Activation of transsulfuration pathway to maintain cysteine is a thermogenic checkpoint for the conservation of energy

Pro-longevity dietary interventions such as caloric restriction (CR)1 and methionine restriction2 (MR) are associated with 'browning' of white adipose tissue in rodents, an adaptive response that increases heat production to maintain core-body temperature for the survival of homeotherms3,4. Here, the analysis of metabolome and transcriptome of adipose tissue of healthy humans5 identified that sustained caloric restriction (CR) decreases methionine cycle and lowers cysteine levels despite elevated expression of enzyme cystathionine γ-lyase (CTH), which catalyzes the synthesis of cysteine in the transsulfuration (TSP) pathway6,7. Cysteine starvation of global, but not adipocyte- or hepatocyte-specific Cth deficient mice, triggered lethal thermogenesis through conversion of white adipose tissue into uncoupled "brown"-like adipocytes. This manifests as depletion of energy reserves and drastic weight-loss. Mechanistically, cysteine starvation-induced thermogenesis and energy expenditure increases adipose noradrenaline bioavailability and induces a UCP1-independent response that partially requires FGF21. Therapeutically, reduction of cysteine reversed obesity by increasing thermogenesis and lowering inflammation. These findings establish that adaptation to dietary restriction requires activation of TSP to defend organismal cysteine levels that serves as a thermogenic checkpoint for regulation of core-body temperature and conservation of energy.


Main
The Comprehensive Assessment of Long-term Effects of Reducing Intake of Energy (CALERIE-II) clinical trial in healthy adults demonstrated that simple 14% reduction of calories for two years without any specific dietary prescription to alter macronutrient intake or meal timings can reprogram the immunometabolic axis to promote healthspan 5,8,9 . Harnessing the homeostatic pathways engaged by CR in humans may expand the current armament of therapeutics against metabolic and immune dysfunction. Induction of negative energy balance and activation of mitochondrial fatty acid oxidation by CR is thought to underlie some of its beneficial effects on healthspan 5 . However, it has also been suggested that CR-induced metabolic effects are simply due to decreased protein intake in food restricted animal models. Adding back individual amino acids to calorie restricted Drosophila, abolished the longevity effects, which were traced to the limitation of methionine, an important node for lifespan extension 10,11 . Indeed, methionine restriction (MR) in rodents increases lifespan 12 with enhanced insulin-sensitivity, adipose tissue thermogenesis and mitochondrial fatty acid oxidation 2 . Surprisingly, in long-lived Drosophila fed an MR diet, adding back methionine did not decrease lifespan and it was hypothesized that activation of methionine synthetic pathway may control longevity 11 . Commercial MR diets contain 0.17% methionine compared to normal levels of 0.86%, but notably, the MR diets also lack cystine 13,14 , another sulfur containing amino acid (SAA) which is a key substrate for protein synthesis, including synthesis of glutathione, taurine and sulfate 15 . Interestingly, in rats, MR induced anti-adiposity and pro-metabolic effects, including reduction of leptin, insulin, IGF1 and elevation of adiponectin were reversed when animals were supplemented with cysteine in the diet 16 . Furthermore, cysteine supplementation in MR rats did not restore low methionine suggesting no increase in methionine cycle 16 , where homocysteine is converted into methionine via the 4 enzyme betaine-homocysteine S-methyltransferase (BHMT) 15 . Existence of the transsulfuration pathways in mammals indicates that in case of dietary cysteine scarcity, the host shuttles homocysteine from methionine cycle, via production of cystathionine which is then hydrolyzed into cysteine by the enzyme cystathionine γ-lyase (CTH) 6,17 . It remains unclear if cysteine specifically controls organismal metabolism and whether sustained CR in healthy humans can help understand the fundamental relationship between energy balance and sulfur-containing amino acid homeostasis pathways that converge to confer longevity.

CR in humans activates the transsulfuration pathway.
Adipose tissue regulates organismal metabolism by orchestrating inter-organ communication required for healthy longevity 3,4 . To study the mechanisms that drive CR's effects on human metabolism, we conducted metabolomics analysis of the subcutaneous adipose tissue (SFAT) of participants in CALERIE-II trial at baseline and one year after 15% achieved CR 5,8,9. The PLSDA analyses of adipose tissue revealed that one year of sustained CR significantly altered the adipose tissue metabolome (Fig.1a). The metabolite sets enrichment analyses demonstrated significant increases in cysteine, methionine and taurine metabolism, which indicates activation of the transsulfuration pathway (TSP) (Fig.1b, c). To confirm the role of TSP in human CR, we reanalyzed our previously reported RNA sequencing data of humans that underwent CR 5 . These analyses revealed that compared to baseline, one and two years of CR in humans increased the adipose expression of CTH (Fig.1d) with a concomitant reduction in the expression of BHMT (Fig.   1e) suggesting reduction in methionine cycle and shift towards activation of TSP (Fig.1d).
Interestingly, even in absence of dietary restriction, prior studies have found that long-lived rodents upregulate metabolites in transsulfuration pathway that generates cysteine from methionine 18,19 .
Furthermore, data from 15 lifespan-extending interventions in rodents, including genetic longevity models, identified upregulation of enzyme cystathionine γ-lyase (CTH) as a common signature or potential biomarker of lifespan extension 20 .
To investigate the activation of the TSP by CR, we utilized unbiased metabolomics to analyze SFAT metabolites after 12 months of sustained CR. Surprisingly, despite an increase in CTH expression at one -and 2 years post CR, adipose cysteine levels were significantly reduced upon CR (Fig.1f), suggesting utilization of cysteine. Consistent with the reduced expression of BHMT, the concentration of dimethylglycine (DMG), a byproduct of BHMT was also lower (Fig.   1f). And while cystathionine, homocysteine, S-adenosylhomocysteine (SAH), Sadenosylmethionine (SAM), and methionine were not changed, CR caused a reduction in cysteine derived metabolites, γ-glutamyl-cysteine (γ-Glu-Cys), glutathione (GSH), and cysteinylglycine (Cys-Gly) (Fig.1f). Collectively, these results suggested reduction in methionine cycle and an activation of the TSP in response to depleted cysteine metabolites in response to CR (Fig.1f).

Cysteine starvation causes weight-loss in mice.
To investigate whether cysteine depletion-induced activation of the TSP plays an integral role in organismal metabolism, we developed a model to induce cysteine deficiency in mice. We created a model where cysteine becomes an essential amino acid requiring acquisition from the diet by deletion of CTH (Cth -/mice) (Extended Data Fig. 1a). Cysteine deficiency was thus induced by feeding adult Cth -/mice a custom amino acid diet that specifically lacks cystine (CysF diet), while control mice were fed an isocaloric diet that contained cystine (CTRL diet) (Fig.1g).
Utilizing this model, we found that mice with cysteine deficiency rapidly lost ~25-30% body weight within 1 week compared to littermate Cth +/+ mice fed a CysF diet or Cth -/fed a control diet ( Fig.1g). Upon clinical examination of the cysteine deficient mice, this degree of weight loss was considered a moribund state that required euthanasia. The weight-loss in mice lacking CTH and cysteine in diet was associated with significant loss of fat mass relative to lean mass in male mice (Extended Data Fig.1b). Similar weight-loss response was observed in female Cth deficient mice pair fed a cysteine free and control diet (Extended Data Fig.1c). This weight-loss is not due to malaise or behavioral alteration, as Cth -/-CysF mice displayed normal activity and food intake (Extended Data Fig.1d) and link of video file of cage activity on day 6 of 25% weight-loss). The Cth deficient mice on control diet were indistinguishable from control littermates in other parameters indicative of health such as nest building, grip strength, gait, ledge test, hindlimb clasping, and displayed no clinical kyphosis (Extended Data Fig.1e). Furthermore, compared to Cth -/mice on control diet, the analyses of heart, lung and kidneys of Cth -/-CysF mice did not reveal any pathological lesions indicative of disease (Extended Data Fig.1f). Restoration of up to 75% cysteine levels in the diet of Cth -/-CysF mice that were undergoing weight-loss was sufficient to completely rescue the weight over three weight-loss cycles, demonstrating the specificity of the response to cysteine availability (Fig.1h).
To identify systemic changes in metabolites upon cysteine deficiency, we conducted serum metabolomics analyses. Compared to Cth deficient mice fed a normal diet, the Cth -/-CysF mice had reduced cystine levels, suggesting that cysteine deficiency is maintained by a reduction in systemic cystine levels (Fig.1i).With the exception of cystathionine, which is significantly increased in CysF diet fed Cth -/mice compared to control diet fed animals, other sulfur amino acid (SAA) metabolites such as methionine, s-adenosyl methionine (SAM), s-adenosyl homocysteine (SAH), and homocysteine (HCys) are not significantly changed (Fig.1i). Notably, cysteine deficiency did not affect glutathione (GSH) concentration, a key downstream substrate in TSP ( Fig.1h). Instead, the gamma-glutamyl peptide analogs of cysteine and GSH such as 2aminobutyric acid (2AB) and ophthalmic acid (OA) were increased in the serum of cysteine deficient mice (Fig.1i). OA is a analog of glutathione in which the cysteine group is replaced by L-2-aminobutyrate. The increased flux towards OA production vs GSH production reveals adaptive changes induced by systemic cysteine deficiency (Fig.1i). Despite reduction of GSH post-CR, the CALERIE-II participants have decreased oxidative stress 9 . The lack of change in GSH levels upon cysteine-starvation in mice suggest that weight-loss is independent of potential GSHmediated redox balance.

Cysteine starvation drives adipose tissue browning.
The decrease in fat mass during cysteine deficiency (Extended Data Fig.1b) is driven by loss of all major fat depots including subcutaneous fat (SFAT), visceral epididymal/ovarian adipose fat (VFAT), and brown adipose tissue (BAT) (Fig.2a). Histological analyses revealed that this reduction in adipose tissue size is associated with transformation of whole white adipose depots into a BAT-like appearance, with the formation of multilocular adipocytes, enlarged nuclei, and high UCP1 expression, a phenomenon known as browning (Fig.2b, Extended Data Fig.2a).
Surprisingly, these morphological changes were quite robust even in the visceral adipose depot which is normally not considered to be amenable to browning responses. Consistent with an increase in the browning of the white adipose tissues, cysteine deficient animals increased their normalized energy expenditure (EE), compared to CTRL diet fed animals, particularly during the dark cycle (Fig.2c). To assess whether the increase in energy expenditure was supported by increased fat utilization, the respiratory exchange ratio (RER) was examined. The cysteine deficient animals (Cth -/mice fed CysF diet), had significantly lower RER through both the light and dark cycle (Extended Data Fig.2b), confirming that increased lipid oxidation is linked to increase in EE (Extended Data Fig.2c). Consistent with the browning of SFAT, the cysteine deficient animals show significantly increased expression of thermogenic genes such as Elovl3, Slc27a2 and Ucp1 mRNA and protein expression (Fig. 2d, Extended Data Fig.2f).
To understand the mechanisms that control the adipose tissues remodeling, we conducted the RNA-sequencing (RNA-seq) of the major adipose depots, SFAT, EFAT, and BAT, during cysteine deficiency. We found that cysteine deficiency drives clear transcriptomic changes in all three tissues, particularly in the SFAT and EFAT as displayed by the heatmap (Extended Data Fig   2d). Gene set enrichment analysis comparing Cth -/-CysF vs CysF CTRL identified that the top downregulated pathways are involved in extracellular matrix and collagen deposition, highlighting the broad remodeling of the adipose tissue (Extended Data Fig.2e). In addition, multiple metabolic pathways appear to be regulated by cysteine deficiency within the SFAT with 'respiratory electron transport chain and heat production' as the top pathway induced during cysteine deficiency (Extended Data Fig.2e). Indeed, numerous genes identified by the 'thermogenesis' GO-term pathway such as Ucp1, Cidea, Cox7a1, Cox8b, Dio2, Eva1, Pgc1, Elovl3, and Slc27a2, are differentially expressed comparing Cth -/-CTRL and CysF in the SFAT (Extended Data Fig.2f).
These results demonstrate that cysteine availability acts as a thermogenic checkpoint and inability to maintain cysteine levels drives unchecked thermogenesis.
To understand the cellular basis of adipose tissue remodeling during cysteine deficiency, we isolated stromal vascular fraction (SVF) by enzymatic digestion and conducted single cell RNA sequencing (scRNA-seq) of SFAT. We isolated SVF cells from Cth +/+ and Cth -/fed CTRL or CysF diet for 4 days to identify changes occurring before full remodeling of the tissue, with each sample pooled from 4 animals (Extended Data Fig.3a). A total of 4,666 cells in Cth +/+ CTRL; 5,658 cells in Cth +/+ CysF; 4,756 cells in Cth -/-CTRL; and 3,786 cells in Cth -/-CysF were analyzed for scRNAseq (Extended Data Fig.3b). Consistent with prior results 21,22 , the unbiased clustering revealed 15 distinct cell populations including αβ T cells, γδ T cells, ILC2s, and NK T cells, B cells, reticulocytes, mesothelial-like cells, Schwann cells, and several myeloid clusters (Extended Data Fig.3b-d). Comparison of Cth -/-CysF with other groups revealed dramatic changes in cellular composition ( Fig.2e-f). Particularly, loss of clusters 0, 1, and 2 were apparent upon cysteine deficiency (Fig.2e). Furthermore, these clusters contained the highest numbers of differentially expressed genes (Extended Data Fig.3e), highlighting them as important populations in mediating the effects of cysteine deficiency. By expression of Pdgfra, we were able to identify these clusters as adipocyte progenitors (Fig.2g). To place these clusters on a trajectory and illuminate their cell lineage, we conducted a pseudotime analysis. Trajectory analysis based on pseudotime suggested that cluster 2 may differentiate into two separate preadipocyte clusters, cluster 0 and 1 (Fig.2g).
Cth -/-CysF animals proportionally lost Clusters 0 and 1, while relatively maintaining cluster 2 compared to the other groups ( Fig.2f, g), suggesting that more differentiated preadipocytes are mobilized during cysteine deficiency. Indeed, cluster 2 expressed Dpp4, an early progenitor marker that has been shown to give rise to different committed preadipoctyes 23 (Fig.2h). Cluster 0 was enriched for both Icam1 and F3, which have been shown to be expressed by committed adipogenic, and antiadipogenic preadipocytes, respectively 21,23 (Fig.2h). Cd9, a fibrogenic marker in preadipocytes 23,24,25 , along with the collagen gene, Col5a3, were broadly expressed across clusters 0 and 1, and was specifically lost by day 4 of inducing cysteine deficiency (Fig.2h, Extended Data Fig.3f). The loss of these preadipocyte clusters were orthogonally validated by FACS (Extended Data Fig.3h). We next sought to identify beige/brown adipocyte precursors in our scRNA-seq dataset to understand whether there was an increased commitment towards brown adipocytes. Examining the expression of a CL-316,243 β3-adrenergic receptor agonist-stimulated gene signature reported by Burl et al, we found that cluster 1 had the most pronounced expression of the β3-adrenergic receptor agonist-stimulated gene signature (Extended Data Fig.3e). Clearly, Tagln, or Sm22, which has been previously described in beige adipocytes 26,27 , is specifically expressed by a subset of cells in cluster 1 (Fig.2h). Interestingly, these Tagln expressing cells are completely lost with cysteine deficiency (Fig.2h). Given the strong browning phenotype observed at day 6, it is possible that these cells become mobilized and differentiate early on during cysteine deficiency, leading to the absence of these cells as mature adipocytes are not captured within the SVF. Indeed, when we performed pathways analysis on cluster1, comparing gene expression of Cth -/-CysF with Cth -/-CTRL, we found that one of the top upregulated pathways was 'adipogenesis' (Extended Data Fig.3g). Furthermore, examination of the expression of stem associated markers and mature adipocyte markers in the adipocyte progenitor clusters revealed a clear downregulation of stem markers and an increase in mature adipocyte markers, suggesting that cysteine deficiency was driving the maturation of progenitor cells into beige-like lineage ( Fig.2i). Given the robust transformation of the white adipose tissue during cysteine starvation towards browning, differentiation of mesenchymal stem cell precursors into beige-like adipocytes may partake in this response.

Cysteine starvation-induced browning increases catecholamine bioavailability
To determine the mechanism of adipose thermogenesis caused by cysteine starvation, we next investigated the processes upstream of increased fatty acid oxidation. We first measured the activation of lipolysis in SFAT by pHSL and ATGL and found that cysteine deficiency increases ATGL expression without consistently affecting pHSL levels (Fig.3a). ATGL preferentially catalyzes the first step of triglyceride hydrolysis whereas HSL has a much broader range of substrates with a preference for diacylglycerols and cholesteryl esters 28,29 . Our data suggests that the mobilization of lipids during cysteine deficiency is mainly mediated by ATGL. Given a dramatic browning response in WAT post-cysteine deficiency, the increased ATGL is consistent with prior work that shows BAT relies heavily on the action of ATGL to mobilize lipid substrates for thermogenesis 30 .
Upstream of lipolysis, non-shivering thermogenesis is mainly activated by the sympathetic nervous system 31,32,33 . We found that noradrenaline (NA) levels were significantly increased in Cth -/-CysF SFAT, compared to cysteine sufficient controls, Cth +/+ CTRL, Cth +/+ CysF, and Cth -/-CTRL ( Fig.3b) highlighting a local increase in sympathetic activity. This increase in sympathetic activity was also observed in the BAT by application of imaging mass spectrometry (IMS) to show increased distribution of noradrenaline throughout the BAT tissue during cysteine deficiency (Fig.   3c). The analyses of catecholamine degradation enzymes revealed that increased adipose tissue NA in cysteine deficiency state was coupled with significant reduction in monoamine oxidase a (Maoa) without affecting catechol-o-methyl transferase (Comt) suggesting reduced NA catabolism ( Fig.3d). These results suggest that cysteine deficiency drives increased sympathetic activity leading to higher NA bioavailability that increases ATGL activity which provides the substrates required for mitochondrial respiration in brown adipocytes. This is further supported by a decrease in most lipid species, particularly triglycerides in the BAT of cysteine deficient mice ( Considering dramatic adipose tissue browning and elevated UCP1 expression upon cysteine starvation, we next sought to investigate whether this is a homeostatic response to defend core-body temperature or if the perturbation of temperature regulation causes hyperthermia. We measured core body temperature utilizing loggers surgically implanted into the peritoneal cavity in Cth -/mice on CTRL or CysF diet over 6 days period when animals lose weight (Fig.3f).
Surprisingly, despite conversion of WAT into brown-like thermogenic fat, the core body temperature was not different between control and cysteine deficient mice (Fig.3f, g). A recent report indicated that energy expenditure remained constant at temperatures higher than the thermoneutral point for mice, and instead core body temperatures began to increase as ambient temperatures rose 33 . Our data suggest that reduction of cysteine signals the host to activate thermogenesis and defend core body temperature within normal physiological range.
To confirm that energy expenditure was reflected in heat production in vivo, we utilized a highly sensitive and specific magnetic resonance spectroscopic imaging (MRSI) method called Biosensor Imaging of Redundant Deviation in Shifts (BIRDS) 34 to image temperature of BAT in Cth +/+ and Cth -/animals after 6 days of CysF diet. This method relies on measuring the chemical shift of the four non-exchangeable methyl groups from an exogenous contrast agent, TmDOTMA, which has a high temperature sensitivity (0.7 ppm/°C) 34 . The TmDOTMAmethyl resonance has ultra-fast relaxation times (<5ms) allowing high signal-to-noise ratio by rapid repetition for superior signal averaging. The temperature was calculated from the chemical shift of the TmDOTMAmethyl resonance according to (eq. 1 methods). We applied this method to measure in vivo local temperature differences of BAT in the context of cysteine deficiency and found that the relative temperature difference between BAT and surrounding tissue was higher in Cth -/mice compared to Cth +/+ mice (Fig.3h, i), suggesting that the adipose tissue was increasing energy expenditure to generate heat and defend core body temperature.

Cysteine-starvation induced thermogenesis and weight-loss is maintained at thermoneutrality.
Cysteine starvation revealed a metabolic crisis that signals the host to activate mechanisms for heat production to maintain core body temperature. However, across animal vivaria, including ours, mice are typically housed at sub-thermoneutral 22°C temperatures and are constantly under thermogenic stress due to slight cold challenge 33 . To further confirm that mice were indeed inducing thermogenesis to defend core body temperature, we housed cysteine deficient animals at 30°C thermoneutrality. The cysteine deficiency in Cth -/mice housed at 30°C also led to similar weight loss as 22°C with significant browning of adipose tissue (Fig.3j, k and Extended Data Fig.4c). The degree of browning and gene expression of Ucp1 and Elovl3 in CysF Cth deficient mice at thermoneutrality was relatively lower than inductions observed at 22°C (Fig.3l). However, cysteine deficient animals still significantly lost weight and maintained a browning phenotype at 30°C. Furthermore, expression of genes involved with lipid regulation and browning such as Prdm16, Ppargc1a, Ppara, Pparg, and cpt1, (Fig.3l) were significantly increased, suggesting that even at thermoneutral temperatures, Cth -/-CysF fed mice activate fat metabolism and have increased thermogenesis caused by cysteine deficiency.

Whole body depletion of cysteine drives browning in a UCP1 independent manner.
In mice, Cth is widely expressed with highest levels in the liver. Given CR in humans lowers cysteine in adipose tissue, we next generated adipocyte as well as hepatocyte specific Cth deficient mice to determine cell type specific mechanism of cysteine in weight-loss ( Fig.4a-f). Neither liver nor adipose specific deletion of Cth caused any fat-mass loss when cysteine was restricted in the diet (Fig.4c, f). These results highlight redundant expression of Cth across multiple tissues defend systemic cysteine concentration to prevent uncontrolled thermogenesis and death. In addition, TSP metabolites can potentially be generated by the gut microbiota. The Cth -/animals co-housed together with Cth +/+ mice still maintained weight loss when fed a CysF diet, suggesting that microbiota derived metabolites shuttling into TSP do not account for the weight-loss (Extended Data Fig.5a). UCP1 controls non-shivering adipose thermogenesis and is responsible for the dissipation of proton gradient across mitochondrial bilayer, thus uncoupling ATP synthesis, elevating the oxidation of lipid substrates and production of heat 35,36,37 . Since cysteine-deficiency induced UCP1 expression in WAT (Fig.2d), we next tested whether UCP1 was required for thermogenesis during cysteine deficiency. To do this, we generated Cth -/-Ucp1 -/double knockout (DKO) mice.
Interestingly, we found that Cth -/-Ucp1 -/-DKO mice lost weight at a similar rate to its Cth -/littermates on a CysF diet and displayed similar browning-like features with multilocular WAT (Fig.4g, h). The ablation of UCP1 in cysteine deficient mice lowered EE but did not affect the CBT (Fig.4i, j). The lack of UCP1 in Cth deficient mice undergoing cysteine starvation, displayed elevated ATGL and tyrosine hydroxylase (Th) expression suggesting increased lipolysis (Fig.4k, l). Despite a lack of UCP1, gene expression of the hallmark thermogenic program are significantly increased in Cth -/-Ucp1 -/-DKO compared to Cth -/in the BAT after 6 days of CysF diet (Fig.4m).
Furthermore, gene expression of other mediators of the thermogenic program such as Acadm, Cox7a1, Elovl3, and Slc27a are also significantly increased in Cth -/-Ucp1 -/-DKO mice compared to Cth -/animals fed cysteine-restricted diet (Fig.4m). These data suggest that Cth -/-Ucp1 -/-DKO mice are increasing their mobilization of lipids for mitochondrial respiration, even more so than the Cth -/-, which already rely on significantly elevated mitochondrial respiration in response to cysteine deficiency. The lack of UCP1 may force Cth -/-Ucp1 -/-DKO mice to rely on less efficient mechanisms of thermogenesis, which could potentially explain the increase in FA metabolism and mitochondrial respiration. UCP1 independent thermogenesis has been reported 3,38 . The creatine futile cycling is proposed to control heat production 38 and recently challenged for its biological relevance in driving UCP1 independent thermogenesis 39 . The creatine synthesis genes, Gatm and Gamt were not differentially induced with cysteine deficiency in Cth -/-Ucp1 -/-DKO mice compared to Cth -/animals, the expression of one of the creatine kinases that utilize ATP, Ckmt2, was significantly increased and the transporter for creatine, Slc6a8 was slightly elevated in BAT (Fig.4n). These data suggest that systemic cysteine deficiency induced thermogenesis depends on a non-canonical UCP1 independent mechanisms.

FGF21 partially mediates cysteine-starvation induced energy expenditure and weight-loss
Changes in the organ and tissue nutritional states induced by caloric restriction, methionine restriction, or low protein diets upregulate the expression of FGF21 40,41 , which is a pro-longevity hormone that regulates energy metabolism 40 . The induction of cysteine deficiency in Cth deficient mice caused dramatic increase in the levels of FGF21 in blood as well as Fgf21 mRNA expression in the liver (Fig.5a, b). To investigate the role of FGF21 during cysteine deficiency we generated Fgf21 -/-Cth -/-DKO mice. Interestingly, in absence of FGF21, cysteine deficiency-induced weightloss and reduction in adiposity in Cth -/mice was blunted but not fully rescued (Fig.5c, d). In addition, loss of FGF21 reduced the energy expenditure triggered by cysteine-starvation in Cth -/mice on CysF diet (Fig.5e). There was maintenance of lipolysis activation, as observed by levels of pHSL and ATGL in Cth -/mice, but reduced UCP1 protein and mRNA expression in WAT of Fgf21 -/-Cth -/- (Fig.5f and Extended Data Fig.5b). In agreement with the findings of UCP-1 independent browning in cysteine deficient state, the WAT of Fgf21 -/-Cth -/-DKO mice maintained classical multilocular browning characteristics (Extended Data Fig.5c). These results suggest that FGF21 is partially required for energy expenditure, thermogenesis, and weight-loss during cysteine deficiency.

Cysteine deficiency reverses western diet induced obesity in mice.
Given cysteine starvation causes potent weight loss in adult chow fed mice, we tested whether cysteine deficiency could be utilized to induce adaptive thermogenesis mechanism for fat mass reduction in an obesity model. The Cth -/mice that had been fed a high fat diet (HFD) for 12 weeks were switched to a HFD containing (HFD-CTRL) or lacking cystine (HFD-CysF). Cth -/mice fed HFD-CysF diet were able to lose approximately 30% body weight within 1 week despite maintaining a high calorie intake (Fig.5g). This weight loss was associated with major reductions in fat mass (Fig.5h). With weight loss, cysteine deficient mice had improved glucose homeostasis, measured by a glucose tolerance test (Fig.5i), and increased EE and reduced RER (Fig.5j).
Histological analysis of the white adipose depots demonstrated that cysteine deficiency induced browning even while on HFD in subcutaneous and visceral adipocytes (Fig.5k). Additionally, consistent with improvement of metabolic function in obesity, the gene expression of inflammasome components Il1b, Il18, Nlrp3, casp1 and pro-inflammatory cytokines Il6 and Tnf were reduced in adipose tissue macrophages of the white adipose tissue (Fig. 5l) These results demonstrate that cysteine deficiency may be harnessed to lose weight and lower associated metabolic dysfunction during obesity.

Discussion
Adipose tissue regulates metabolism by orchestrating inter-organ communication required for healthy longevity. Analyses of adipose tissue of humans that underwent 15% moderate CR in has highlighted mechanisms that link energy metabolism and inflammation to influence healthspan 5,8 . In rodents reared in sub-thermoneutral temperature in most animal vivaria, restriction of calories up to 40% is known to reduce core-body temperature (CBT) and induce browning of the adipose tissue 1 . Uncoupling of WAT to produce heat from ATP during negative energy balance in rodents is an enigmatic phenomenon and its relevance to human physiology is unclear. Humans undergoing ~15% CR for two years in the CALERIE-II trial do not show reduction in CBT 5 . The transcriptomic analyses of abdominal subcutaneous adipose tissue of CALERIE-II participants revealed upregulation of lipolytic and fatty acid oxidation pathways but UCP1 or browning response was not detected 5 . Similarly, obese individuals upon weight-loss by CR do not show adipose tissue browning 42 . This suggest that rodent adipose tissue browning and thermogenesis in response to CR is related to extreme degree of restriction of calories (>40%) or other phenomenon including reduction of specific macronutrients. In this regard, pro-longevity interventions, caloric, protein restriction and methionine restriction induce adipose browning in mice and that is linked with induction of FGF21 43 . Our data demonstrates that induction of FGF21 by cysteine starvation partakes in increased energy expenditure. Detection of increased activity of transsulfuration pathway in adipose tissue of healthy humans undergoing CR and subsequent systematic analyses in genetic mouse models has revealed that protection of cysteine in mammals is an unexpected regulator of adipose-tissue thermogenesis.

Expression and activity of TSP genes CBS and CTH increase when cysteine is low 15 .
Indeed, during CR, the TSP is induced to defend against depletion of cysteine levels. Methionine restriction dietary regimens that improve lifespan are also restricted or deficient in cysteine [19][20][21] .
Thus, to understand the requirement of dietary non-essential amino acid such as cysteine to metabolism, a genetic mouse model is required that lacks Cth in conjunction with restriction of cysteine from the diet to effectively deplete this SAA from host. The previously reported Cth mutant mice generated on a 129SvEv mouse strain maintained on normal diet display hypertension and have motor-dysfunction characteristic of neurodegenerative changes in corpus striatum despite having cysteine in food [44][45][46] . The same strain of Cth deficient mice displayed normal body weight, food intake, body composition, and blood glucose levels on CR diet sufficient in cysteine 19 . Thus, to rigorously investigate tissue specific role of cysteine, we created new Cth conditional deletion (on pure C57/B6 background) in adipose tissue and liver and restricted cysteine in the diet. Surprisingly, we discovered that systemic cysteine starvation drives rapid uncontrolled mobilization and utilization of lipids in adipose tissues to increase heat production via adipose browning that results in loss of adipose TGs and weight-loss induced death without pathology. These results are consistent with the data that cysteine is a dietary nonessential amino acid, as our Cth deficient mice on normal diet were indistinguishable from control littermates.
The cysteine depletion induced browning in the white adipose depots is mediated through an increase in sympathetic nervous system derived NA bioavailability that drives increased lipolysis. While it is still unclear why cysteine deficiency triggers the activation of adipose browning, it appears that thermogenesis partially requires FGF21 and can be successfully maintained even in the absence of UCP1. Several non-canonical UCP-1 independent mechanisms have been proposed to control adipose beiging and thermogenesis 3,38 some of these mechanisms such a creatine cycle are also debated to be biologically not relevant for thermogenesis 39 . Despite significant induction of UCP1 upon cysteine deficiency, our data demonstrates that UCP1 is not required for thermogenesis and weight-loss. Future studies using this model of cysteine loss that produces lethal browning response can allow the discovery of alternate mechanism of adipose tissue thermogenesis.
In humans undergoing CR, a major redox regulator and cysteine substrate, glutathione was reduced in adipose tissue. Despite reduction in glutathione the CALERIE-II participants have improvement in markers of oxidative stress 9 and healthspan 5 . The Cth deficient mice on cysteine free diet did not show reduced GSH or organ dysfunction and displayed improved metabolic health and reduced inflammation. Cystine starvation in cancer cells induces an unexpected accumulation of γ-glutamyl-peptides, driven by elevated glutamate cysteine ligase catalytic subunit (GCLC) activity 47 . T h e γ -glutamyl-peptides that are similar to glutathione tripeptide (γ-glutamyl-2aminobutyrylglycine (g-Glu-2AB-Gly), can protect against ferroptosis 47 . Consistent with this, cysteine deficient mice that display adipose browning have elevated levels of γ-glutamyl-peptides like ophthalmic acid and 2-aminobutyric acid suggesting activation of alternative pathways that are engaged upon cysteine deficiency in the thermogenic adipose tissue.
Taken together, this study expands our understanding of pathways activated by prolongevity dietary interventions that confer metabolic adaptation required to maintain tissue homeostasis. Manipulation of TSP activity to drive adipose tissue browning also has implications for developing interventions that control adiposity and promote longevity. Here, based on human dietary-restriction studies, we identify that reduction of methionine cycle and defense of cysteine by induction of TSP is a critical determinant of organismal metabolism to prioritize energy utilization. Inability to maintain cysteine leads to uncontrolled adipose tissue thermogenesis. These findings, assign function to cysteine as an unexpected regulator of lipolysis and fatty acid oxidation that drives lipids into pathways that produce heat by uncoupling of white adipocytes.    Statistical differences were calculated by 2-way ANOVA with Sidak's correction for multiple comparisons, and by unpaired t-test *p<0.05, **p<0.01, ***p<0.001.

Diet studies
For cysteine deficiency studies, mice were fed either a control diet, CysF diet, HFD-CTRL diet, or HFD-CysF diet purchased from Dyets, for 6 days unless specified otherwise. For pair feeding studies, mice were provided with either ad libitum or 2.22-2.27g of diet daily.

Western blot analysis
Cell lysates were prepared using RIPA buffer and optionally frozen and stored at -80°C. Samples were left on ice, vortexing every ten min for 30 min. For tissue samples, snap frozen tissues were given by i.p. injection based on body weight (0.8U/kg). Blood glucose levels were measured by handheld glucometer (Breeze, Bayer Health Care).

Flow Cytometry
Adipose tissue was digested in HBSS (Life Technologies) + 0.1% collagenase I or II (Worthington Biochemicals). The stromal vascular fraction was collected by centrifugation, and washed and filtered using 100um and 70um strainers. Cells were stained with LIVE/DEAD™ Fixable Aqua Dead Cell Stain Kit (Thermo Fisher Scientific) and then for surface markers including CD45, CD3, B220, CD11b, F4/80, Ly6G, Siglec F, CD163, CD24, F3, CD31, Pdgfra, Dpp4, and CD9 and all antibodies were purchased from eBioscience or Biolegend. Cells were fixed in 2% PFA. Samples were acquired on a custom LSR II and data was analyzed in FlowJo.

Single-cell RNA sequencing
For stromal vascular fraction, female Cth +/+ and Cth -/mice were fed CTRL of CysF diet for 4 days. SFAT was collected, with lymph nodes removed, pooled, and digested. For myeloid cells, Cth +/+ and Cth -/mice were fed CTRL of CysF diet for 3 days. Intravascular labeling was performed by i.v. injection of 2.5ug CD45.2 diluted in PBS for 3 min. SFAT was collected, with lymph nodes removed, pooled, digested, and stained for live/dead, CD11b, CD45, and dump (CD3, B220, Siglec F). Cells were sorted on BD FACSAria, for CD45+ CD11b+ cells that were negative for CD45.2 and the dump channel. Isolated cells were subjected to droplet-based 3' end massively parallel single-cell RNA sequencing using Chromium Single Cell 3' Reagent Kits as per manufacturer's instructions (10x Genomics). The libraries were sequenced using a HiSeq3000 instrument (Illumina). Sample demultiplexing, barcode processing, and single-cell 3' counting was performed using the Cell Ranger Single-Cell Software Suite (10x Genomics).
Cellranger count was used to align samples to the reference genome (mm10), quantify reads, and filter reads with a quality score below 30. The Seurat package in R was used for subsequent analysis (Butler et al., 2018). Cells with mitochondrial content greater than 0.05 percent were removed and data was normalized using a scaling factor of 10,000, and nUMI was regressed with a negative binomial model. Principal component analysis was performed using the top 3000 most variable genes and t-SNE analysis was performed with the top 20 PCAs. Clustering was performed using a resolution of 0.4. The highly variable genes were selected using the FindVariableFeatures function with mean greater than 0.0125 or less then 3 and dispersion greater than 0.5. These genes are used in performing the linear dimensionality reduction.
Principal component analysis was performed prior to clustering and the first 20 PC's were used based on the ElbowPlot. Clustering was performed using the FindClusters function which works on K-nearest neighbor (KNN) graph model with the granularity ranging from 0.1-0.9 and selected 0.4 for the downstream clustering. For identifying the biomarkers for each cluster, we have performed differential expression between each cluster to all other clusters identifying positive markers for that cluster. To understand the trajectory of the adipocyte progenitors, we used Monocle2 to analyze scRNA-seq data of Clusters 0, 1, and 2 (Trapnell 2014).

Whole tissue RNA sequencing and transcriptome Analysis
Snap frozen tissues were ground by mortar and pestle in liquid nitrogen and resuspended in STAT-60. RNA was extracted using Zymo mini kits. RNA was sequenced on a HiSeq2500. The pathways (adjusted p-value < 0.05) was done using an in-house Python script in the following way. We considered all ordered pairs of pathways, where the first pathway had normalized enrichment score equal to or greater than the second pathway. For each ordered pair of pathways, we analyzed the leading gene sets of these pathways. The leading gene sets were obtained using fgsea [fgsea]. If at least one of the leading gene sets in a pair of pathways had more than 60% of genes in common with the other leading gene set, then we eliminated the second pathway in the pair.

Sample preparation for metabolome analysis
Frozen tissues or serum samples, together with internal standard compounds (mentioned below), was subjected to sonication in 500 μL of ice-cold methanol. To this, an equal volume of ultrapure water (LC/MS grade, Wako, Japan) and 0.4 volume of chloroform were added. The resulting suspension was centrifuged at 15,000 × g for 15 minutes at 4 °C. The aqueous phase was then filtered using an ultrafiltration tube (Ultrafree MC-PLHCC, Human Metabolome Technologies, Japan), and the filtrate was concentrated by nitrogen spraying (aluminum block bath with nitrogen gas spraying system, DTU-1BN/EN1-36, TAITEC, Japan). The concentrated filtrate was dissolved in 50 μL of ultrapure water and utilized for IC-MS and LC-MS/MS analysis. Methionine sulfone and 2-morpholinoethanesulfonic acid were employed as internal standards for cationic and anionic metabolites, respectively. The recovery rate (%) of the standards in each sample measurement was calculated to correct for the loss of endogenous metabolites during sample preparation.

IC-MS metabolome analysis
Anionic metabolites were detected using an orbitrap-type MS (Q-Exactive focus; Thermo Fisher Scientific, USA) connected to a high-performance ion-chromatography (IC) system (ICS-5000+, Thermo Fisher Scientific, USA) that allows for highly selective and sensitive metabolite quantification through IC separation and Fourier transfer MS principle. The IC system included a modified Thermo Scientific Dionex AERS 500 anion electrolytic suppressor, which converted (CO2). Mouse activity was detected by infrared sensors, and food intake and water consumption were measured via weight sensors on food and water dispensers located in the cage.

EchoMRI
The parameters of body composition were measured in vivo by magnetic resonance imaging (EchoMRI; Echo Medical Systems). The amount of fat mass, lean mass and free water were measured by the analysis. For the analysis, each mouse was placed in an acrylic tube with breathing holes and the tube was inserted in the MRI machine. The analysis per mouse takes approximately 90 sec and automatically calculated numerical results were analyzed.

Climate chambers
Mice were acclimated in climate chambers (model 7000-10, Caron) at either 30°C or 22°C, with humidity maintained at 50% under 12 h light/dark cycles. After one week acclimation, mice were switched to custom amino acid diets, either CTRL or CysF diet for 6 days, while maintained in the climate chambers. Mice were handled daily to measure body weight.

Histology
Tissues were collected in 10% formalin, embedded in paraffin and sectioned into 5um thick sections. Tissues were stained with hematoxylin and eosin (H&E) or stained for UCP1 (Abcam) and Goat anti-rabbit HRP (DAKO) and developed for color using Abcam DAB substrate kit.