A familial natural short sleep mutation promotes healthy aging and extends lifespan in Drosophila

Summary Sleep loss typically imposes negative effects on animal health. However, humans with a rare genetic mutation in the dec2 gene (dec2P384R) present an exception; these individuals sleep less without the usual effects associated with sleep deprivation. Thus, it has been suggested that the dec2P384R mutation activates compensatory mechanisms that allows these individuals to thrive with less sleep. To test this directly, we used a Drosophila model to study the effects of the dec2P384R mutation on animal health. Expression of human dec2P384R in fly sleep neurons was sufficient to mimic the short sleep phenotype and, remarkably, dec2P384R mutants lived significantly longer with improved health despite sleeping less. The improved physiological effects were enabled, in part, by enhanced mitochondrial fitness and upregulation of multiple stress response pathways. Moreover, we provide evidence that upregulation of pro-health pathways also contributes to the short sleep phenotype, and this phenomenon may extend to other pro-longevity models.


1
Sleep loss typically imposes negative effects on animal health. However, humans with a rare 2 genetic mutation in the dec2 gene (dec2 P384R ) present an exception; these individuals sleep less 3 without the usual effects associated with sleep deprivation. Thus, it has been suggested that 4 the dec2 P384R mutation activates compensatory mechanisms that allows these individuals to 5 thrive with less sleep. To test this directly, we used a Drosophila model to study the effects of 6 the dec2 P384R mutation on animal health. Expression of human dec2 P384R in fly sleep neurons 7 was sufficient to mimic the short sleep phenotype and, remarkably, dec2 P384R mutants lived 8 significantly longer with improved health despite sleeping less. The improved physiological 9 effects were enabled, in part, by enhanced mitochondrial fitness and upregulation of multiple 10 stress response pathways. Moreover, we provide evidence that upregulation of pro-health 11 pathways also contributes to the short sleep phenotype, and this phenomenon may extend to 12 other pro-longevity models. Sleep is an ancient behavior that is universally conserved among the animal kingdom 1 .
2 However, a high degree of variability exists in the amount of time different species spend 3 sleeping 2 . Some species, such as C. elegans, only sleep during critical developmental 4 transitions or injury 3 , while others, including many bat species, spend most of their life sleeping 5 2 . Although work in the past few decades has led to a better understanding of the molecular 6 mechanisms governing sleep homeostasis [4][5][6] , why organisms require a certain amount of sleep 7 is still a fundamental mystery.

8
In most species, it is evident that sleep is important for maintaining physiological health 9 as inadequate sleep correlates with numerous health issues, such as hypertension, heart 10 disease, metabolic disorders, cognitive impairment, neurodegenerative diseases, and even 11 premature mortality 7-16 . Moreover, a bidirectional relationship between aging and sleep exists;  been suggested that individuals harboring the dec2 P384R mutation may employ compensatory 8 mechanisms that allow them to thrive with chronic sleep loss. However, whether the dec2 P384R 9 mutation directly confers global health benefits has not yet been tested experimentally in any 10 system.

11
In this study, we used a Drosophila model to understand the role of the dec2 P384R 12 mutation on animal health and elucidate the mechanisms driving these physiological changes.

13
We found that the expression of the mammalian dec2 P384R transgene in fly sleep neurons was 14 sufficient to mimic the short sleep phenotype observed in mammals. Remarkably, dec2 P384R 15 mutants lived significantly longer with improved health despite sleeping less. In particular, 16 dec2 P384R mutants were more stress resistant and displayed improved mitochondrial fitness in 17 flight muscles. Differential gene expression analyses further revealed several altered 18 transcriptional pathways related to stress response, including detoxification and xenobiotic 19 stress pathways, that we demonstrate collectively contribute to the increased lifespan and 20 improved health of dec2 P384R mutants. Finally, we provide evidence that the short sleep 21 phenotype observed in dec2 P384R mutants may be a result of their improved health rather than 22 altered core sleep programs. Taken together, our results highlight the dec2 P384R mutation as a 23 novel pro-longevity factor and suggest a link between pro-health pathways and reduced sleep  The human dec2 genes were PCR-amplified from dec2 WT and dec2 P384R transgenic strains using 11 the following primers:5'-GGGGACAACTTTGTATACAAAAGTTGTAATGGACGAAGGAATTCCT 12 CATTTGC-3' and 5'-GGGGACCACTTTGTACAAGAAAGCTGGGTATCAGGGAGCTTCCTTTC 13 CTGGCTGC-3'. dec2 WT and dec2 P384R transgenes were subsequently cloned into the pDONR 14 P5-P2 Gateway vector (Invitrogen) using BP clonase (ThermoFisher Scientific, Cat# 11789020),

21
Injection services of Genetivision (Houston, TX) were used to insert transgenes at the vk27: (3r) 22 89e11 site using phiC31-mediated insertion. Transgenic lines were maintained over the 3 rd 23 chromosome balancer TM6B using standard genetic crossing procedures. Trikinetics (Waltham, MA). In flies, sleep is defined as a quiescent period of five mins or longer 2 27 . Male flies (7-10 days old) were loaded in 5 × 65 mm glass tubes with food on one side and 3 were allowed to acclimate for approximately 24 hrs. Baseline sleep was measured as bouts of 4 5 min of rest and was recorded for 5 days. During the analysis, flies were subjected to a 12:12h 5 L:D cycle in an incubator at 25 o C. Sleep data was analyzed using ShinyR-DAM software 28 . For 6 sleep rebound, baseline sleep was recorded for 24h before flies were subjected to 24h of sleep 7 deprivation using a sleep nullifying apparatus 29 , which tilts asymmetrically from -60° to +60° 8 angle to mechanically displace flies 10 times per min. Rebound sleep was then recorded for 24h 9 post-deprivation.

24
To test memory function, we used an Aversive Phototaxis Suppression Assay. This assay is 25 based on the principle that flies are naturally attracted to light, except when an aversive odor 26 (Quinine hydrochloride dihydrate, MP Biomedicals) is simultaneously present. For each 1 experiment, ∼20 adult male flies were transferred to an empty vial and starved for 6h before 2 conducting the experiment to promote active foraging during the experiment. Before each 3 experimental trial, flies were tested to determine whether they were positively phototaxic under 4 normal conditions; flies were acclimated to the dark chamber for 30s, and flies that failed to 5 migrate towards the light chamber after 25s were considered non-phototaxic and were censored 6 from the experiment. For the remaining flies that were phototaxic, a filter paper soaked with 7 quinine solution was inserted into the light chamber and 12 training trials were conducted. For 8 each trial, flies were allowed 60s to migrate towards the light chamber. Flies that migrated 9 towards the light chamber within 60s were scored as "Fail" and flies that stayed in the dark 10 chamber scored as "Pass". Immediately after 12 training trials, five test trials were performed to 11 test short-term memory. In the test trials, the light chamber contained filter paper soaked with 12 water. Flies that migrated towards the light chamber within 10s were scored as "Fail" and flies 13 that remained in the dark chamber were scored as "Pass". For long-term memory, the same 14 flies were kept in vials with food for 4-5h before conducting five more test trials. For each test 15 trial, the average pass rate for the five test trials was calculated for each individual fly.   For the Nanopore reads, we mapped the reads to the reference genome using Minimap2 31 with 11 arguments (p=80 and N=100) as described in 32 . We then used Salmon 33 to quantify gene 12 expression in alignment-based mode. For both the Illumina and Nanopore data, differential  Hochberg's approach for controlling the false discovery rate. Genes with an adjusted p-value ≤ 18 0.05 and fold-change ≥ 1.5 found by DESeq2 were assigned as differentially expressed. We

4
RNA was extracted from whole Drosophila animals and converted to cDNA using iScript™ 5 cDNA Synthesis Kit (Bio-Rad, cat#1708891). Primers for qPCR were designed using IDT 6 PrimerQuest (Table S1). Three experimental replicates per strain were analyzed and Actin was 7 used as a housekeeping gene. qPCR was conducted using PowerUp TM SYBR TM Green Master 8 Mix (ThermoFisher Scientific). qPCR was performed on a QuantStudio 6 Real-Time PCR 9 system. Data were analyzed using standard ΔCT method. The 2 -DDCT method was used to 10 estimate the relative changes in gene expression. Data were normalized to the WT control.

10
Nevertheless, these data indicate that dec2 P384R mutants still exhibit age-dependent changes in 11 sleep.

12
We also examined the effect of dec2 P384R expression on sleep homeostasis, a regulatory 13 mechanism that governs the timing and amount of sleep in a 24hr circadian period 52 . Normally, 23 mutants did not display a significant change in sleep consolidation ( Fig. 1H and 1I)

4
In humans, patients with a rare genetic disease called Fatal Familial Insomnia lose the ability to 5 sleep around mid-life and only survive on average 18 months after diagnosis 15 . In less severe 6 instances, chronic sleep deprivation is associated with developmental disorders, cognitive 7 impairments, metabolic dysfunctions, physiological deficits, cardiovascular diseases, and 8 neurodegenerative diseases 7,9-13,16,24,53,54 . This prompted us to explore whether chronic reduced 9 sleep in the dec2 P384R mutants had any negative health impacts. Remarkably, we found that 10 mutant dec2 P384R flies lived significantly longer compared to control flies ( Fig. 2A and Table S2).

11
Thus, despite sleeping less, the dec2 P384R mutation might in fact confer longevity to the 12 organism.

13
To investigate whether dec2 P384R mutants have improved physiology, we assessed  Table S2).

6
Sleep deprivation has also been linked to poor memory consolidation; in flies, 6-12 hours 7 of sleep deprivation is sufficient to cause learning impairment 62 . This suggests that altered 8 sleep architecture can negatively impact memory encoding. Therefore, we examined whether 9 dec2 P384R mutants displayed significant memory impairment in either early-and/or mid-age by 10 performing an aversive phototaxis suppression (APS) assay (Fig. 2E), which is commonly used 11 to assess short and long-term memory in Drosophila 63,64 . At one-week of age, there was no 12 significant difference among mutants and controls (Fig. 2F). However, at three weeks of age 13 (mid-life), dec2 P384R mutants displayed significantly improved short and long-term memory 14 compared to both control groups (Fig. 2G)  if dec2 P384R mutants display altered energy production, we first measured mitochondrial 24 respiratory fluxes across the primary substrate-coupling pathways in homogenized flight 25 muscles ( Fig. 3A and 3B). dec2 P384R mutants exhibited normal OXPHOS capacity supported by 1 3C,3D and 3H). Strikingly, there was a significant increase in OXPHOS capacity supported by 2 flavin adenine dinucleotide (FAD) linked substrates, namely succinate (61% vs. dec2 WT , 81% vs.

3
WT) and glycerol-3-phosphate (25% vs. dec2 WT , 47% vs. WT) in dec2 P384R relative to both WT 4 and dec2 WT flies (Fig. 3E-3G). Consistently, the FAD pool was also depleted in dec2 P384R flies 5 relative to both controls (Fig. 3I), indicative of decreased FAD/FADH2 ratio, favoring oxygen 6 consumption and ATP synthesis. The increased respiratory flux observed in dec2 P384R mutants 7 was not attributed to a change in mitochondrial supercomplex structure and formation in 8 dec2 P384R and dec2 WT flies (Fig. S2A); however, we found a significant decrease in citrate 9 synthase activity, a marker of mitochondrial abundance, in dec2 P384R flight muscles compared to 10 controls (Fig. S2B). Thus, it is even more remarkable that dec2 P384R mutants have improved 11 mitochondrial capacity despite having less mitochondrial content. These data suggest that 12 dec2 P384R mutants exhibit improved mitochondrial respiratory function and ATP production 13 capacity of substrates linked to reduction of FAD.
14 Based upon our observations that dec2 P384R mutants display enhanced mitochondrial 15 functional capacity, we tested resistance to stress induced by complex-specific OXPHOS 16 inhibitors. We found that dec2 P384R mutants survived significantly longer when fed high doses 17 (500µM) of Rotenone ( Fig. 3J

24
Multiple stress response genes are upregulated in dec2 P384R mutants 25 1 and improved health of dec2 P384R mutants might be due to global changes in gene expression.

2
To examine this possibility, we performed Illumina-based RNA-sequencing to identify 3 differentially expressed genes (DEGs) in dec2 P384R vs. WT and dec2 WT flies. One week old flies 4 were collected at ZT3, the time in which we observed the most significant difference in their 5 daytime sleep (Fig. 1A), and RNA was extracted from whole animals for sequencing. In parallel, 6 we also performed long-read sequencing using Nanopore technology, which enables whole 7 transcript sequencing and can identify isoform variants and limits amplification biases 73 .

8
Significantly, the two analyses shared ~50% overlap in the DEGs identified (Fig. 4A), 9 underscoring the confidence and reproducibility of our datasets. Principal component analyses 10 were plotted to visualize the difference in gene expression among the three groups: WT, dec2 WT 11 and dec2 P384R (Fig. S3A). The Illumina analyses obtained RNA-seq profiles for 17,972 genes 12 with 323 DEGs in dec2 P384R vs WT and 121 DEGs in dec2 P384R vs dec2 WT (Fig. 4B, 4C, and 13 Table S3), while the long-read Nanopore sequencing obtained RNA-seq profiles for 15,488 14 genes with 136 DEGs in dec2 P384R vs WT and 43 DEGs in dec2 P384R vs dec2 WT (Fig. S3B, S3C,   15 and Table S4). To begin deciphering the molecular pathways that may be contributing to the 16 improved health and extended lifespan of dec2 P384R mutants, we performed gene ontology (GO) 17 and KEGG pathway enrichment analyses ( Fig. S4A and S4B). Notably, multiple gene clusters 18 related to stress resistance were upregulated in dec2 P384R mutants (Fig. 4B-4D), which could 19 account for the improved physiological health observed in dec2 P384R mutants. Moreover, we 20 identified several uncharacterized and orphan genes that were differentially expressed in 21 dec2 P384R mutants (Fig. 4B-4E), which could represent novel pro-longevity factors.

22
To validate the RNA-seq data, we selected the top ten upregulated genes as well as a 23 subset of related gene family members to quantify expression by qPCR (Fig. 5A-5C and S5A-24 S5F). We first examined the metallothionein (MT) gene family, which consists of five paralogs 25 (mtnA-E) that reside in a gene cluster on Ch. 3R (Fig. 5A). MT proteins have known cytoprotective functions and promote cell survival with increased expression [74][75][76][77] . Consistent 1 with the RNA-seq data, mtnB and mtnD transcripts were increased in dec2 P384R compared to 2 both controls (Fig. 5A). Methuselah-like 8 (mthl8) is an uncharacterized gene that is predicted to 3 encode a G protein-coupled receptor 78 and was the most upregulated gene in dec2 P384R vs WT 4 in both Illumina and Nanopore datasets (Fig. 4B and S3B). Notably, a related homolog 5 methuselah has been linked to lifespan regulation in flies 79 . Strikingly, mthl8 transcripts 6 measured by qPCR were increased >1000-fold in both dec2 P384R and dec2 WT compared to WT 7 (Fig. 5B). Finally, we examined expression of CG11699, which was upregulated in dec2 P384R 8 compared to both controls; CG11699 transcripts were increased in dec2 WT compared to WT and 9 further upregulated in dec2 P384R (Fig. 5C). Although CG11699 is not well-characterized, it has 10 been linked to lifespan regulation; a transposable element insertion in the 3'UTR increases 11 CG11699 expression and extends lifespan 80 . CG11699 is also related to human TMEM242, a phenotype like the GAL4/UAS system and, indeed, we observed a similar short sleep 25 phenotype when dec2 P384R was expressed using the LexA/LexAop system (Fig. S6A-S6E).
GR23E10-lexA/lexAop-dec2 P384R transgenic flies were also resistant to Rotenone (Fig.   1   5D and Table S2) but still do not survive for more than three days. Therefore, we performed 2 lifespans in the presence of Rotenone as a faster means of screening through candidate genes 3 initially. We first examined the MT gene family mtnA-E. Upregulation of MT genes in neurons 4 promotes longevity 82 ; thus, we used the pan-neuronal driver elav-gal4 to inhibit MT gene 5 expression in the brain of dec2 P384R flies. Strikingly, inhibition of mtnB alone was sufficient to 6 diminish the lifespan extension effect of dec2 P384R mutants back to control lifespans ( Fig. 5D and   7 Table S2), while there was no significant difference in lifespan with suppression of mtnC or 8 mtnD (Fig. 5E, 5F and Table S2). This is consistent with the DGE analyses, as mtnB was the 9 most differentially expressed MT gene compared to both WT and dec2 WT controls ( Fig. 4B and   10 4C). We next tested CG11699 and nmdmc by inhibiting their expression ubiquitously using 11 tubulin-gal4 (tub-gal4). However, we did not obtain any viable progeny, suggesting that global 12 inhibition of these genes is lethal. We then inhibited CG11699 or nmdmc in neurons using elav-13 gal4. Inhibiting CG11699 in dec2 P384R mutants reduced the lifespan back to WT (Fig. 5G and 14 Table S2), while there was no reduction in lifespan with nmdmc gene suppression ( Fig. 5H and 15 Table S2). Finally, we examined whether mtnB is required for dec2 P384R mutant lifespan increasing orexin expression in mammals 22 , it is puzzling that over-expression of mammalian 2 dec2 P384R in Drosophila can still induce a short-sleep phenotype given that the orexin system 3 does not exist in invertebrates 83 . This suggests that dec2 P384R is capable of reducing sleep by 4 an orexin-independent mechanism. This led us to postulate that perhaps the dec2 P384R -5 dependent short sleep phenotype is not directly related to altered core sleep mechanisms, but 6 rather a byproduct of their increased longevity. Based on this idea, we hypothesized that 7 inhibiting the pro-health pathways triggered by Dec2 P384R would reverse the short sleep 8 phenotype. Thus, we inhibited mtnB pan-neuronally in dec2 P384R mutants, which reduces the 9 lifespan of dec2 P384R mutants back to WT (Fig. 5I), and assessed their sleep length. In accord 10 with our hypothesis, we found that mtnB inhibition increased sleep of dec2 P384R mutants back to 11 WT levels (Fig. 6A-6B). Moreover, inhibition of mtnB also suppressed the sleep fragmentation 12 phenotype of dec2 P384R mutants (Fig. 6C-6D). Thus, these data suggest that the improved 13 health of dec2 P384R mutants may also contribute to the short sleep phenotype.
14 Based on our result that reducing pro-health pathways in dec2 P384R mutants reverses the 15 short sleep phenotype, it is intriguing to speculate that improving organismal health may reduce  (Fig. 6E and 6F). We also observed increased sleep fragmentation in 1 both mutant models compared to controls (Fig. 6G and 6H). Interestingly, fat body 2 overexpression of foxo displayed increased nighttime sleep, which is consistent with the 3 previous observation that inhibiting insulin signaling reduces daytime sleep, but promotes a 4 compensatory increase in nighttime sleep 86 . However, we did not observe similar 5 compensatory increases in nighttime sleep of AMPK or dec2 P384R models, suggesting that sleep 6 may be differentially influenced in these models. Nevertheless, these results lend further 7 support to the notion that inducing pro-longevity pathways may reduce sleep pressure. or Rotenone (Fig. 6I-6L). Taken together, these data lend support to the idea that improving 16 health might reduce sleep need.

19
In this study we identified a familial natural short sleep mutation as a pro-longevity factor.

20
While it has been suggested that human natural short sleepers are able to thrive with chronic 21 short sleep, this has never been directly tested experimentally. Using a Drosophila model, we  Although there are likely multiple genes that collectively contribute to the lifespan 7 extension of dec2 P384R mutants, we found that increased expression of mtnB, a metallothionein 8 protein, is one critical gene required for the full lifespan extension effects of dec2 P384R mutants.

9
Metallothionein proteins are small proteins that mediates cellular stress responses and are 10 linked to longevity 97,98 . Notably, increased expression of metallothionein results in resistance to 11 mitochondrial induced stress and prevention of apoptotic signaling 82,97-99 . This is consistent with 12 our observations that dec2 P384R mutants are resistant to mitochondrial inhibitors ( Fig. 3J and   13 3K). We also observed upregulation of the CG11699 gene, which transcribes a protein that is 14 not fully characterized. However, in flies, increased expression of CG11699 confers xenobiotic 15 stress resistance through increased aldehyde dehydrogenase type III (ALDH-III) activity 80 .

16
ALDH oxidizes aldehydes to non-toxic carboxylic acids mitigating both intrinsic and pathological 17 cellular stress, thus promoting overall survival 100 . Additionally, a closely related human homolog 18 of CG11699, TMEM242, is required for the assembly of the c-8 ring of human ATP synthase, 19 which is essential for ATP production 81 . Consistently, we found that dec2 P384R mutants have 20 increased mitochondrial respiratory capacity ( Fig. 3A-3I). Specifically, we found improved FAD-21 linked capacity with a concomitant decrease in the FAD-pool, indicating an overall increase in 22 FAD oxidation and ATP production. While increased FAD oxidation can also result in increased 23 oxidative stress, we have found that the dec2 P384R mutants are able to capitalize on the 24 increased capacity while mitigating the potential deleterious effects of oxidative stress through 25 upregulation of multiple stress-response mechanisms.

26
Our results also indicate that expressing the dec2 P384R mutation in neurons alters cellular 1 physiology in other non-neuronal tissues, such as muscles (Fig. 3). These data suggest that 2 Dec2 P384R triggers cellular responses in a cell non-autonomous manner to elicit systemic 3 changes. How might this occur? Dec2 is a transcription factor that regulates multiple circadian 4 genes 101 , many of which are known to affect organismal health and survival. For example,

19
The Drosophila genome encodes a single gene, clockwork orange (cwo), that is 20 orthologous to mammalian dec1 and dec2 107 . Similar to DEC proteins, CWO also antagonizes 21 CLOCK/BMAL transcription factors at an E-box site to attenuate period expression 108 . Although

22
CWO is structurally similar to Dec2, containing a basic helix-loop helix domain, there is less 23 than 18% amino acid sequence similarity with Dec proteins, and the proline 384 residue is not 24 conserved in the CWO protein. Thus, there are likely to be functional distinctions between the 25 orthologs. Nevertheless, the fact that mammalian dec2 P384R induces short sleep and impacts 1 multiple aspects of physiology when expressed in flies, signifies that it is acting in a dominant 2 negative fashion and could interfere with expression of endogenous CWO target genes.

3
Alternatively, the proline mutation could produce a more dramatic structural alteration to Dec2, 4 causing it to bind ectopic sites in the genome and alter transcription of non-native CWO target 5 genes. Having a deeper understanding of endogenous Dec2 and CWO target genes, perhaps 6 with a focus on non-circadian regulatory networks, will be important to decipher how dec2 7 orthologs and their variants influence non-sleep phenotypes.

8
Finally, our results also suggest that the improved health in dec2 P384R mutants may also 9 contribute to the short sleep phenotype. Typically sleep loss is associated with reduced health 10 and lifespan 14,109,110 ; however, there is evidence to suggest that this may not always be the 11 case. In a study using a sleep inbred panel in which flies were sorted based on their natural

16
The fact that some species evolved mechanisms to virtually eliminate the need for sleep, while 17 maintaining a similar lifespan as related species that require sleep 19 , lends support to this idea.

18
Perhaps these species have naturally adapted sleep-independent pro-health mechanisms that 19 allows them to survive with less sleep. This might also help explain why expression of the 20 mammalian dec2 P384R transgene can still induce a short sleep phenotype in flies, despite lacking 21 an orexin ortholog. Thus, we hypothesize that the pro-health pathways that are ectopically 22 induced in dec2 P384R mutants may also contribute to the short sleep phenotype in flies. Whether 23 similar mechanisms occur in mammals will be important future studies.

24
Sleep loss is becoming endemic in our modern society; it is estimated that 30% of adults 1 in the U.S. sleep an average of 6hrs/night or less and are chronically sleep deprived 111,112 .
2 These sleep disturbances are becoming even more prevalent due to certain occupational, and 3 lifestyle demands (i.e., shiftwork, cross time-zone travel). Thus, sleep loss has become a major 4 public health concern and uncovering mechanisms that can sustain health in sleep-deprived 5 states is of critical importance. Studying the genetic mechanisms regulated by these rare short 6 sleep mutations could provide a unique opportunity to not only understand how these exceptional 7 individuals offset the negative effects of sleep deprivation, but also uncover novel pro-longevity 8 pathways that could be co-opted to sustain health in sleep-deprived states as well as promote 9 health more generally.

12
We thank all lab members for critically reading the manuscript and providing helpful feedback.