the control from each individual experimental value. These differences were then directly compared between males and females using the appropriate statistical analysis (the test used in each case is mentioned in the figure legends for the respective figures). In case of experiments with two parental controls, the average value to calculate the difference would be the average of the Gal4 and UAS control genotypes.
Immunohistochemistry
The brains of adult male or female flies were dissected in ice-cold Schneider’s insect media (S2) and fixed immediately after dissection in 2% paraformaldehyde (PFA) in S2 media for 30 min at room temperature. The fixed brains were washed (three washes of 10 min each) with 0.3% phosphate-buffered saline-Triton X 100 (PBS-TX) and then treated with blocking solution (5% normal goat serum made in 0.3% PBS-TX) for 1 h at room temperature. The brains were then incubated with primary antibodies at 4 °C for 24 h. The primary antibodies used were anti-PDF (mouse, 1:3000, C7, DSHB) and anti-RFP (rabbit, 1:2,000, Rockland Immunochemicals). After incubation, the brains were subjected to six washes with 0.3% PBS-TX and incubated with Alexa Fluor-conjugated secondary antibodies overnight at 4 °C. The following secondary antibodies were used: goat anti-mouse 488 (1:3,000, Invitrogen) and goat anti-rabbit 568 (1:3,000, Invitrogen). After incubation, the brain samples were washed six times with 0.3% PBS-TX, cleaned and mounted on a clean glass slide using Vectashield mounting media.
Image acquisition and analysis
The brains were imaged via a confocal microscope (Olympus FV3000) with an Olympus UPLanXApo 20× or 40× objective. Image analysis was performed via Fiji software [89]. In the samples, small and large LNvs were classified on the basis of their anatomical locations and expression of the PDF. PDF intensities in these cells were measured by selecting the slice of the Z-stack that showed the maximum intensity, drawing a region of interest (ROI) around the cells, and measuring their intensities. Three to four separate background values were also measured around each cell, and the final intensity was taken as the difference between the cell intensity and the average background.
For quantification of the PDF in the dorsal projections, a rectangular box was drawn as the ROI starting from the point where the PDF projection turns into the dorsal brain, and the intensity is measured. Three to four background values were also measured around the projection. The intensity values obtained from both hemispheres for each cell type for each brain were averaged and used for statistical analysis. PDF intensity from the s-LNv was compared between the experimental and control genotypes via a Mann–Whitney test. To estimate different aspects of rhythmicity in PDF oscillations in the dorsal termini of s-LNv in males and females, we used a COSINOR-based curve-fitting method [Cornelissen, 2014]. COSINOR analysis was implemented via the CAT Cosinor function from the CATkit package written for R [90].
Supporting information
Acknowledgments
We are grateful to Charlotte Helfrich-Forster, Abhilash Lakshman, Cahir O’Kane, Orie Shafer, Jeff Price, and Paul Taghert for helpful discussions and Orie Shafer and members of the Fernandez Lab for helpful comments on the manuscript. We also thank Justin Blau, Aljoscha Nern, Gerry Rubin, Michael Rosbash, and Paul Taghert for sharing fly lines. The mouse anti-PDF antibody was obtained from the Developmental Studies Hybridoma Bank, created by the NICHD of the NIH and maintained at The University of Iowa, Department of Biology, Iowa City, IA 52242. Stocks obtained from the Bloomington Drosophila Stock Center (NIH P40OD018537) were used in this study. We used FlyBase (release FB2025_01) to find information on phenotypes, stocks, and gene expression.
Abbreviations
Data Availability
Data are freely available without restrictions in a public repository. The relevant doi information is included in the corresponding figure legends. The relevant URLs can be also found below. Fig 1: https://doi.org/10.6084/m9.figshare.28601348.v1↗ Fig 2: https://doi.org/10.6084/m9.figshare.28598054.v1↗ Fig 3: https://doi.org/10.6084/m9.figshare.28601354.v1↗ Fig 4: https://doi.org/10.6084/m9.figshare.28601357.v1↗ Fig 5: https://doi.org/10.6084/m9.figshare.28601363.v1↗ S1 Fig: https://doi.org/10.6084/m9.figshare.28601366.v1↗ S2 Fig: https://doi.org/10.6084/m9.figshare.28601372.v1↗ S3 Fig: https://doi.org/10.6084/m9.figshare.28601375.v1↗ S4 Fig: https://doi.org/10.6084/m9.figshare.28601378.v1↗ S5 Fig: https://doi.org/10.6084/m9.figshare.28601381.v1↗.
Funding Statement
This work was supported by an Arnold and Mabel Beckman Foundation award to E.S-C (Scholars Program 2023) and a National Science Foundation grant (NSF CAREER Award IOS-2239994) to M.P.F. The funders had no role in the study design, data collection, analysis, decision to publish, or preparation of the manuscript. https://www.beckman-foundation.org/↗, https://www.nsf.gov/↗.
References
Associated Data
Supplementary Materials
Data Availability Statement
Data are freely available without restrictions in a public repository. The relevant doi information is included in the corresponding figure legends. The relevant URLs can be also found below. Fig 1: https://doi.org/10.6084/m9.figshare.28601348.v1↗ Fig 2: https://doi.org/10.6084/m9.figshare.28598054.v1↗ Fig 3: https://doi.org/10.6084/m9.figshare.28601354.v1↗ Fig 4: https://doi.org/10.6084/m9.figshare.28601357.v1↗ Fig 5: https://doi.org/10.6084/m9.figshare.28601363.v1↗ S1 Fig: https://doi.org/10.6084/m9.figshare.28601366.v1↗ S2 Fig: https://doi.org/10.6084/m9.figshare.28601372.v1↗ S3 Fig: https://doi.org/10.6084/m9.figshare.28601375.v1↗ S4 Fig: https://doi.org/10.6084/m9.figshare.28601378.v1↗ S5 Fig: https://doi.org/10.6084/m9.figshare.28601381.v1↗.