Sleep prevents brain phosphoproteome disruption to safeguard survival

Abstract

Prolonged sleep deprivation (Pr-SD) causes death in many species. While various mechanisms related to sleep regulation or this fatal consequence of sleep loss have been identified, the core molecular basis linking Pr-SD-induced lethality and sleep homeostasis remains unknown in mammals. A critical "point of no return (PONE)" status in Pr-SD subjects is highlighted in classic research, and characterizing PONE status could help uncover this mystery. Using a Pr-SD model and a reliable PONE status prediction method, we show that mice in PONE exhibit an inability to enter natural sleep, and significant disruptions in brain phosphoproteome, independent of deprivation time but closely linked to PONE status. Brain kinase or phosphatase dysfunction influences PONE status development and leads to corresponding sleep aberration concurrently. Daily 80-min recovery sleep significantly delays PONE onset and restores brain phosphoproteome. The harmful effects of excessive kinase activity on PONE development can be eliminated by combining recovery sleep and compensatory phosphatase expression. We conclude that sleep is crucial for maintaining brain phosphoproteome homeostasis, whose disruption may impact both Pr-SD-induced lethality and sleep regulation.

Fig. 1. Evaluation of the PONE status during Pr-SD.

a Schematic diagram of SWAM. b, c Survival analysis of SWAM-treated mice: C57BL/6 J (B6J, n = 15) vs C57BL/6 N (B6N, n = 16) mice (b), B6J cohorts aged 8 weeks (8 W, n = 16) vs 24 weeks (24 W, n = 14) (c). d Schematic of SWAM-treated mice undergoing continuous EEG/EMG recording. Baseline day (BLD), SWAM onset day (SOD), last full day (LFD). e, f Analysis of duration (e) and absolute EEG power spectra (f) of NREMS, REMS and wake states of B6J mice (n = 18 per group) on BLD, SOD and LFD. g Schematic of PONE status evaluation and PONE index calculation. h Representative PONE index development for the mice died on the 6th SWAM day during ZT0–5 (yellow, n = 7), ZT6-11 (red, n = 2) or ZT12-23 (blue, n = 5). i Cumulative mean PONE index value of B6J (n = 15) and B6N (n = 15) mice (top), and B6J mice aged 8 W (n = 14) and 24 W (n = 17) (bottom). j, k Relative timeline (j) and simulating analysis by segmental linear regression (k) for the last four PONE index rating in three scenarios. l A plotting of the PONE index values against the remaining survival duration from each rating time to the actual death point (n = 60). Red line represents mean survival. m Survival analysis of SWAM interruption test for mice with PONE index ≤ 2 (n = 26) and PONE index ≥ 6 (n = 17). Data are mean ± s.e.m. Log-rank (Mantel-Cox) test (b, c, m); two-way ANOVA with Dunnett’s test (e, f right); two-way ANOVA with Sidak’s test (f); Exponential growth equation Least squares fit (i). *P < 0.05; †P < 0.01; ‡P < 0.001; NS, not significant, P > 0.05. n refers to the number of biological replicates.

Fig. 2. Functional changes in the brains of mice with PONE status.

a, b Heat map plot of absolute EEG power (a), mean absolute EEG power (b, top), and theta/delta ratio (b bottom) of BLD, SOD and the last 24 h (L-24h) before the actual death time point (n = 15 per group). c Representative trace of action potentials of motor cortex layer 5 pyramidal neurons from AD mice (top, n = 14), PD mice with normal resting membrane potential (RMP) (middle, n = 18), and PD mice with RMP > -58 mV (bottom, n = 14). d, e RMP (d left), percentage of neurons with RMP > -58 mV (d right) and action potential frequency (e) were calculated for AD or PD neurons. f The paradigm of SWAM interruption test and relating parameter definitions. g Time within none-sleep period of PS and PD mice (n = 13 per group). h Sleep time (left) and sleep latency (right) within 24 h after SWAM interruption of PS (n = 13) and PD (n = 14) mice. i Mean absolute EEG power (top) and theta/delta ratio (bottom) during 24-h periods before (left) and after (right) SWAM interruption in PS (n = 13) and PD (n = 14) mice. Data are mean ± s.e.m. Two-way ANOVA with Dunnett’s test (b, e); one-way ANOVA with Dunnett’s test (d left); two-sided Chi-square (d right); unpaired t-test (g, h); two-way ANOVA with Sidak’s test (i). *P < 0.05; †P < 0.01; ‡P < 0.001; NS, not significant, P > 0.05. n refers to the number of biological replicates.

Fig. 3. Disturbance of brain phosphoproteome during PONE status.

a Experimental design for quantitative proteomic and phosphoproteomic studies of Pr-SD model. b–d Volcano plots displaying changed phosphopeptides in PS/AD (b), PD/PS (c) and PD/AD (d) groups. In, increased [log₂(ratio) > 0.3]; De, decreased [log₂(ratio) < −0.3]. Multiple unpaired t-test (P value) followed by false discovery rate (FDR) (Q value) analysis. e, f Analysis of mean abundance of increased (e) or decreased (f) phosphopeptides in PD/AD comparison (d). g, h Venn diagram (g), hierarchical cluster and classification analysis (h) of 317 significantly changed (Q < 0.2) phosphopeptides among three groups with total number in parentheses (g). Pr-SD represents all significantly changed peptides in PD/PS or PD/AD groups. i–l Global ΔPs analysis of phosphoproteins in PD/PS (i) and PD/AD (j) groups, and quantitative ΔPs analysis of SNIPPs in PD/PS (k) and PD/AD (l) comparisons. Dotted lines represent ΔPs = ±2.4. Numbers of total, hyperphosphorylated (Hyper) and hypophosphorylated (Hypo) proteins are shown. m Venn diagram showing overlaps of Hyper-phosphoproteins (ΔPs > 2.4, left) and Hypo phosphoproteins (ΔPs < −2.4, right) among three models. n Synaptic percentage analysis for class A–C phosphoproteins (left), Hyper-, Hypo- and PD-SNIPPs phosphoproteins (right). The dot size represents the number of overlapped kinase motifs and dot color shows enrichment factors. One-way ANOVA with Tukey’s test (e, f); two-tailed unpaired t-test (k, l); two-sided Chi-square (n). *P < 0.05; †P < 0.01; ‡P < 0.001.

Fig. 4. The effect of SLP on sleep and PONE status development.

a Schematic diagram of the protein structure of SIK3 and truncated proteins for AAV expression, and the experimental design for EEG/EMG recording and survival tests. b–d Sleep analyses for hSyn-GFP, hSyn-A221 and hSyn-sSLP group (n = 13 per group). 24-h absolute delta power of NREMS before (pre) and after (post) hSyn-sSLP injection (b). Mean absolute NREMS delta power (left) and the change ratio [(‘post’- ‘pre’)/ ‘pre’] (right) in mice injected with hSyn-GFP, hSyn-A221 and hSyn-sSLP (c). Duration (left) and time difference (right) of NREMS in mice injected with hSyn-GFP, hSyn-A221 and hSyn-sSLP (d). e, f Analysis of cumulative mean PONE index (e) and survival (f) of SWAM mice injected with hSyn-GFP (n = 17), hSyn-A221 (n = 16) and hSyn-sSLP (n = 14). g–i Sleep analyses for hSyn-GFP (n = 6), Camk2a-sSLP (n = 9), mDlx-sSLP (n = 7) and GfaABC1D-sSLP (n = 7). 24-h absolute delta power (g); change ratio of mean absolute NREMS delta power (h); time difference of NREMS duration (i). j, k Analysis of cumulative mean PONE index (j) and survival (k) of SWAM mice injected with hSyn-GFP (n = 13), Camk2a-sSLP (n = 18), mDlx-sSLP (n = 13) and GfaABC1D-sSLP (n = 13). Data are mean ± s.e.m. Two-way ANOVA with Sidak’s test (b, g); two-tailed paired t-test (c, d, left); one-way ANOVA with Tukey’s test (c, d, right); log-rank (Mantel-Cox) test (f, k); one-way ANOVA with Dunnett’s test (h, i). *P < 0.05; †P < 0.01; ‡P < 0.001; NS, not significant, P > 0.05. n refers to the number of biological replicates.

Fig. 5. The effect of PP2A on sleep and PONE status development.

a Volcano plots of quantified phosphopeptides of PP2A specific substrates in PD/AD comparison. b Relative abundances of the specific substrates for three phosphatases in sleep-related comparisons. c Volcano plots of quantified phosphopeptides of PP2A family in PD/AD comparison. d 24-h (left) and mean (right) relative NREMS delta power of Ppp2r5d (S565A) homozygous (HOMO, n = 14) and WT (n = 13) littermates. e, f Analysis of cumulative mean PONE index (e) and survival (f) for Ppp2r5d (S565A) littermates (n = 15 per group). g 24-h absolute delta power (left) and the change ratio (right) of NREMS in mice injected with hSyn-GFP (n = 10) and Camk2a-SV40 small T antigen (n = 8). h, i Analysis of cumulative mean PONE index (h) and survival (i) of SWAM mice injected with hSyn-GFP (n = 12) and Camk2a-SV40 (n = 11). Data are mean ± s.e.m. Two-way ANOVA with Sidak’s test (d, g, left); two-tailed unpaired t-test (d, g, right); log-rank (Mantel-Cox) test (f, i). *P < 0.05; †P < 0.01; ‡P < 0.001; NS, not significant, P > 0.05. n refers to the number of biological replicates.

Fig. 6. The effect of brain phosphoproteomic reconstruction on PONE status development.

a Schematic of the 4-h recovery period (RP4h) experiment. b, c Analysis of hourly duration of NREMS (b) or REMS (c) for BLD (ZT0-3) and the average of all RP4 (n = 16). d Total sleep time on BLD (ZT0–3), 1st-RP4, L-RP4 (last-RP4) and the average (Ave) of all RP4 (n = 16). e, f Cumulative PONE index (e) and population survival (f) of SWAM mice without (SWAM) or with recovery (RP4, n = 10 per group). g Schematic diagram of the RP4h phosphoproteomic study. h Volcano plot showing all phosphorylation changes in the RP0/PS group. In, increased; De, decreased. i, j Mean abundance analysis of increased (i) and decreased (j) phosphopeptides in the RP0/PS comparison under three conditions. k Global ΔPs analysis in RP0/PS comparison. l Comparative ΔPs analysis of 47 PD-SNIPPs between RP0/PS and RP4/PS comparisons. m Global kinase motif enrichment analysis of phosphoproteomic data from RP4/PS groups. En, enriched; De, depleted. n–q Cumulative PONE index (n, p) and survival analyses (o, q) of SWAM mice with RP4h and injected with AAV. n = 11 (n, o, GFP, sSLP and A221); n = 10 (p, q GFP and GFP&sSLP); n = 9 (p, q GFP&PPP2CA). r A model illustrating homeostasis of brain phosphoproteome linking sleep regulation and functions. Data are mean ± s.d. Two-way ANOVA with Sidak’s test (b, c); one-way ANOVA with Tukey’s test (i, j); log-rank (Mantel-Cox) test (o, q); two-tailed unpaired t-test (l). *P < 0.05; †P < 0.01; ‡P < 0.001; NS, not significant, P > 0.05. n refers to the number of biological replicates.