Daywake, an Anti-siesta Gene Linked to a Splicing-Based Thermostat from an Adjoining Clock Gene

Abstract

Sleep is fundamental to animal survival but is a vulnerable state that also limits how much time can be devoted to critical wake-dependent activities [1]. Although many animals are day-active and sleep at night, they exhibit a midday nap, or "siesta," that can vary in intensity and is usually more prominent on warm days. In humans, the balance between maintaining the wake state or sleeping during the day has important health implications [2], but the mechanisms underlying this dynamic regulation are poorly understood. Using the well-established Drosophila melanogaster animal model to study sleep [3], we identify a new wake-sleep regulator that we term daywake (dyw). dyw encodes a juvenile hormone-binding protein [4] that functions in neurons as a day-specific anti-siesta gene, with little effect on sleep levels during the nighttime or in the absence of light. Remarkably, dyw expression is stimulated in trans via cold-enhanced splicing of the dmpi8 intron [5] from the reverse-oriented but slightly overlapping period (per) clock gene [6]. The functionally integrated dmpi8-dyw genetic unit operates as a "behavioral temperate acclimator" by increasingly counterbalancing siesta-promoting pathways as daily temperatures become cooler and carry reduced risks from daytime heat exposure. While daily patterns of when animals are awake and when they sleep are largely scheduled by the circadian timing system, dyw implicates a less recognized class of modulatory wake-sleep regulators that primarily function to enhance flexibility in wake-sleep preference, a behavioral plasticity that is commonly observed in animals during the midday, raising the possibility of shared mechanisms.

Keywords: 0.9 gene; Daywake; Drosophila; dmpi8 intron; juvenile hormone-binding protein; midday siesta; period clock gene; pre-mRNA splicing; sleep-wake behavior.

Figure 1.. Effect of dmpi8 splicing efficiency on daytime sleep requires 0.9 protein function but not dPER.

(A) Schematic of the dmpi8WT and dmpi8UP transgenes [10] modified to include a premature stop codon in the open reading frame of dper, 0.9, or both. The transgenes are based on a dper cDNA/genomic hybrid and include approximately 825bp 5’ upstream of the 0.9 transcription unit, which is transcribed in the opposite direction to dper. The 3’ untranslated regions of both dper and 0.9 overlap by 49bp. Rectangles; black (dper circadian regulatory sequence), gray (untranslated regions), green (coding region); black lines (introns). (B-G) Young adult male (B-E) or female (F-I) flies were exposed to 4 days of 12hr light/12 hr dark cycles (LD) followed by three days of continuous light (LL) at 25°C. For each genotype, graphs represent data from 96 individual flies (three independent transgenic lines × 32 individual flies/line) and data pooled to get the population average. Shown are the relative sleep levels during the last 2 days of LD followed by 2 days of LL for transgenic flies carrying the dmpi8UP (red) or dmpi8WT (blue) versions of the wildtype per-0.9 (B, F) per-STOP (C, G), 0.9-STOP (D, H) or double stop (E, I). Grey rectangle, 12 hr dark period. The following lines were used to obtain group averages; dmpi8WT, kx-f4c, m38-k, f9; dmpi8UP, m17, m32, f13; per-STOP[dmpi8WT], m53, f18, f46; per-STOP[dmpi8UP], m55, m131, m138; 0.9-STOP[dmpi8WT], m50, m62, m73; 0.9-STOP[dmpi8UP], m55, f50, f59; per/0.9-STOP[dmpi8WT], m94, m100, f38; per/0.9-STOP[dmpi8UP], m42, m81, f5. Representative examples are shown and similar results were obtained with other independent lines (see STAR Methods and KEY RESORCE TABLE for identity of lines analyzed). See also Figures S1 and S2 for additional results.

Figure 2.. Increased dmpi8 splicing efficiency is associated with higher 0.9 transcript levels.

(A-F) Flies were kept at the indicated temperature and entrained by 4 days of LD, followed by collection on the fifth day at the indicated times [where Zeitgeber 0 (ZT0) is defined as lights-on]. RNA was extracted from head extracts and used to measure the relative levels of 0.9 transcripts and dmpi8 splicing efficiency, as indicated. Significant differences were observed for daily 0.9 levels and dmpi8 splicing efficiency, as follows: (A, B) Comparing dmpi8UP and dmpi8WT (two-sided t-test); 0.9 levels; ZT2, 4.3 × 10⁻³; ZT8, 8.2 × 10⁻⁵; ZT14, 2.7 × 10⁻⁴; ZT20, 9.9 × 10⁻⁶; dmpi8 splicing; ZT2, 2.8 × 10⁻⁸; ZT8, 7.1 × 10⁻⁹; ZT14, 1.8 × 10⁻⁶; ZT20, 5.2 × 10⁻¹¹. The results are based on pooling data from at least three independent transgenic lines for each genotype; dmpi8WT (f9, m38-k, kx-f4-c), dmpi8UP (m17, f13, m32). (C, D) Comparing w,norpA^[36] mutant to its genetic background control (two-sided Student’s t-test); 0.9 levels; ZT2, 7.4 × 10⁻⁶; ZT8, 8.0 × 10⁻³; ZT14, 1.6 × 10⁻³; ZT20, 3.9 × 10⁻³; dmpi8 splicing; ZT2, 7.9 × 10⁻³; ZT8, 8.7 × 10⁻⁵; ZT14, 3.4 × 10⁻⁵; ZT20, 5.4 × 10⁻³. (E, F) Comparing three temperatures (ANOVA); 0.9 levels; ZT2, 1.7 × 10⁻⁷; ZT8, 5.4 × 10⁻⁴; ZT14, 3.6 × 10⁴; ZT20, 1.7 × 10⁻²; dmpi8 splicing; ZT2, 4.8 × 10⁻²; ZT8, 8.4 × 10⁻⁴; ZT14, 2.4 × 10⁻²; ZT20, 6.0 × 10⁻³. For each experiment, approximately 50-100 flies were used for each timepoint. Graphs shown are the average of three independent experiments. See also Figure S1D, E for further results.

Figure 3.. The 0.9 gene suppresses daytime sleep with little to no effect on nighttime sleep levels.

(A, B) Flies were kept for 5 days in LD at 25°C, and shown are total day and night sleep levels (min) for male adult progeny from crosses between per-Gal4 and the indicated UAS-RNAi lines (blue bars); the gene targeted by RNAi is shown at bottom of panels. Contemporaneous control crosses between w¹¹¹⁸ (w; the background for both per-Gal4 and the UAS-RNAi lines) and per-Gal4 (black) or the UAS-RNAi lines (gray). (C) Flies were kept for 5 days in LD at either 18°C or 25°C, and shown are total daytime sleep (min) for adult male or female from per-Gal4>RNAi-0.9 (blue); and the parental control crosses, w¹¹¹⁸ with per-Gal4 (black) or UAS-RNAi-0.9 (gray). Sleep values are an average from the last 3 days of LD based on 32 individual flies for each cross. **p < 0.001 for experimental group compared to both control crosses. (D-F) Flies were kept for 5 days in LD at 18°C, and shown are the daily sleep levels for female adult progeny for the indicated driver and UAS-RNAi-0.9 (red), and the two parental control crosses between w¹¹¹⁸ and the driver (green) or UAS-RNAi-0.9 (blue). RNAi-0.9(a) and RNAi-0.9(b) refers to stock no. 105930 (VDRC) and 56988 (BDSC), respectively. The sleep profiles are an average of the last days of LD based on 32 individual flies for each cross. (G-J) Flies were kept for 5 days in LD at 25°C, and shown are the daily sleep levels for male (except panel J) adult progeny for the indicated driver and UAS-0.9 (red), and the two parental control crosses between w¹¹¹⁸ and the driver (green) or UAS-0.9 (blue). For each cross, activity data from 32 individual flies was used, and the sleep profiles shown are an average of the last three days of LD based on pooling results from two independent crossing experiments using a different UAS-0.9 line (f57, f79). Similar results were obtained in other experiments (n = 3). See also Figure S3 and Table S1 for additional results using RNAi-0.9, and see Figure S4 for additional results using UAS-0.9.

Figure 4.. Model for how the dmpi8/daywake genetic unit functions in the thermal adaptation of midday siesta in D. melanogaster.

As daily temperatures decline, the splicing efficiency of dmpi8 progressively increases (top), leading to an increase in dper mRNA levels (left). Although it is not clear how dmpi8 splicing regulates dper mRNA levels, active splicing of dmpi8 is required [5]. This suggests that the seeding of splicing factors at the dmpi8 intron stimulates (directly, or indirectly via other interacting factors) the production and/or stability of dper transcripts [5]. The 3’ UTRs of dper and daywake overlap by 49 bp, raising the possibility that this proximity allows the stimulatory mechanism initiated during splicing of the dmpi8 intron to also function over short distances and regulate daywake transcript levels in-trans. Increases in dyw expression promote daytime wakefulness during the midday, leading to a reduction in midday siesta. On warm days, other systems (e.g., heat-activated sleep-promoting DN1s? [33]) become increasingly dominant and evoke a strong midday siesta in a manner that is little influenced by dmpi8 splicing efficiency [7] (bottom, right). This thermal adaptation system provides D. melanogaster the key survival response of mounting a strong siesta on warm days (default state for a tropically originating species), yet the flexibility to increase activity on days when the risks posed by exposure to heat are reduced. For dper and dyw transcripts, blue rectangle (coding region) and green rectangle (3’ UTR); gray double helix (genomic DNA); spliceosome binding is at the dmpi8 intron.