Description

Sleep is intertwined with metabolic function in vertebrates (Tsuneki et al. 2016; Herrera et al. 2017; Aalling et al. 2018; Wilms et al. 2018) and invertebrates (Kempf et al. 2019; Ki and Lim 2019; Yurgel et al. 2019; Grubbs et al. 2020), but the molecular underpinnings of this connection are not well understood. We recently reported that the salt inducible kinase (SIK) homolog KIN-29 is required in a subset of sensory neurons for the metabolic regulation of sleep in Caenorhabditis elegans (Grubbs et al. 2020). However, since our genetic manipulations made use of mutations that were present throughout the life of the animal, it is possible that the sleep defect of kin-29 mutants reflects a requirement for KIN-29 activity during development rather than during sleep. Indeed, because kin-29 is expressed throughout development as well as adulthood (Maduzia et al. 2005), and kin-29 mutants show altered expression of targets influential in larval development (Van Der Linden et al. 2008), a role for kin-29 during development remained plausible. Distinguishing a developmental early role from a role during the time of sleep is important for constraining models for how KIN-29 regulates sleep.

To determine the KIN-29 time window of action, we depleted the KIN-29 protein in a temporally-controlled fashion. We used CRISPR/Cas9 genome editing to introduce a DEGRON fused to GFP just before the kin-29 stop codon (Fig. 1A). After confirming successful editing by Sanger sequencing, we crossed the kin-29::degron::GFP into the previously generated Peft-3::TIR1::mRuby strain (Zhang et al. 2015) to allow for rapid and reversible degradation of KIN-29 when exposed to the phytohormone auxin (Ruegger et al. 1998; Gray et al. 1999). We examined stress-induced sleep (SIS) following exposure to UVC-irradiation (Debardeleben et al. 2017) in the first day of adulthood. We first verified that auxin exposure did not disrupt SIS in the parental strains, and that the progeny containing both transgenes, from here on referred to as KIN-29 Auxin-Inducible Degron (KIN-29 AID) animals, show normal SIS in the absence of auxin (Fig. 1B). We also confirmed the expression of both transgenes in KIN-29 AID animals (Fig. 1C). We then measured SIS in KIN-29 AID transgenic worms cultivated on agar containing auxin (2 mM) from the time of hatching. KIN-29 AID transgenic animals cultivated on auxin throughout development mimicked the kin-29(oy38) null mutant SIS phenotype (Fig. 1D-E). Such animals also showed undetectable expression of endogenous GFP-tagged KIN-29 (Fig. 1F).

To determine if KIN-29’s regulation of sleep has a developmental component, we used two paradigms. First, we cultivated young adult animals for four hours on agar containing auxin (2 mM). We then administered a UVC treatment (1500 J/m²) and recorded movement quiescence of animals in the continued presence of auxin. Four hours in the presence of auxin at the young adult stage was sufficient to reduce GFP-tagged KIN-29 expression (Fig. 1F), and was able to reduce SIS to a similar extent as that seen in animals cultivated on auxin their whole life (Fig. 1F-G). Second, to restore KIN-29 function to adult animals, we cultivated KIN-29 AID transgenic animals in the presence of auxin since hatching, but then removed them from auxin-containing plates at the L4 stage. We waited twelve hours before assessing SIS using UVC-irradiation treatment. These animals showed significantly increased movement quiescence in comparison to animals that were cultured their whole life on auxin, although these quiescence levels were lower than the quiescence phenotype of wt animals (Fig. 1H). Consistent with these intermediate behavioral results, expression of GFP-tagged KIN-29 in these animals was intermediate between that of animals never exposed to auxin, and that of animals exposed to auxin throughout life (Fig. 1F). Our results are consistent with previous observations that even after 24 hours off of auxin, expression of a DEGRON-tagged protein may not fully normalize (Zhang et al. 2015).

In summary, our results demonstrate that KIN-29’s function in SIS is not developmental. Rather, these findings suggest that kin-29 functions at the time of sleep. These results fit with our proposed model of kin-29 acting as a real-time energy sensor in sensory neurons to promote sleep (Grubbs et al. 2020). Based on this analysis, we propose that the sleep-regulatory effects of SIK3 in mammals (Funato et al. 2016) are also unlikely to be developmental.