**Background:** Age-associated sleep changes, including fragmentation and excessive daytime sleepiness, are well-documented in both humans and mice. The hypothalamus, particularly the dorsomedial hypothalamus (DMH), plays a critical role in sleep-wake regulation and aging. Previous work showed that Sirt1 signaling in the DMH/lateral hypothalamus delays aging and reduces sleep fragmentation. The transcription factor Prdm13, regulated by Sirt1 and enriched in the DMH, was hypothesized to be causally involved in age-related sleep alterations. Dietary restriction (DR) is a known anti-aging intervention, but its effect on sleep and the role of Prdm13 in mediating these effects was unknown.
**Methods:** The authors used EEG/EMG recordings to assess sleep-wake patterns in young (4 mo) and old (20 mo) C57BL/6J mice, DMH-specific Prdm13-knockout (DMH-Prdm13-KO) mice, and mice with chemogenetic inhibition of Prdm13+ DMH neurons (using hM4Di/CNO). Sleep deprivation (SD) was performed for 6 hours by gentle handling. cFos immunohistochemistry and RNAscope in situ hybridization identified neuronal activation. DR was implemented as 60% of ad libitum intake for 14-28 days in old mice. Prdm13 was overexpressed in the DMH of old mice via lentivirus. Transcriptional activity was assessed using luciferase reporter assays in NIH3T3 cells. Lifespan was analyzed by Kaplan-Meier curves.
**Key Results:** Old mice showed increased sleep fragmentation (higher number of wake/NREM episodes, shorter episode durations) and spent less time awake and more time in NREM sleep during the dark period. During SD, old mice exhibited significantly more sleep attempts than young mice (repeated measures ANOVA: factor age F(1,9)=5.989, P=0.0369). The initial increase in SWA after SD was significantly higher in old mice (repeated measures ANOVA: factor age × time F(5,45)=7.162, P<0.0001). cFos+ cells in the DMH increased during SD, and 58% of Prdm13+cFos+ neurons were in the medial DMH. Chemogenetic inhibition of Prdm13+ DMH neurons significantly increased sleep attempts during SD in Prdm13-CreERT2 mice (repeated measures ANOVA: F(1,20)=7.744, P=0.0115). DMH-Prdm13-KO mice at 4 months recapitulated old WT phenotypes: shorter wake/NREM episode durations and significantly more sleep attempts during SD (repeated measures ANOVA: factor genotype F(1,25)=9.131, P=0.0057). At 20 months, DMH-Prdm13-KO mice showed exacerbated sleep fragmentation, significantly lower EEG power during wakefulness (4-12 Hz, P=0.0497), more sleep attempts during SD (P=0.0462), higher SWA after SD (P=0.0010), increased body weight (P<0.05), larger adipocytes (P<0.01), lower wheel-running activity (P=0.014), and shortened lifespan (P=0.0178 by log-rank test). DR in old WT mice significantly reduced the number of wake/NREM episodes, increased episode durations, and suppressed sleep attempts during SD (repeated measures ANOVA: factor diet F(1,9)=5.131, P=0.0498). These DR benefits were abrogated in DMH-Prdm13-KO mice (sleep attempts: factor diet F(1,14)=2.918, P=0.1097). Overexpression of Prdm13 in the DMH of old mice significantly reduced sleep attempts during SD (P<0.05). Prdm13 was localized to the nuclear scaffold fraction, confirming its role as a transcription factor. Prdm13 upregulated promoters of Cck, Grp, and Pmch in a dose-dependent manner via its zinc finger domain. Prdm13+Cck+ DMH neurons showed significantly higher cFos activation during SD in young mice (P<0.001), but this response was blunted in old mice.
**Clinical Implications:** This study identifies Prdm13 signaling in the DMH as a critical mediator of age-related sleep fragmentation and excessive sleepiness, and as a necessary component for the beneficial effects of dietary restriction on sleep. The findings suggest that targeting Prdm13+ DMH neurons or their downstream targets (e.g., Cck) could represent a novel therapeutic strategy for improving sleep quality in the elderly. The data also provide a mechanistic link between sleep dysfunction and broader age-related physiological decline, including increased adiposity and reduced physical activity.