Single-cell profiling of skeletal muscle reveals a novel senolytic target: CRYAB


 Skeletal muscle mass and function can decline with aging, resulting in a syndrome known as sarcopenia. This decline is linked to functional alterations in critical cell types within mature muscle, including fibro-adipogenic progenitors (FAPs) and satellite cells (SCs), driven in part by cellular senescence. We utilized single-cell RNA sequencing and isolated FAPs and SCs to identify novel targets responsible for senescent cell killing - senolysis. We identified the small alpha-crystalline heat shock protein CRYAB as a novel senolytic target. Using chemical inhibitor screening of CRYAB, we identified 25-hydroxycholesterol (25HC), an endogenous metabolite of cholesterol biosynthesis, as a potent senolytic capable of killing senescent cells. We validated 25HC as a senolytic in mouse and human cells in culture and in vivo in mouse skeletal muscle. Thus, 25HC represents a potential new class of senolytics, which may be useful in combating diseases or physiologies in which cellular senescence is a key driver.

Cellular senescence causes proliferative cells to undergo an essentially permanent cell-1 cycle arrest upon encountering stressful stimuli, such as genotoxic chemotherapy, radiation or 2 telomere shortening [1][2][3] . Senescent cells accumulate with age in many organs and create a pro-3 inflammatory milieu via a senescence-associated secretory phenotype (SASP) 4 . The widely 4 used chemotherapeutic agent Doxorubicin (Doxo) induces senescence in mice and cell culture 5 1 , and is associated with severe side effects, such as cardiotoxicity, muscle weakness, fatigue 6 and cognitive impairment [5][6][7][8] . Genetic ablation of senescent cells in mice improves function in 7 several tissues 1 . Consequently, there is an active search for drugs that kill senescent cells 8 (senolytics). This new class of drugs improves health span in multiple contexts of aging 9-14 .

13
Skeletal muscle is among the largest organs in the human body and provides a means to   analyses. Single-cell sequencing and cell identity assignation via transcriptional profiling is 21 increasingly applied to diverse tissues of heterogeneous cell makeup [23][24][25][26] , including skeletal 22 muscle, heart and brain [27][28][29] . We leveraged the power of this approach to determine 23 transcriptional profiles of sub-populations of skeletal muscle cells isolated from mature tissue to 24 identify senescent cell sub-populations. We also refined this approach to identify novel 25 senolytic targets and drugs.

26
In mice, misfolding and aggregation of the crystallin proteins CRYAA and CRYAB leads to 27 cataracts 30 . 25-Hydroxycholesterol (25HC) slowed cataract development by blocking CRYAB 28 aggregation in mice 30 . Cataracts are a significant age-related pathology 31 , and considerable 29 evidence implicates senescence as a general driver of aging pathology 32 . CRYAB has been 30 studied in other aging pathologies, including Parkinson's disease 33 , cardiomyopathy 34 and 31 skeletal muscle decline 35 , suggesting that CRYAB might be a therapeutic target.

32
We performed single-cell RNA sequencing (scRNA seq) on SCs and FAPs from mice 33 treated with vehicle, Doxo or Doxo+ABT. Our results demonstrate that Doxo treatment induces 34 4 senescence in FAPs and SCs, and that ABT reduces the fraction of both senescent cell types.

1
In addition, using qRT-PCR and shRNA knockdowns, we identified CRYAB and HMOX1 as 2 novel and common senolytic targets in both cell types. Our results show that 25HC, which 3 blocks CRYAB aggregation 30 , is a novel senolytic drug for FAPs and SCs. Further, we 4 demonstrated the senolytic activity of 25HC in multiple other cell types from diverse human and 5 mouse organs. Finally, we showed that 25HC resulted in senolysis using a novel in vivo 6 biomarker in mice treated with Doxo and naturally aged mice. Our results suggest that scRNA 7 seq can be used to discover novel therapeutic targets, inhibitors of CRYAB can lead to potent 8 senolysis, and 25HC is a promising novel senolytic to treat senescence-associated pathologies.

2
We identified specific clusters expressing Cdkn2A (encoding p16) or CDKN1A (encoding p21) in 3 libraries from FAPs and SCs (Figure 1b,c, Table 1). To further identify potential senescent sub-4 populations, we created lists of differentially expressed genes for FAPs and SCs using the  Table 2 and 3). Genes showing this 9 expression profile were considered high-priority candidates for critical involvement in driving 10 senescence. Genes from these lists were then examined for their role in biological processes,

11
based on published studies. Genes known to induce apoptosis upon silencing or inhibitor 12 treatments were selected as potential novel senolytic targets ( Supplementary Fig. 4). We 13 identified 10 such targets with this approach, 4 from FAPs and 6 from SCs (

17
Fxyd3, Itga8 and Cav1 (Table 2), which also regulate apoptosis [42][43][44][45][46][47][48][49][50][51][52] . Cryab and Rnf7 also play 18 a role in cellular antioxidant defense 47,53 . Our next step was to validate these targets from FAPS      Next, we determined the senolytic potential for these genes by suppressing Cryab or Hmox1 31 expression using shRNA, and measuring the viability of Doxo-or DMSO-treated FAPs (both 32 non-senescent and senescent). We first optimized shRNAs for each gene by evaluating the 33 efficacy of knockdown for up to 5 sequences specific for each gene. shRNAs for Cryab or 34 6 Hmox1 significantly decreased cell viability 9 days after senescence induction ( Figure 2). In 1 agreement with our viability results, shRNAs for each gene dramatically increased extracellular 2 LDH levels in senescent FAPs at Day 9 after senescence induction, indicating robust cell killing 3 in conjunction with senescence ( Figure 2). Similar results were not possible for SCs due to an 4 inability to maintain SCs for prolonged periods in culture. LDH release was not detected in non-5 senescent FAP cells transfected with the shRNAs, indicating that senolysis was specific for 6 senescent cells expressing the target genes.

23
Collectively, these data suggest that 25HC is a broadly acting senolytic, independent of cell type

22
Our findings suggest that CRYAB is a novel gene to target for senolysis, and 25HC is a useful 23 tool compound to develop senolytic drugs. CRYAB is upregulated upon senescence in multiple 24 cell types, suggesting it can be a common senolytic target for diseases related to multiple 25 organs. We speculate that CRYAB-mediated aging pathology may be due to induction of 26 cellular senescence resulting in higher expression of CRYAB. Therefore, targeting senescent 27 cells via senolytics such as 25HC may lead to novel treatments for aging pathologies.

28
We showed broad utility for the senolytic potential of 25HC in multiple cell types, with the 29 most robust outcomes on FAPs. 25HC is an oxysterol, an oxygenated derivative of 30 cholesterol 69 . In addition to blocking CRYAB aggregation 30 , 25HC has also been tested as an 31 antiviral drug 70,71 . Notably, 25HC increases the inflammatory response, which leads to tissue 32 damage in mice after influenza infection 72 . 25HC inhibits SARS-CoV-2 replication by depleting 33 membrane cholesterol [73][74][75] . Further, the mRNA encoding the 25HC synthesizing enzyme 9 CH25H increases upon senescence induction in retina 76 . Considering 25HC has multiple roles 1 in immune response, including beneficial and harmful effects, it will be important to test 25HC in 2 mice for safety and efficacy as a senolytic in vivo. Other oxysterols, including 24HC and 3 27HC 69 , may also exert senolytic effects.

4
25HC showed stronger effects in skeletal muscle tissue compared to other tissues, and it 5 also rescued loss of muscle mass with aging. Aging is associated with a chronic state of 6 increased inflammatory cytokines, such as interleukin 6 (IL-6), which are known inducers of 7 muscle atrophy. 25HC substantially inhibited the expression of several SASP factors, including 8 IL6 in skeletal muscle (Supplemental Figure 17 and 18). This observation suggests that 25HC 9 improves age-related muscle atrophy via its senolytic effects.

10
We also identified another potential senolytic target, Heme oxygenase 1. HMOX1 belongs 11 to the heme oxygenase family, which mediates the first step in heme catabolism and plays 12 crucial roles in heme degradation 77 . We demonstrated that HMOX1 inhibition by shRNAs and 13 inhibitors causes senolysis in culture; however, the effects in vivo have yet to be characterized.
14 Further analysis using more potent inhibitors or genetic modulation is required to address this 15 issue in vivo. 16 We speculate that oxysterols may be a fruitful new class of promising senolytics with

19
In-vivo studies. All mice were maintained according to National Institutes of Health guidelines