《人体的社交网络——共生微生物》课外文献阅读：Cold-induced conversion of cholesterol to bile acids in mice shapes the gut microbiome and promotes adaptive thermogenesis

ARTICLES natre medicine Cold-induced conversion of cholesterol to bile acids in mice shapes the gut microbiome and promotes adaptive thermogenesis Anna Worthmann,, Clara John,, Malte C Ruhlemann20, Miriam Baguhl', Femke-Anouska Heinsen2o Nicola Schaltenbergl, Markus Heine, Christian Schlein, Ioannis Evangelakos', Chieko Mineo,, Markus Fischer 3 Maura Dandri, Claus Kremoser6, Ludger Scheja, Andre Franke 2o, Philip W Shaul& Joerg Heerenl( 2 Adaptive thermogenesis is an energy-demanding process that is mediated by cold-activated beige and brown adipocytes, and it g entails increased uptake of carbohydrates, as well as lipoprotein-derived triglycerides and cholesterol, into these thermogenic cells. ere we report that cold exposure in mice triggers a metabolic program that orchestrates lipoprotein processing in brown adipose tissue( BAT) and hepatic conversion of cholesterol to bile acids via the alternative synthesis pathway. This process is dependent on s hepatic induction of cytochrome P450, family 7, subfamily b, polypeptide 1(CYP7B1)and results in increased plasma levels,as G well as fecal excretion, of bile acids that is accompanied by distinct changes in gut microbiota and increased heat production 8 bile acid excretion, changed the bacterial composition of the gut and modulated thermogenic responses. These results identify bile acids as important metabolic effectors under conditions of sustained BAT activation and highlight the relevance of cholesterol 3 metabolism by the host for diet-induced changes of the gut microbiota and energy metabolism The gut microbiota contributes to energy homeostasis, and altera- form or after its conversion into bile acids. The integrative regulation ions in its composition are associated with cardiovascular and meta- of these processes is not fully understood but is of clinical relevance, a bolic diseases such as atherosclerosis, thrombosis, type 2 diabetes as an imbalance leads to increased plasma concentrations of athero- g certain cancers2-7. Microbial colonization of the gut begins directly Bile acids are exclusively synthesized in the liver by a number after birth and develops in response to genetic and environmental of enzymatic reactions using two different routes. The classical bile factors, especially the amount and composition of the diet. 9. Recent acid synthesis pathway starts with the rate-limiting enzyme choles- tudies in mice indicate that decreasing the housing temperature alters terol 7-a-hydroxylase(encoded by CYP7AI), and it prevails under the gut microbiota, which in turn enhances the thermogenic capacity normal conditions. Bile acids can also be formed by the alternative of adipose tissues and, hence, energy expenditure of the host o, The bile acid synthesis pathway, which is initiated by the action of sterol exposure of mammals to temperatures below their thermoneutral 27-hydroxylase(encoded by CYP27AI)followed by 25-hydroxycholes- N threshold (-30oC for mice and C for humans)is regarded as a terol7-a-hydroxylase(encoded by CYP7B1)8.Both synthesis routes cold stimulus, which activates BAT and promotes the appearance of principally generate the same bile acid species, which are subsequently brown-like beige adipocytes in white adipose tissue(WAT), increas- conjugated with glycine or taurine. The physiological relevance of the ing energy expenditure through non-shivering thermogenesis2-14. alternative pathway is not well understood, but it has been postulated Because this process needs ample amounts of energy, BaT activity is to be important for the metabolism of hydroxylated cholesterol that is a major determinant of plasma glucose and triglyceride levels 5, 6. derived from extrahepatic organs, such as the brain 8, 9. In addition, activated BAT protects from atherosclerosis by acceler- One important physiological role of bile acids after their biliary ating the apolipoprotein-E-dependent clearance of cholesterol-rich secretion in response to food ingestion is to mediate the emulsifica remnant particles by the liver. The resulting excess cholesterol in tion and absorption of dietary lipids20. In addition, bile acids function hepatocytes is recycled to the blood circulation as part of lipoproteins. as signaling molecules in various tissues. In enterocytes, they acti Alternatively, cholesterol is secreted into bile either in an unmodified vate the transcription factor farnesoid X receptor(FXR)to generate logy, Department of Pediatrics, University of Texas Southwestern Me nter, Dallas, Texas, USA -Institute of Food Chemistry, Germany. Department of Internal Medicine I, University Medical Center Hamburg-Eppendorf, Hamburg, Germany 5Phenex Pharmaceuticals AG, Heidelberg, Germany. These authors contributed equally to this work. Correspondence should be addressed to J H.(heeren @uke. de) Received 27 June 2016; accepted 17 May 2017: published online 12 June 2017; doi: 10.1038/nm. 4357 DNLINE PUBLICATION

© 2017 Nature America, Inc., part of Springer Nature. All rights reserved. a r t i c l e s nature medicine advance online publication  The gut microbiota contributes to energy homeostasis1, and altera￾tions in its composition are associated with cardiovascular and meta￾bolic diseases such as atherosclerosis, thrombosis, type 2 diabetes and non-alcoholic fatty liver disease, as well as the development of certain cancers2–7. Microbial colonization of the gut begins directly after birth and develops in response to genetic and environmental factors, especially the amount and composition of the diet8,9. Recent studies in mice indicate that decreasing the housing temperature alters the gut microbiota, which in turn enhances the thermogenic capacity of adipose tissues and, hence, energy expenditure of the host10,11. The exposure of mammals to temperatures below their thermoneutral threshold (~30 °C for mice and ~24 °C for humans) is regarded as a cold stimulus, which activates BAT and promotes the appearance of brown-like beige adipocytes in white adipose tissue (WAT), increas￾ing energy expenditure through non-shivering thermogenesis12–14. Because this process needs ample amounts of energy, BAT activity is a major determinant of plasma glucose and triglyceride levels15,16. In addition, activated BAT protects from atherosclerosis by acceler￾ating the apolipoprotein-E-dependent clearance of cholesterol-rich remnant particles by the liver17. The resulting excess cholesterol in hepatocytes is recycled to the blood circulation as part of lipoproteins. Alternatively, cholesterol is secreted into bile either in an unmodified form or after its conversion into bile acids. The integrative regulation of these processes is not fully understood but is of clinical relevance, as an imbalance leads to increased plasma concentrations of athero￾genic lipoproteins. Bile acids are exclusively synthesized in the liver by a number of enzymatic reactions using two different routes. The classical bile acid synthesis pathway starts with the rate-limiting enzyme choles￾terol 7-α-hydroxylase (encoded by CYP7A1), and it prevails under normal conditions18. Bile acids can also be formed by the alternative bile acid synthesis pathway, which is initiated by the action of sterol 27-hydroxylase (encoded by CYP27A1) followed by 25-hydroxycholes￾terol 7-α-hydroxylase (encoded by CYP7B1)18. Both synthesis routes principally generate the same bile acid species, which are subsequently conjugated with glycine or taurine. The physiological relevance of the alternative pathway is not well understood, but it has been postulated to be important for the metabolism of hydroxylated cholesterol that is derived from extrahepatic organs, such as the brain18,19. One important physiological role of bile acids after their biliary secretion in response to food ingestion is to mediate the emulsifica￾tion and absorption of dietary lipids20. In addition, bile acids function as signaling molecules in various tissues. In enterocytes, they acti￾vate the transcription factor farnesoid X receptor (FXR) to generate 1Department of Biochemistry and Molecular Cell Biology, University Medical Center Hamburg-Eppendorf, Hamburg, Germany. 2Institute of Clinical Molecular Biology, Christian-Albrechts-University Kiel, Kiel, Germany. 3Center for Pulmonary and Vascular Biology, Department of Pediatrics, University of Texas Southwestern Medical Center, Dallas, Texas, USA. 4Institute of Food Chemistry, University of Hamburg, Hamburg, Germany. 5Department of Internal Medicine I, University Medical Center Hamburg-Eppendorf, Hamburg, Germany. 6Phenex Pharmaceuticals AG, Heidelberg, Germany. 7These authors contributed equally to this work. Correspondence should be addressed to J.H. (heeren@uke.de). Received 27 June 2016; accepted 17 May 2017; published online 12 June 2017; doi:10.1038/nm.4357 Cold-induced conversion of cholesterol to bile acids in mice shapes the gut microbiome and promotes adaptive thermogenesis Anna Worthmann1,7, Clara John1,7, Malte C Rühlemann2 , Miriam Baguhl1, Femke-Anouska Heinsen2 , Nicola Schaltenberg1, Markus Heine1, Christian Schlein1, Ioannis Evangelakos1, Chieko Mineo3, Markus Fischer4, Maura Dandri5, Claus Kremoser6, Ludger Scheja1, Andre Franke2 , Philip W Shaul3 & Joerg Heeren1 Adaptive thermogenesis is an energy-demanding process that is mediated by cold-activated beige and brown adipocytes, and it entails increased uptake of carbohydrates, as well as lipoprotein-derived triglycerides and cholesterol, into these thermogenic cells. Here we report that cold exposure in mice triggers a metabolic program that orchestrates lipoprotein processing in brown adipose tissue (BAT) and hepatic conversion of cholesterol to bile acids via the alternative synthesis pathway. This process is dependent on hepatic induction of cytochrome P450, family 7, subfamily b, polypeptide 1 (CYP7B1) and results in increased plasma levels, as well as fecal excretion, of bile acids that is accompanied by distinct changes in gut microbiota and increased heat production. Genetic and pharmacological interventions that targeted the synthesis and biliary excretion of bile acids prevented the rise in fecal bile acid excretion, changed the bacterial composition of the gut and modulated thermogenic responses. These results identify bile acids as important metabolic effectors under conditions of sustained BAT activation and highlight the relevance of cholesterol metabolism by the host for diet-induced changes of the gut microbiota and energy metabolism

ARTICLES a negative endocrine feedback signal to th ia fibroblast growth warm-housed mice(Fig.). On the genus level, the abundance of factor(fgf)9inh Parabacteroides spp. was higher, whereas the abundance of undlassi cold- ances at the versus differ- cold in an ghe the 16S rR ent with studie exposurel0.. We examine that th Porphyrom uction

© 2017 Nature America, Inc., part of Springer Nature. All rights reserved. Ar t i c l e s  advance online publication nature medicine a negative endocrine feedback signal to the liver via fibroblast growth factor (FGF) 19 in humans or via the mouse ortholog FGF15 (ref. 21). The mechanism involves the activation of the transcriptional co-repressor small heterodimeric partner (SHP; encoded by Nr0b2) in hepatocytes, which confers suppression of CYP7A1 expression and hence lowers bile acid synthesis. Both in liver and adipose tissue, bile￾acid-activated FXR has additional effects, including such as those on cholesterol, fatty acid and glucose metabolism20,21. Bile acids also stimulate signaling at the plasma membrane via the G-protein-cou￾pled bile acid receptor 1 (GPBAR1; also known as TGR5), thereby regulating diverse metabolic pathways22. In the context of energy metabolism, TGR5 stimulation triggers the secretion of the insulin secretagogue and anorexic hormone glucagon (encoded by Gcg) in L-cells of the intestine23. Moreover, in both mice and humans, bile acids increase energy expenditure via the stimulation of TGR5 on brown adipocytes24,25, which suggests a role of liver-derived bile acids as regulators of adaptive thermogenesis. Bile acids show mutual interactions with gut microbiota. On the one hand, the gut bacteria modify bile acids through both dehydroxy￾lation, to form secondary bile acids such as deoxycholic acid (DCA), and through deconjungation via the enzymatic activity of bile salt hydrolases26. On the other hand, bile acids exert species-dependent bacteriostatic effects, thereby affecting the composition of the intes￾tinal microbiota27. As previously mentioned, cold exposure has been found to induce alterations in the gut microbiota of mice10,11, and transplantation of feces from cold-exposed mice was found to improve the metabolic profile of recipient mice that were housed at ambient temperature10. Thus, these studies provided evidence that cold exposure has distinct effects on the bacterial composition of the intestine and can promote a healthy metabolic phenotype. However, how cold conditions change the composition of the gut microbiota remains elusive. Here we demonstrate that during BAT activation by cold expo￾sure, bile-acid-driven alterations in systemic cholesterol metabolism integrate the organ-specific responses in thermogenic adipose tis￾sues, liver and intestine. Cholesterol homeostasis was achieved by the stimulation of lipoprotein flux, increased bile acid synthesis via the alternative pathway and massive fecal excretion of conjugated bile acids, which ultimately shaped the gut microbiome and promoted adaptive thermogenesis in a cold environment. RESULTS Cold alters the gut microbiome Dietary composition is a determinant of the gut microbiome28,29, and cold exposure requires higher food intake to meet the increased energy demands for thermogenesis12. Here we combined dietary and cold intervention by challenging mice with a cholesterol-enriched high-fat diet (HFD) and keeping the mice under thermoneutral (30 °C; hereafter referred to as warm) versus cold (6 °C) conditions, respectively. First, we profiled the fecal microbiota from warm-housed and cold-housed mice by analyzing the DNA sequence encoding the 16S rRNA. Multidimensional scaling (MDS) based on weighted UniFrac distances revealed distinct clustering between microbiota from warm-housed versus cold-housed mice (Fig. 1a), which is consist￾ent with studies using mice that were subjected to prolonged cold exposure10,11. We examined the changes at the family level and found that the abundance of Lachnospiraceae and Deferribacteraceae fam￾ily members was higher, whereas the abundance of Clostridiales and Porphyromonadaceae family members was lower, in cold- versus warm-housed mice (Fig. 1b). On the genus level, the abundance of Parabacteroides spp. was higher, whereas the abundance of unclassi￾fied Porphyromonadaceae family members was clearly lower in cold￾housed versus warm-housed mice (Fig. 1b). These differences could be traced to operational taxonomic unit (OTU) abundances at the species level, which showed distinct patterns in warm-housed versus cold-housed mice (Fig. 1c). Although we could not observe differ￾ences in α-diversity at the genus level (Fig. 1d), we did note that cold exposure resulted in lower richness and a decreased Shannon index at the species level (Fig. 1e). We detected similar cold-dependent differences in chow-fed wild-type (WT) mice (Supplementary Fig. 1a–e) and in HFD-fed diabetic db/db mice (Supplementary Fig. 2a–e). Thus, these data clearly indicated that cold-housing resulted in an altered gut microbiome in mice that were fed either a chow diet or a cholesterol-enriched HFD. Cold induces the alternative bile acid synthesis pathway In accordance with increased energy demands, we observed higher food intake and fecal mass in cold-exposed mice than in mice that were housed at thermoneutrality (Fig. 1f and Supplementary Table 1a). Despite higher dietary lipid uptake, the cold-exposed mice had lower plasma levels of triglycerides and total cholesterol (Fig. 1g), owing to reduced levels of triglyceride-rich lipoproteins (Fig. 1h) and choles￾terol-transporting low-density lipoprotein (LDL) and high-density lipoprotein (HDL) (Fig. 1i). Next we administered radiolabeled cho￾lesterol by oral gavage and observed higher cholesterol uptake in all BAT depots in cold- versus warm-housed mice, despite the slightly lower levels of radioactivity in plasma (Fig. 1j). Notably, cholesterol uptake was also higher in the liver, which took up more than 10% of the administered dose in cold-exposed mice (Fig. 1k). Consistent with the known inefficient absorption of dietary cholesterol, the main proportion of radioactive tracer was still present in the small intes￾tine, irrespective of cold exposure (Fig. 1k). Taken together, these data demonstrate both accelerated lipoprotein processing in BAT and increased cholesterol transport from thermogenic adipose tissues toward the liver, which is consistent with previous studies17. To determine the fate of cholesterol, we measured bile acid levels in the liver by quantitative liquid chromatography coupled to mass spec￾trometry (LC–MS). We observed that the levels of most of the unconju￾gated bile acid (UBA; Fig. 2a) and conjugated bile acid (CBA; Fig. 2b) species were significantly higher in the cold-housed mice than in the warm-housed control mice and noted that the most prominent induc￾tions were for cholic acid (CA), ursodeoxycholic acid (UDCA) and muricholic acids (MCAs). However, the composition of the bile acid pool was not significantly affected (Supplementary Fig. 3a). In gall bladder, higher levels of CBA were observed in cold- versus warm￾housed mice (Fig. 2c), suggesting that there was efficient biliary CBA excretion via the bile salt export pump (BSEP; encoded by Abcb11). Consistent with normal liver function, cold exposure did not provoke liver inflammation or damage (Supplementary Fig. 3b,c). To determine the basis for cold-induced hepatic bile acid synthesis, we performed mRNA analysis of liver tissue to evaluate the expression of components of both the classical and alternative bile acid syn￾thesis pathways (Fig. 2d,e). Whereas Cyp7a1 expression was unal￾tered, Cyp7b1 expression was fourfold higher in cold-housed mice than in warm-housed mice (Fig. 2e), indicating specific upregulation of the alternative bile acid synthesis pathway. To unravel the poten￾tial role of FXR in cold-induced bile acid synthesis, we treated mice with the FXR agonist Phenex20606 (hereafter referred to as PX)30. As expected, treatment with PX resulted in significant induction

ARTICLES ●cod oTU 32 Palndibacter oTU 897 oTU 3 orphyromonadaceae aludibacter oTU 998 Osciibacter OTU 15 Deferribacteraceae oTu 7 Porphyromonadaceae Parabacteroides OT Ds1(52%) Clostridium xIva OTU 68 小导4杂P98 Genera Mean relative abundance (% Mean relative abundance(%) t Clostridiales ■Unc. Porphyromonas■Uc Anaerotruncus Uncl. Lachnospiraceae UncL. ■Unc. Erysipelotrichac.■uncL zEz9 Cold g8E8型 Food intake Feces Fracto k Figure 1 BAT activation alters the gut microbiome, lipoprotein levels and cholesterol uptake (a-e)MDS plot of weighted UniFrac distance(a), mean relative sundance of gut microbiota on family and genus level( b), hierarchical clustering of significant altered oTUs (c), as well as alpha diversity presented at the (w), n=9 mice)or 6C(cold (c), n=5 mice). (f)Food intake and amount of feces per 24 h period (n= 6 mice per group). (g) Plasma triglyceride (g)and cholesterol(Chol)levels(n= 5 mice per group).(h, i) Profiles of triglyceride-rich lipoproteins(TRLs)and cholesterol-rich LDL and hdl by measurement of triglycerides(h)and cholesterol (i)in FPLC fractions, respectively (pooled plasma samples of n=5 mice per group).( k)Organ uptake of 3H-radiolabeled P<0.05, ""P<0.01.*P<0.001: n.s., not significant; by pairwise Wilcoxon rank-sum test(d, e)or unpaired two-tailed Student's -test (f,gk-mgWAT holesterol 4 h after oral gavage, expressed as fold induction (i or as percentage of injected dose for liver and small intestine ( k)(n=7 mice per group). ing guinal WAT; epiwAT, epididymal WAT; ScBAT, subscapular BAT; iBAT, interscapular BAT: deBAT, deep cervical BAT. Throughout, data are mean +se

© 2017 Nature America, Inc., part of Springer Nature. All rights reserved. a r t i c l e s nature medicine advance online publication  MDS2 (20%) MDS1 (52%) Warm Cold Warm Warm *** *** *** *** ** *** *** *** * * Food intake Feces Cold Cold Warm Cold Warm Cold Warm Cold TRL Warm HDL LDL Cold Warm Cold Warm Cold Warm Cold Warm Cold Warm Warm Cold Warm Families Mean relative abundance (%) Porphyromonadaceae Mean relative abundance (%) Parabacteroides Uncl. Porphyromonad. Uncl. Ruminococc. Mucispirillum Anaerotruncus Other Uncl. Marinilabiliac. Uncl. Clostridiales Uncl. Synthrophom. Uncl. Lachnospiraceae Uncl. Erysipelotrichac. Genera Cold Cold log10 scaled abundance Family Lachnospiraceae Lachnospiraceae Porphyromonadaceae Porphyromonadaceae Porphyromonadaceae Ruminococcaceae Ruminococcaceae Erysipelotrichaceae Erysipelotrichaceae Other 30 2.0 160 3 2 1 0 120 80 40 0 1.5 1.0 0.5 0.0 20 10 0 5 200 3 10 8 6 4 2 0 2 1 0 0 10 20 Fraction 30 0 10 20 Fraction 30 150 100 TG Chol 50 0 4 3 2 1 0 5 4 3 2 1 0 n.s. n.s. ** ** Observed genera Amount (g/24 h) Organ uptake (cpm/organ, fold) Organ uptake (% of given dose) Concentration (mg/dl) TG (mg/dl) Chol (mg/dl) Shannon (genera) Observed OTUs Shannon (OTU) Marinilabiliaceae Clostridiales Synthrophomonadaceae Deferribacteraceae Deferribacteraceae Porphyromonadaceae Lachnospiraceae w6 Plasma Liver Spleen Kidney Muscle epiWAT ingWAT scBAT iBAT dcBAT Liver Duodenum Jejunum Ileum Intestine total w1w2 w9 w3 w7 w5w4 w8c2 c3 c1 c4c5 Genus OTU OTU_32 OTU_897 OTU_3 OTU_998 OTU_15 OTU_23 OTU_7 OTU_1 OTU_68 incertae sedis Paludibacter Paludibacter Paludibacter Oscillibacter Allobaculum Mucispirillum Parabacteroides Clostridium XIVa –2 –1 0 1 2 0.10 0.1 0.2 –0.10 –0.2 –0.1 0.05 –0.05 0 0 a b d f j k g h i c Figure 1 BAT activation alters the gut microbiome, lipoprotein levels and cholesterol uptake. (a–e) MDS plot of weighted UniFrac distance (a), mean relative abundance of gut microbiota on family and genus level (b), hierarchical clustering of significant altered OTUs (c), as well as alpha diversity presented at the genus level (d) or at the OTU level (e), on the basis of 16S-rRNA-encoding sequences in feces collected from mice that were housed at thermoneutrality (warm (w), n = 9 mice) or 6 °C (cold (c), n = 5 mice). (f) Food intake and amount of feces per 24 h period (n = 6 mice per group). (g) Plasma triglyceride (TG) and cholesterol (Chol) levels (n = 5 mice per group). (h,i) Profiles of triglyceride-rich lipoproteins (TRLs) and cholesterol-rich LDL and HDL by measurement of triglycerides (h) and cholesterol (i) in FPLC fractions, respectively (pooled plasma samples of n = 5 mice per group). (j,k) Organ uptake of 3H-radiolabeled cholesterol 4 h after oral gavage, expressed as fold induction (j) or as percentage of injected dose for liver and small intestine (k) (n = 7 mice per group). ingWAT, inguinal WAT; epiWAT, epididymal WAT; scBAT, subscapular BAT; iBAT, interscapular BAT; dcBAT, deep cervical BAT. Throughout, data are mean ± s.e.m. *P < 0.05, **P < 0.01, ***P < 0.001; n.s., not significant; by pairwise Wilcoxon rank-sum test (d,e) or unpaired two-tailed Student’s t-test (f,g,j,k)

ARTICLES O warm●ca owam●cad O Warm 20 001● e 冒 g :: O wT. ● WT cold ● WT cold o cyp7br--cold 0.0 0-a Figure 2 BAT activation induces the alternative bile acid synthesis pathway independently of FXR. (a, b)Relative levels of unconjugated bile acid (UBA) species(a)and conjugated bile acid(CBA)species()in the livers of mice that were housed at thermoneutrality (warm, n=9 mice)or 6C (cold, n=5 nice).(c)Total UBA and CBa levels in gall bladders of warm- housed(n=8)and cold- housed(n= 10)mice. (d) Schematic diagram of the classical and ternative bile acid synthesis pathways. (e, f)Expression of genes involved in cholesterol and bile acid transport and metabolism(e)and hepatic UBA and CBA levels(f)in warm-housed and cold-housed mice that were treated with the fXr agonist PX20606 (PX)or with vehicle for 3 d (warm-housed: vehicle (warm), n=6 mice; PX, n=6 mice; cold-housed: vehicle (cold), n=5 mice; PX, n=6 mice). (g, h)Hepatic mRNA expression of genes involved in bile acid etabolism(g), as well as UBA (left)and CBa (right)levels(h), in mice that were housed at 22C and treated without(mock)or with CL316, 243(CL) for 7 d(n=8 mice per group). (i)Bile acid levels in livers of mice that were housed at thermoneutrality and treated with either AAv-GFP or AAV-Cyp7bl (n=7 mice per group). () Hepatic levels of bile acids in warm- housed and cold-housed WT and Cyp7b1--mice(n= 3 mice per group).(k, l)Expression of Cyp27al in BAT (n= 5 mice per group)(k) from, and plasma levels of 27-hydroxycholesterol (n= 4 mice per group)()in, warm-housed and cold housed mice (m)Levels of 27-hydroxycholesterol in the feces of warm housed and cold-housed WT (n=5 mice per group) and Cyp7b1--(n=4 mice per group)mice. CDCA, chenodeoxycholic acid; CA, cholic acid; DCA, deoxycholic acid; GCDCA, glycochenodeoxycholic acid; GCA, glycocholic acid; GDCA, glycodeoxycholic acid; HDCA, hyodeoxycholic acid; TCDCA, taurochenodeoxycholic acid; TCA, taurocholic acid; TDCA, taurodeoxycholic acid; THDCA urohyodeoxycholic acid; TLCA, taurolithocholic acid; TUDCA, tauroursodeoxycholic acid; T-a-MCA, tauro-a-muricholic acid; T-B-MCA, tauro-B-muricholic id: UDCA, ursodeoxycholic acid; a-MCA, a-muricholic acid; B-MCA, B-muricholic acid; o-MCA, o-muricholic acid. Throughout, data are mean ts.e. m. P 0.05, P<0.01, P<0.001; by unpaired two-tailed Students t-test (a-c, g-i, k, I or two-way analysis of variance(ANOvA)(e, f, j, m)

© 2017 Nature America, Inc., part of Springer Nature. All rights reserved. Ar t i c l e s  advance online publication nature medicine Warm CA β-MCA UDCA α-MCA HDCA T-α/βMCA THDCA TUDCA TCDCA/TDCA TLCA TCA GCDCA GDCA GCA UBA CBA ω-MCA 60 8 200 150 100 50 0 6 4 2 0 40 20 0 Bile acids (ng/mg, fold) Bile acids (ng/mg, fold) Bile acids (mM) *** *** *** ** ** ** ** * * a Cold b Warm Cold c Warm Cold Cyp7a1 Cyp8a1 Cyp7b1 Cyp27a1 Baat Abcb11 Nr0b2 Cholesterol 5-Cholesten-3-β, 7α-diol 5-Cholesten-7-α-ol-one 7-Hydroxycholesterol 27�-Hydroxycholesterol Unconjugated bile acid Conjugated bile acids Baat Cholic acid �-Muricholic acid Classical pathway Alternative pathway Cyp27a1 Cyp7b1 Cyp8b1 Cyp7a1 Gene expression (fold) Bile acids (ng/mg) *** *** *** ** ** ** ** ** ** ** ** ** * * * * * * * * * * Warm 20 15 200 150 100 50 0 10 5 0 15 10 5 0 Warm PX Cold PX UBA CBA Cold d e f Cyp7a1 Cyp8b1 Cyp7b1 Cyp27a1 Baat Abcb11 Nr0b2 Gene expression (fold) Bile acids (ng/mg) Bile acids (ng/mg) * * 200 150 100 50 0 200 150 100 50 0 10 8 6 4 2 0 20 15 10 5 0 4 3 2 1 0 Mock CL Mock CL AAV-GFP AAV-Cyp7b1 UBA CBA UBA CBA g h i Gene expression (fold) Concentration (nmol/mg cholestrol) 27-OH-Chol (ng/mg) Bile acids (ng/mg) ** ** ** * * * * 15 500 WT warm Cyp7b1−/− warm WT cold Cyp7b1−/− cold 400 300 200 100 10 5 0 0 UBA CBA Warm Cold Cyp27a1 27-OH-Chol Feces WT warm WT cold Cyp7b1−/− warm Cyp7b1−/− cold 2.0 0.3 80 60 40 20 0 0.2 0.1 0.0 1.5 1.0 0.5 0.0 Warm Cold j k l m Figure 2 BAT activation induces the alternative bile acid synthesis pathway independently of FXR. (a,b) Relative levels of unconjugated bile acid (UBA) species (a) and conjugated bile acid (CBA) species (b) in the livers of mice that were housed at thermoneutrality (warm, n = 9 mice) or 6 °C (cold, n = 5 mice). (c) Total UBA and CBA levels in gall bladders of warm-housed (n = 8) and cold-housed (n = 10) mice. (d) Schematic diagram of the classical and alternative bile acid synthesis pathways. (e,f) Expression of genes involved in cholesterol and bile acid transport and metabolism (e) and hepatic UBA and CBA levels (f) in warm-housed and cold-housed mice that were treated with the FXR agonist PX20606 (PX) or with vehicle for 3 d (warm-housed: vehicle (warm), n = 6 mice; PX, n = 6 mice; cold-housed: vehicle (cold), n = 5 mice; PX, n = 6 mice). (g,h) Hepatic mRNA expression of genes involved in bile acid metabolism (g), as well as UBA (left) and CBA (right) levels (h), in mice that were housed at 22 °C and treated without (mock) or with CL316,243 (CL) for 7 d (n = 8 mice per group). (i) Bile acid levels in livers of mice that were housed at thermoneutrality and treated with either AAV-GFP or AAV-Cyp7b1 (n = 7 mice per group). (j) Hepatic levels of bile acids in warm-housed and cold-housed WT and Cyp7b1−/− mice (n = 3 mice per group). (k,l) Expression of Cyp27a1 in BAT (n = 5 mice per group) (k) from, and plasma levels of 27-hydroxycholesterol (n = 4 mice per group) (l) in, warm-housed and cold￾housed mice. (m) Levels of 27-hydroxycholesterol in the feces of warm-housed and cold-housed WT (n = 5 mice per group) and Cyp7b1−/− (n = 4 mice per group) mice. CDCA, chenodeoxycholic acid; CA, cholic acid; DCA, deoxycholic acid; GCDCA, glycochenodeoxycholic acid; GCA, glycocholic acid; GDCA, glycodeoxycholic acid; HDCA, hyodeoxycholic acid; TCDCA, taurochenodeoxycholic acid; TCA, taurocholic acid; TDCA, taurodeoxycholic acid; THDCA, taurohyodeoxycholic acid; TLCA, taurolithocholic acid; TUDCA, tauroursodeoxycholic acid; T-α-MCA, tauro-α-muricholic acid; T-β-MCA, tauro-β-muricholic acid; UDCA, ursodeoxycholic acid; α-MCA, α-muricholic acid; β-MCA, β-muricholic acid; ω-MCA, ω-muricholic acid. Throughout, data are mean ± s.e.m. *P < 0.05, **P < 0.01, ***P < 0.001; by unpaired two-tailed Student’s t-test (a–c,g–i,k,l) or two-way analysis of variance (ANOVA) (e,f,j,m)

ARTICLES of AbcblI expression, as well as that of Nrob2, and downregulation transporter(ASBT), which is required for efficient CBA re-uptake as of the SHP target genes Cyp 7al and Cyp8b1 in warm-housed, as well part of the enterohepatic circulation- Cold exposure led to higher as cold-housed, mice(Fig. 2e), whereas other genes important for ASBT expression at the mRNA (SIc10a2; Fig 3f), as well as protein synthesis( Cyp27a1)and conjugation(Baat)were unaffected by PX (Fig. 3g), level, which was lost in Cyp7b1-/-mice(Supplementary treatment but were slightly higher after cold exposure. Notably, the Fig. 7c, d). Furthermore, cold-housing resulted in a trend toward upregulation of Cyp bl that we observed after cold treatment was higher bile acid levels in portal(Fig. 3h) and systemic(Fig. 3i)blood. not affected by a FXR-mediated negative feedback loop( Fig. 2e). Taken together, these data argue against diminished CBa re-uptake Consistent with sustained Cyp7b1 expression, PX treatment did not capacity as a cause for higher fecal bile acid levels influence the cold-induced increase in hepatic bile acid levels(Fig Another mechanism for increased fecal concentrations of cBa and only species that arose by the action of the PX-sensitive enzyme could be due to reduced bile acid deconjugation by bile salt hydrolases CYP8B1, such as CA, were slightly lower(Supplementary Fig 4a-c).(BSHs), which are expressed by a number of gut bacteria 26. To address In a human cohort, we found lower hepatic expression of CYP7BI this hypothesis we depleted the intestinal microbiota by using an ant in obese subjects with type 2 diabetes than in non-obese controls, biotic cocktail. Relative to that in the untreated controls, this interven whereas expression of CYP7AI and CYP8BI was not different tion resulted in higher fecal CBA content both in warm-housed and (Supplementary Fig. 5a-c), indicating that there was metabolic cold-exposed mice(Fig. 3j, k), which could be explained by the com- 9 regulation of the alternative pathway in humans. plete eradication of gut BSH activity by antibiotic treatment(Fig. 31) To determine the role of BAT in cold-induced bile acid synthesis, we Although antibiotic treatment did not affect the expression of genes g used the B3-adrenergic receptor agonist CL316, 243 to pharmacologi- related to bile acid metabolism(Supplementary Fig. 8a), hepatic e cally activate BAT in mice that were housed at room temperature. As UBA but not CBA levels were lower(Supplementary Fig 8b, c) w compared to the untreated controls, expression of Cyp7bl, but of no Of note, higher BSH activity was observed in cold-exposed mice other genes involved in hepatic bile acid synthesis, was higher in the in the absence of antibiotics than in warm-housed mice(Fig. 31), CL316, 243-treated mice(Fig. 2g)and was associated with increased resulting in higher fecal taurine levels(Fig. 3m). Cumulatively, these levels of hepatic bile acids( Fig. 2h and Supplementary Fig. 6a, b). results suggest that saturation of ASBT rather than diminished decon Genetic intervention by adeno-associated virus(AAV)-mediated jugation by BSH activity is responsible for higher fecal CBAexcretion E overexpression of Cyp 7bl in WT mice resulted in moderately higher after cold exposure z hepatic bile acid levels at thermoneutral conditions, as compared to those in mice that were treated with GFP-expressing control AAv Bile acid excretion depends on a BAT-liver cholesterol axis E(AAV-GFP)(Fig. 2i and Supplementary Fig. 6c). Conversely, the Cold-activated BAT efficiently processes dietary lipids carried by p a cold-dependent induction in hepatic bile acid levels was blunted prandial lipoproteins and promotes receptor-mediated uptake of the in Cyp7b1-- mice, as compared to that in WT mice(Fig. 2j and respective cholesterol-rich remnants by the liver 5, 17. Accordingly, in induced the a CyP27aI in BAT(Fig. 2k)and led to increased amounts of plasma cold(Fig.1g-i), we observed higher amounts of circulating cholesterol 27-hydroxycholesterol(Fig. 21), the substrate for hepatic CYP7B1. rich lipoproteins in cold-housed versus warm-housed mice that were Of note, 27-hydroxycholesterol did not accumulate in the liver and deficient in the LDL receptor (LDLr), which is the main receptor for g plasma of cold-exposed Cyp7b1-/-mice(Supplementary Fig. 6f, g), hepatic remnants(Fig 4a, b). When we combined LdIr-genotype with which could be because much of it ed in the feces( Fig. 21 knockout of the the alternative In conclusion, these data reveal that under conditions of BAT activa- lipoprotein receptor, LDLR-related protein 1 (LrpI), we observed on, the alternative pathway is selectively upregulated and increases even higher lipoprotein levels(Fig. 4c, d), indicating blunted remnant bile acid sy clearance in the absence of both receptors. Notably, the cold-induced increase in fecal bile acid excretion in LdIr-mice was only 50% of that old accelerates fecal bile acid excretion via cyP7b1 in WT mice, and it was nearly abolished in mice that lacked LDLR and To further explore the fate of bile acids, we quantified their levels hepatic LRPI (Fig. 4e-g). Taken together, these findings indicate that N in stool samples from warm-housed and cold-housed mice. The fecal bile acid excretion is dependent on hepatic cholesterol that is deliv- amount of excreted bile acids was much higher in cold-housed ered by postprandial lipoproteins generated by cold-activated BAT than in warm-housed mice(Fig 3a, b). Notably, some CBA sp To assess the contribution of dietary cholesterol to cold-induced cies, especially tauro-a/B-MCA (T-a/B-MCA) and tauro-ca bile acid synthesis, we first blocked dietary cholesterol resorption by TCA), were up to 40-fold higher UBA levels were less affected and inhibiting the intestinal cholesterol transporter Niemann-Pick-1-like were increased up to twofold(Fig. 3a, b ) In Cyp7b1-/- mice, the 1(NPCIL1)with ezetimibe(EZ). Notably, combining cold-exposure cold-induced rise in fecal CBA was abrogated, and concentrations and EZ treatment resulted in lower levels of plasma lipids and choles- of some bile acid species were even lower(Fig. 3c, d). Conversely, terol-rich lipoproteins, even more so than with cold exposure alone AAV-mediated Cyp7b1 overexpression resulted in higher fecal bile (Fig 5a, b). Consistent with diminished dietary cholesterol uptak acid species, as compared to that in the AAv-GFP controls(Fig. 3e and hepatic lipid levels after EZ treatment(Fig. 5c), we observed and Supplementary Fig. 7a, b). Notably, the effect was not present compensatory hepatic upregulation of the gene encoding the rate- at thermoneutral conditions but at 22C, and it was even more pro- limiting enzyme of the cholesterol synthesis pathway(Hmgcr),as bounced at 16C(Fig. 3e and Supplementary Fig. 7a, b), indicat- well as of Ldlr(Fig. 5d), whereas expression of Cyp7b1 was unaf ing that increased dietary intake and BAT-dependent processing of fected( Supplementary Fig 9a). Liver bile acid levels of cold-housed cholesterol at lower temperatures is critical for efficient CYP7B1- EZ-treated mice were comparable to those of cold-housed control mediated bile acid production mice(Fig. 5e and Supplementary Fig 9b, c). In contrast, the fecal mice, wea sate the basis of higher fecal CBA levels in cold-housed bile acid concentrations of EZ-treated mice were not higher after rst quantified the ileal expression of apical sodium bile cold exposure, and bile acid levels were almost identical to those in

© 2017 Nature America, Inc., part of Springer Nature. All rights reserved. a r t i c l e s nature medicine advance online publication  of Abcb11 expression, as well as that of Nr0b2, and downregulation of the SHP target genes Cyp7a1 and Cyp8b1 in warm-housed, as well as cold-housed, mice (Fig. 2e), whereas other genes important for synthesis (Cyp27a1) and conjugation (Baat) were unaffected by PX treatment but were slightly higher after cold exposure. Notably, the upregulation of Cyp7b1 that we observed after cold treatment was not affected by a FXR-mediated negative feedback loop (Fig. 2e). Consistent with sustained Cyp7b1 expression, PX treatment did not influence the cold-induced increase in hepatic bile acid levels (Fig. 2f), and only species that arose by the action of the PX-sensitive enzyme CYP8B1, such as CA, were slightly lower (Supplementary Fig. 4a–c). In a human cohort, we found lower hepatic expression of CYP7B1 in obese subjects with type 2 diabetes than in non-obese controls, whereas expression of CYP7A1 and CYP8B1 was not different (Supplementary Fig. 5a–c), indicating that there was metabolic regulation of the alternative pathway in humans. To determine the role of BAT in cold-induced bile acid synthesis, we used the β3-adrenergic receptor agonist CL316,243 to pharmacologi￾cally activate BAT in mice that were housed at room temperature. As compared to the untreated controls, expression of Cyp7b1, but of no other genes involved in hepatic bile acid synthesis, was higher in the CL316,243-treated mice (Fig. 2g) and was associated with increased levels of hepatic bile acids (Fig. 2h and Supplementary Fig. 6a,b). Genetic intervention by adeno-associated virus (AAV)-mediated overexpression of Cyp7b1 in WT mice resulted in moderately higher hepatic bile acid levels at thermoneutral conditions, as compared to those in mice that were treated with GFP-expressing control AAV (AAV-GFP) (Fig. 2i and Supplementary Fig. 6c). Conversely, the cold-dependent induction in hepatic bile acid levels was blunted in Cyp7b1−/− mice, as compared to that in WT mice (Fig. 2j and Supplementary Fig. 6d,e). Cold exposure induced the expression of Cyp27a1 in BAT (Fig. 2k) and led to increased amounts of plasma 27-hydroxycholesterol (Fig. 2l), the substrate for hepatic CYP7B1. Of note, 27-hydroxycholesterol did not accumulate in the liver and plasma of cold-exposed Cyp7b1−/− mice (Supplementary Fig. 6f,g), which could be because much of it was excreted in the feces (Fig. 2m). In conclusion, these data reveal that under conditions of BAT activa￾tion, the alternative pathway is selectively upregulated and increases bile acid synthesis. Cold accelerates fecal bile acid excretion via CYP7B1 To further explore the fate of bile acids, we quantified their levels in stool samples from warm-housed and cold-housed mice. The amount of excreted bile acids was much higher in cold-housed than in warm-housed mice (Fig. 3a,b). Notably, some CBA spe￾cies, especially tauro-α/β-MCA (T-α/β-MCA) and tauro-CA (TCA), were up to 40-fold higher. UBA levels were less affected and were increased up to twofold (Fig. 3a,b). In Cyp7b1−/− mice, the cold-induced rise in fecal CBA was abrogated, and concentrations of some bile acid species were even lower (Fig. 3c,d). Conversely, AAV-mediated Cyp7b1 overexpression resulted in higher fecal bile acid species, as compared to that in the AAV-GFP controls (Fig. 3e and Supplementary Fig. 7a,b). Notably, the effect was not present at thermoneutral conditions but at 22 °C, and it was even more pro￾nounced at 16 °C (Fig. 3e and Supplementary Fig. 7a,b), indicat￾ing that increased dietary intake and BAT-dependent processing of cholesterol at lower temperatures is critical for efficient CYP7B1- mediated bile acid production. To investigate the basis of higher fecal CBA levels in cold-housed mice, we first quantified the ileal expression of apical sodium bile transporter (ASBT), which is required for efficient CBA re-uptake as part of the enterohepatic circulation20. Cold exposure led to higher ASBT expression at the mRNA (Slc10a2; Fig. 3f), as well as protein (Fig. 3g), level, which was lost in Cyp7b1−/− mice (Supplementary Fig. 7c,d). Furthermore, cold-housing resulted in a trend toward higher bile acid levels in portal (Fig. 3h) and systemic (Fig. 3i) blood. Taken together, these data argue against diminished CBA re-uptake capacity as a cause for higher fecal bile acid levels. Another mechanism for increased fecal concentrations of CBA could be due to reduced bile acid deconjungation by bile salt hydrolases (BSHs), which are expressed by a number of gut bacteria26. To address this hypothesis we depleted the intestinal microbiota by using an anti￾biotic cocktail. Relative to that in the untreated controls, this interven￾tion resulted in higher fecal CBA content both in warm-housed and cold-exposed mice (Fig. 3j,k), which could be explained by the com￾plete eradication of gut BSH activity by antibiotic treatment (Fig. 3l). Although antibiotic treatment did not affect the expression of genes related to bile acid metabolism (Supplementary Fig. 8a), hepatic UBA but not CBA levels were lower (Supplementary Fig. 8b,c). Of note, higher BSH activity was observed in cold-exposed mice in the absence of antibiotics than in warm-housed mice (Fig. 3l), resulting in higher fecal taurine levels (Fig. 3m). Cumulatively, these results suggest that saturation of ASBT rather than diminished decon￾jugation by BSH activity is responsible for higher fecal CBA excretion after cold exposure. Bile acid excretion depends on a BAT–liver cholesterol axis Cold-activated BAT efficiently processes dietary lipids carried by post￾prandial lipoproteins and promotes receptor-mediated uptake of the respective cholesterol-rich remnants by the liver15,17. Accordingly, in contrast to the decline in circulating lipids that occurs in WT mice with cold (Fig. 1g–i), we observed higher amounts of circulating cholesterol￾rich lipoproteins in cold-housed versus warm-housed mice that were deficient in the LDL receptor (LDLR), which is the main receptor for hepatic remnants (Fig. 4a,b). When we combined Ldlr−/− genotype with a liver-specific knockout of the gene encoding the alternative hepatic lipoprotein receptor, LDLR-related protein 1 (Lrp1), we observed even higher lipoprotein levels (Fig. 4c,d), indicating blunted remnant clearance in the absence of both receptors. Notably, the cold-induced increase in fecal bile acid excretion in Ldlr−/− mice was only 50% of that in WT mice, and it was nearly abolished in mice that lacked LDLR and hepatic LRP1 (Fig. 4e–g). Taken together, these findings indicate that fecal bile acid excretion is dependent on hepatic cholesterol that is deliv￾ered by postprandial lipoproteins generated by cold-activated BAT. To assess the contribution of dietary cholesterol to cold-induced bile acid synthesis, we first blocked dietary cholesterol resorption by inhibiting the intestinal cholesterol transporter Niemann-Pick-1-like 1 (NPC1L1) with ezetimibe (EZ). Notably, combining cold-exposure and EZ treatment resulted in lower levels of plasma lipids and choles￾terol-rich lipoproteins, even more so than with cold exposure alone (Fig. 5a,b). Consistent with diminished dietary cholesterol uptake and hepatic lipid levels after EZ treatment (Fig. 5c), we observed compensatory hepatic upregulation of the gene encoding the rate￾limiting enzyme of the cholesterol synthesis pathway (Hmgcr), as well as of Ldlr (Fig. 5d), whereas expression of Cyp7b1 was unaf￾fected (Supplementary Fig. 9a). Liver bile acid levels of cold-housed EZ-treated mice were comparable to those of cold-housed control mice (Fig. 5e and Supplementary Fig. 9b,c). In contrast, the fecal bile acid concentrations of EZ-treated mice were not higher after cold exposure, and bile acid levels were almost identical to those in

ARTICLES ●cold owam●cold 2,500 2,000 200 Cold Cyp7b1-- o WT cold 4.000 Cold 250- Cold 81001头鲁 UBA CBA Housing temperature (C) Sic 10a2 UBA CBA O Warm k O War●cod● Warm AB● Cold AB =1 80 是0001 actvity Figure 3 Cold exposure promotes fecal excretion of CYP7Bl-derived bile acids. (a, b)Relative levels of bile acid species (a)and the sum of the amounts of UBA and CBa species(b)in the feces of cold- housed and warm-housed mice(n=5 mice per group). (c, d)Heat map representation of the fecal levels of bile acid species in cold-housed WT and Cyp/- mice, relative to those in the respective warm-housed controls, (c)and total levels of UBA g and CBA species in warm-housed and cold-housed WT and Cyp 7b1--mice(d)(n= 3 mice per group). (e)Bile acid levels in the feces of mice that were 9 of S/c10a2, which encodes apical sodium-dependent bile transporter(ASBT)(n=8 mice per group), (f)and ASBT protein (n=3 mice per group)(g)in arm-housed and cold- housed mice. One representative out of two technical replicates of three biological replicates is shown. Uncropped western blot images are shown in Supplementary Figure 15.(h, i) Bile acid concentrations in plasma from portal vein blood (warm-housed, n=8 mice; cold-housed 9 mice)(h)and in plasma from systemic blood (n= 4 mice per group)(i)of warm-housed and cold- housed mice. gj, k)Heat map showing fecal bile acid species (relative to those in warm-housed controls)(), as well as total UBA and CBa levels (k), from warm-housed and cold-housed mice that were not or were treated with antibiotics(AB)(warm-housed: no AB (warm), n=l0 mice; +AB, n= 10 mice; cold- housed: no AB(cold), n=10 mice: +AB n=9 mice).(1)Cecal BSH activity in warm-housed and cold-housed mice that were not or were treated with antibiotics (warm-housed: no AB (warm), n=7 mice: +AB, n= 6 mice; cold- housed: no AB(cold), n=6 mice; +AB, n=5 mice).(m) Fecal taurine concentration in warm- housed and cold- housed mice(n= 5 mice per group). Throughout, data are mean +s.e. m 'P<0.05, P<0.01, P<0.00l; by unpaired two tailed Students t-test (a, b, d-f, h, i, m)or two-way ANOVA (k, i) warm-housed control mice(Fig. 5f-h), indicating that dietary cho- Bile acids shape the gut microbiome in cold-housed mice lesterol uptake determined cold-induced fecal bile acid excretion. It is conceivable that cold-related endogenous metabolites generated Treatment with EZ resulted in a higher fecal cholesterol content with- by the host, such as bile acids, have an effect on gut bacteria27.We ut affecting food intake or fecal mass( Supplementary Fig. 9d-f). therefore postulated that cold-induced bile acid levels would allow Elevated dietary cholesterol intake induced by cold exposure was selection of intestinal bacteria and thus determine the microbiome in (Fig. 1j, k), we calculated the absorption of cholesterol and its conver- housed and cold-housed mice that were or were not treated with EZ sion to bile acids, which we determined from feces(Supplementary MDS analysis revealed a clear separation of the cold-housed and Table 1b), and observed 17. 1% conversion in warm-housed mice, warm-housed groups( Fig. 5i), whereas EZ-treated cold-housed mice whereas cold exposure caused a 70% increase in the conversion rate; clustered similarly to the untreated warm-housed group( Fig. 5i). thus, 29.2% of absorbed cholesterol was converted to bile acids in The EZ-treated warm-housed group clustered in a position opposite cold-housed mice. Taking into account the higher amount of choles- to the untreated cold-housed group, which is consistent with lower terol absorption, cold-housed mice produced 5 8-fold more amounts fecal bile acid levels observed in the EZ-treated warm-housed mice, of bile acids than warm-housed controls compared to that in the untreated warm-housed group( Fig. 5f)

© 2017 Nature America, Inc., part of Springer Nature. All rights reserved. Ar t i c l e s  advance online publication nature medicine warm-housed control mice (Fig. 5f–h), indicating that dietary cho￾lesterol uptake determined cold-induced fecal bile acid excretion. Treatment with EZ resulted in a higher fecal cholesterol content with￾out affecting food intake or fecal mass (Supplementary Fig. 9d–f). Elevated dietary cholesterol intake induced by cold exposure was accompanied by higher fecal content of cholesterol (Supplementary Table 1b). On the basis of metabolic studies using a radioactive tracer (Fig. 1j,k), we calculated the absorption of cholesterol and its conver￾sion to bile acids, which we determined from feces (Supplementary Table 1b), and observed 17.1% conversion in warm-housed mice, whereas cold exposure caused a 70% increase in the conversion rate; thus, 29.2% of absorbed cholesterol was converted to bile acids in cold-housed mice. Taking into account the higher amount of choles￾terol absorption, cold-housed mice produced 5.8-fold more amounts of bile acids than warm-housed controls. Bile acids shape the gut microbiome in cold-housed mice It is conceivable that cold-related endogenous metabolites generated by the host, such as bile acids, have an effect on gut bacteria27. We therefore postulated that cold-induced bile acid levels would allow selection of intestinal bacteria and thus determine the microbiome in cold-housed mice. To test this possibility, we performed sequencing of the gene encoding the 16S rRNA using fecal samples from warm￾housed and cold-housed mice that were or were not treated with EZ. MDS analysis revealed a clear separation of the cold-housed and warm-housed groups (Fig. 5i), whereas EZ-treated cold-housed mice clustered similarly to the untreated warm-housed group (Fig. 5i). The EZ-treated warm-housed group clustered in a position opposite to the untreated cold-housed group, which is consistent with lower fecal bile acid levels observed in the EZ-treated warm-housed mice, as compared to that in the untreated warm-housed group (Fig. 5f). 1,000 Warm Cold Warm Cold Warm Cold 55 Mr (kDa) 40 Mr (kDa) 250 200 150 100 50 0 * * * * * * ** * * * 800 2,500 2,000 1,500 1,000 500 0 2,000 2,000 3,000 4,000 2.0 1.5 ASBT β-actin Gene expression (fold) 1.0 0.5 0.0 Slc10a2 * * * * WT cold AAV-GFP AAV-Cyp7b1 1,500 1,000 500 1,000 0 0 30 22 16 UBA CBA UBA CBA UBA CBA Warm AB Cold AB UBA 6 4 2 * 0 CBA Housing temperature (°C) Cold wild type 0 1 >7 0 1 >19 Cold Cyp7b1–/– Cyp7b1–/– cold Bile acids (ng/mg) Bile acids (ng/mg) Bile acids (ng/mg) Bile acids (µM) Bile acids (ng/mg) Bile acids (µM) 600 400 200 0 CA β-MCA DCA/CDCA UDCA α/ω-MCA T-α/β-MCA THDCA TUDCA TCDCA/TDCA TLCA TCA GCDCA GCA CA β-MCA α-MCA CDCA/DCA UDCA ω TDCA -MCA T-α/β-MCA THDCA TUDCA TCDCA TLCA TCA GCDCA GCA CA β-MCA α-MCA CDCA/DCA UDCA ω-MCA T-β-MCA T-α-MCA THDCA TUDCA TDCA TLCA TCA GCDCA GCA Warm Cold Warm Cold Warm Cold Warm Cold Warm AB Cold AB UBA CBA Bile acids (ng/mg) 8,000 0.08 100 80 60 40 20 0 0.06 0.04 0.02 0.00 BSH activity Taurine (mmol/10 min) *** Concentration (ng/mg) *** *** *** *** *** *** ** Taurine * *** *** ** 6,000 4,000 2,000 0 a b c d e f g h i j k l m Figure 3 Cold exposure promotes fecal excretion of CYP7B1-derived bile acids. (a,b) Relative levels of bile acid species (a) and the sum of the amounts of UBA and CBA species (b) in the feces of cold-housed and warm-housed mice (n = 5 mice per group). (c,d) Heat map representation of the fecal levels of bile acid species in cold-housed WT and Cyp7b1−/− mice, relative to those in the respective warm-housed controls, (c) and total levels of UBA and CBA species in warm-housed and cold-housed WT and Cyp7b1−/− mice (d) (n = 3 mice per group). (e) Bile acid levels in the feces of mice that were housed at the indicated ambient temperatures after infection with either AAV-GFP or AAV-Cyp7b1 (n = 7 mice per group). (f,g) Ileal mRNA expression of Slc10a2, which encodes apical sodium-dependent bile transporter (ASBT) (n = 8 mice per group), (f) and ASBT protein (n = 3 mice per group) (g) in warm-housed and cold-housed mice. One representative out of two technical replicates of three biological replicates is shown. Uncropped western blot images are shown in Supplementary Figure 15. (h,i) Bile acid concentrations in plasma from portal vein blood (warm-housed, n = 8 mice; cold-housed, n = 9 mice) (h) and in plasma from systemic blood (n = 4 mice per group) (i) of warm-housed and cold-housed mice. (j,k) Heat map showing fecal bile acid species (relative to those in warm-housed controls) (j), as well as total UBA and CBA levels (k), from warm-housed and cold-housed mice that were not or were treated with antibiotics (AB) (warm-housed: no AB (warm), n = 10 mice; +AB, n = 10 mice; cold-housed: no AB (cold), n = 10 mice; +AB, n = 9 mice). (l) Cecal BSH activity in warm-housed and cold-housed mice that were not or were treated with antibiotics (warm-housed: no AB (warm), n = 7 mice; +AB, n = 6 mice; cold-housed: no AB (cold), n = 6 mice; +AB, n = 5 mice). (m) Fecal taurine concentration in warm-housed and cold￾housed mice (n = 5 mice per group). Throughout, data are mean ± s.e.m. *P < 0.05, **P < 0.01, ***P < 0.001; by unpaired two-tailed Student’s t-test (a,b,d–f,h,i,m) or two-way ANOVA (k,i)

ARTICLES b o+-warm c o Ldi-Lup7-Cre+warm d o Ld-Lrp1-Cre+ ●Ldr ●Ld-cold ●Lal-Lrp1-Cre+co 00 TRL iii elis ile. atus. i,ies lato sh i CDCA ●Ld-wamo f8 ●Ld-ccd● 4000 ● Lair"col O Ldr-Lrp1-Cre+warm 音 Figure 4 Hepatic uptake of cholesterol-rich lipoproteins determines fecal bile acid excretion in cold-exposed mice (a-d)Plasma lipid analysis in a (LdIr-Lrpl-Cre, n=6 mice per group)(c, d). Plasma triglyceride and cholesterol levels (a, c)and lipoprotein profiles from corresponding pooled plasma cold-housed mice(WT, n= 13 mice per group; warm-housed Ldirk-, n= 13 mice; cold-housed Ldlr-, n= 12 mice; Ldir-Lrpl-Cre, n=6 mice per group). Throughout, data are mean ts.e. m. P<0.05, " P<0.01, **P<0.00l; by unpaired two-tailed Students t-test (a, c)or two way ANova (e-g) 2 However, hierarchical cluster analysis showed a separation from the Unexpectedly, we observed a clear separation of the gut microbiota other groups for only the cold-housed mice that were not treated with from WT mice versus those from Cyp7b1--mice, irrespectively of a without manipulating its absorption rate, we performed a study using on the gut microbiome seen in the mDS plot may explain the rela mice that were fed a chow diet supplemented without or with choles- tively weak effect of cold also in the WT mice. Notably, Cyp Zbl over- terol Relative to the warm-housed controls, cold-exposed mice showed expression was associated with changes in the composition of the gut substantial changes in the gut microbiome on both diets(Fig. 5j microbiome. We observed that the effects on gut bacteria were present and Supplementary Fig. 10a, b). Notably, even in the absence of dietary in mice that were housed at 22C but not in mice that were housed at cholesterol, significantly higher fecal bile acid excretion was observed thermoneutrality(Fig. 51, m and Supplementary Fig. 1la-c), indi- in cold-housed versus warm-housed mice(Supplementary Fig 10c, d ), cating that both Cyp 7bI upregulation and accelerated cholesterol indicating that endogenously synthesized cholesterol is sufficient metabolism via activated BAT are important for changes in the for some cold-induced bile acid production. Despite Cyp 7bl induc- microbiome. To investigate the relevance of cold-induced bile acids tion and higher bile acid levels in liver in response to B3-adrenergic for changes in the gut microbiome by an independent approach,we receptor activation(Fig. 2g, h), we observed only a trend toward performed experiments with an additional genetic model, Abcb4-- higher fecal bile acid levels without a clear effect on the gut microbi-(also known as Mdr2--)mice, which lack the canalicular phospholi ome(Supplementary Fig. 10e-g) Taking into account the diurnal pid transporter MDR2, resulting in abrogated cold-induced changes food intake patterns of mice, the resulting intermittent B3-adren- in fecal bile acid levels(Supplementary Fig. lld ). Consistent with ergic receptor stimulation by the CL316, 243 supplied with the diet the observation of similar bile acid levels in the feces of warn likely explains the less pronounced effects. Next we asked whether housed and cold-housed Mdr2-/- mice, we did not observe differ the loss of cold-induced fecal bile acids observed in Cyp7b1-/-mice ences in the gut microbiota( Fig. 5n and Supplementary Fig. lle) (Fig 3c, d) had an effect on the composition of the gut microbiota. In summary, these data show that the cold-induced changes in the DNLINE PUBLICATION

© 2017 Nature America, Inc., part of Springer Nature. All rights reserved. a r t i c l e s nature medicine advance online publication  However, hierarchical cluster analysis showed a separation from the other groups for only the cold-housed mice that were not treated with EZ (Supplementary Fig. 9g). To address the role of dietary cholesterol without manipulating its absorption rate, we performed a study using mice that were fed a chow diet supplemented without or with choles￾terol. Relative to the warm-housed controls, cold-exposed mice showed substantial changes in the gut microbiome on both diets (Fig. 5j and Supplementary Fig. 10a,b). Notably, even in the absence of dietary cholesterol, significantly higher fecal bile acid excretion was observed in cold-housed versus warm-housed mice (Supplementary Fig. 10c,d), indicating that endogenously synthesized cholesterol is sufficient for some cold-induced bile acid production. Despite Cyp7b1 induc￾tion and higher bile acid levels in liver in response to β3-adrenergic receptor activation (Fig. 2g,h), we observed only a trend toward higher fecal bile acid levels without a clear effect on the gut microbi￾ome (Supplementary Fig. 10e–g). Taking into account the diurnal food intake patterns of mice, the resulting intermittent β3-adren￾ergic receptor stimulation by the CL316,243 supplied with the diet likely explains the less pronounced effects. Next we asked whether the loss of cold-induced fecal bile acids observed in Cyp7b1−/− mice (Fig. 3c,d) had an effect on the composition of the gut microbiota. Unexpectedly, we observed a clear separation of the gut microbiota from WT mice versus those from Cyp7b1−/− mice, irrespectively of housing temperature (Fig. 5k). The dominant effect of the genotype on the gut microbiome seen in the MDS plot may explain the rela￾tively weak effect of cold also in the WT mice. Notably, Cyp7b1 over￾expression was associated with changes in the composition of the gut microbiome. We observed that the effects on gut bacteria were present in mice that were housed at 22 °C but not in mice that were housed at thermoneutrality (Fig. 5l,m and Supplementary Fig. 11a–c), indi￾cating that both Cyp7b1 upregulation and accelerated cholesterol metabolism via activated BAT are important for changes in the gut microbiome. To investigate the relevance of cold-induced bile acids for changes in the gut microbiome by an independent approach, we performed experiments with an additional genetic model, Abcb4−/− (also known as Mdr2−/−) mice, which lack the canalicular phospholi￾pid transporter MDR2, resulting in abrogated cold-induced changes in fecal bile acid levels (Supplementary Fig. 11d). Consistent with the observation of similar bile acid levels in the feces of warm￾housed and cold-housed Mdr2−/− mice, we did not observe differ￾ences in the gut microbiota (Fig. 5n and Supplementary Fig. 11e). In summary, these data show that the cold-induced changes in the 3,000 Ldlr –/– warm Ldlr –/– cold Ldlr –/–Lrp1-Cre + warm Ldlr –/–Lrp1-Cre + cold Ldlr –/–Lrp1-Cre + warm Ldlr –/–Lrp1-Cre + cold 400 300 200 100 0 400 300 200 100 0 0 10 Fraction 20 30 0 10 Fraction 20 30 LDL TRL LDL TRL Cholesterol (mg/dl) Cholesterol (mg/dl) 2,000 1,000 0 3,000 4,000 2,000 1,000 0 *** TG Chol TG * Chol *** Lipids (mg/dl) Lipids (mg/dl) Ldlr –/– warm Ldlr –/– cold a b c d Ldlr –/– warm Ldlr –/– cold Ldlr –/–Lrp1-Cre + warm Ldlr –/–Lrp1-Cre + cold WT warm WT cold 10 *** *** *** *** *** *** *** *** *** *** *** ** *** *** * *** ** 8 Bile acids (ng/mg, fold) 6 4 2 0 CA β-MCA DCA UDCA α-MCA ω-MCA CDCA e Ldlr –/– warm Ldlr –/– cold Ldlr –/–Lrp1-Cre + warm Ldlr –/–Lrp1-Cre + cold WT warm WT cold *** *** *** *** *** *** *** *** *** *** *** *** *** *** * 6,000 4,000 2,000 0 *** * ** * * Bile acids (ng/mg, fold) Bile acids (ng/mg) 200 150 100 50 0 T-α/β-MCA TUDCA TDCA TCDCA GCA Ldlr –/–Lrp1-Cre + warm Ldlr –/–Lrp1-Cre + cold WT warm WT cold Ldlr –/– warm Ldlr –/– cold f g Figure 4 Hepatic uptake of cholesterol-rich lipoproteins determines fecal bile acid excretion in cold-exposed mice. (a–d) Plasma lipid analysis in warm-housed and cold-housed Ldlr−/− mice (n = 7 mice per group) (a,b), as well as in mice that lacked both LDLR and hepatic LRP1 expression (Ldlr−/−Lrp1-Cre, n = 6 mice per group) (c,d). Plasma triglyceride and cholesterol levels (a,c) and lipoprotein profiles from corresponding pooled plasma samples (b,d) are shown. (e–g) Fecal levels of UBA (e) and CBA (f) species, as well as the total amount of fecal bile acids (g) from warm-housed and cold-housed mice (WT, n = 13 mice per group; warm-housed Ldlr−/−, n = 13 mice; cold-housed Ldlr−/−, n = 12 mice; Ldlr−/−Lrp1–Cre, n = 6 mice per group). Throughout, data are mean ± s.e.m. *P < 0.05, **P < 0.01, ***P < 0.001; by unpaired two-tailed Student’s t-test (a,c) or two-way ANOVA (e–g)

ARTICLES b d Wam● Cold o Warm EZ。 Cold ez ● Coldez 意16n邑 6i是 TG Chol CBA ng) ug owam。WamE g owam。 Warm EZ o Warm o Warm EZ ● Cold eZ ●cod● Cold EZ ●cod● Cold ez 芦芹 g28·88:-201·是g8目e量8e.。201 i..dit △Wam △wam ◆ Warm EZ口coEz ◆ Warm Chol口 Cold cho ◇cp7 bl warm口 ◇ 8△◇ ◇ 0.00● ◇ 0051◇ °公°△ 0.10 △ -0.10-0050000050.10 .3-02.10.00.102 MDs1(37%) Ds1(52%) s1(80%) △AAV-GFP●AAv-Cyp7b △ AAV-GFP● AAV-Cyp7b1 △Mdr2wam●Md2'co 0.1 ● -0030.000.03006 10-0050.000.05 .1000.1 MDs1(43%) Figure 5 Cold-induced bile acid excretion determines the composition of the gut microbiome (a-h) Plasma lipids (a), lipoprotein profiles of poole lasma(b), hepatic lipid content (c), hepatic expression of HMG-CoA reductase( Hmgcr) and Ldlr mRNAs(d), hepatic UBA and CBA levels (e), total fecal UBA and CBA levels(f), as well as fecal unconjugated (g) and conjugated and unconjugated a-MCA (h) bile acid species, as determined in arm-housed and cold-housed mice that were treated without or with ezetimibe(Ez)(for a-d, f -h, n=8 mice per group; for e, warm-housed: no ez warm), n=6 mice: +EZ, n=7 mice; cold-housed: no EZ(cold), n=8 mice: +EZ, n=8 mice). (i-n)MDS plots of weighted UniFrac-distance-based 6s rRNA-encoding sequence analysis of fecal samples from warm-housed and cold- housed mice that were treated without or with Ez (warm-housed: no EZ (warm), n=8 mice; +EZ, n= 10 mice; cold-housed: no EZ (cold), n=7 mice; +EZ, n=9 mice)(i), warm-housed and cold- housed mice that were fed a chow diet supplemented without or with cholesterol (warm-housed, no added Chol (warm)n=8 mice; +Chol, n=8 mice; cold- housed: no added 4 mice per group)(k), mice that were infected with either AAV-GFP or AAV-Cyp7bl and housed at 30C (AAV-GFP, n= 7 mice; AAv-Cyp7bl, n=5 mice)(I)or 22C(AAV-GFP, n=7 mice per group)(m), and warm-housed and cold- housed Mdr2--mice(warm-housed, n=7 mice; cold-housed n=8 mice)(n). Throughout, data are mean ts.e. m. 'P<0.05, *P<0.01,"*P<0.001; by two-way ANovA (a-h)

© 2017 Nature America, Inc., part of Springer Nature. All rights reserved. Ar t i c l e s  advance online publication nature medicine Warm Cold Warm EZ Cold EZ 8 6 4 2 0 Gene expression (fold) Bile acids (ng/mg, fold) Bile acids (µg/mg) Bile acids (ng/mg) Bile acids (ng/mg, fold) MDS2 (19%) MDS2 (19%) MDS2 (19%) MDS2 (21%) MDS2 (17.1%) MDS2 (24%) Liver lipids (per mg/protein) Cholesterol (mg/dl) Lipids (mg/dl) 2.0 *** *** ** *** ** *** * 1.5 1.0 0.5 0.0 Fraction TG (mg) Chol (µg) 40 30 20 10 0 0 10 20 30 250 200 150 100 50 0 TG UBA UBA CBA CBA CA β-MCA CDCA/DCA UDCA ω-MCA α-MCA T-α/β-MCA THDCA TUDCA TDCA TCDCA TLCA TCA Chol 20 15 10 5 0 5 4 3 2 1 0 10 8 6 4 2 0 100 80 60 40 20 0 200 150 250 100 50 0 0.10 0.05 0.00 –0.05 0.10 0.05 0.00 –0.05 –0.10 0.2 0.1 0.0 –0.1 0.05 0.00 –0.05 0.10 0.05 0.00 –0.05 –0.10 –0.3 –0.2 –0.1 0.0 0.1 0.2 –0.2 –0.1 0.0 0.1 0.2 –0.1 0.0 0.1 0.2 –0.10 –0.05 0.00 0.05 0.10 –0.03 0.00 0.03 0.06 MDS1 (37%) –0.10 –0.05 0.00 0.05 MDS1 (30%) MDS1 (43%) MDS1 (52%) MDS1 (80%) MDS1 (42%) Hmgcr Ldlr Warm Warm EZ Cold Cold EZ Warm Cold Warm EZ Cold EZ Warm Warm EZ Cold Cold EZ Warm Warm EZ Cold Cold EZ Warm Warm EZ Cold Cold EZ Warm Cold AAV–GFP AAV–Cyp7b1 AAV–GFP AAV–Cyp7b1 Mdr2–/– warm Mdr2–/– cold Warm EZ Cold EZ Warm Cold Warm Chol 30 °C 22 °C Cold Chol WT warm WT cold Cyp7b1–/– warm Cyp7b1–/– cold 0.050 –0.050 0.025 –0.025 0.000 a b c d e f g h i j k l m n *** *** *** *** * *** *** *** *** * *** *** * *** ** *** *** *** *** *** * *** ***** *** * ***** *** **** * * ** ** * ** **** *** *** *** ** ** ** ** ** ** ** ** ** *** ** *** *** *** *** Figure 5 Cold-induced bile acid excretion determines the composition of the gut microbiome. (a–h) Plasma lipids (a), lipoprotein profiles of pooled plasma (b), hepatic lipid content (c), hepatic expression of HMG-CoA reductase (Hmgcr) and Ldlr mRNAs (d), hepatic UBA and CBA levels (e), total fecal UBA and CBA levels (f), as well as fecal unconjugated (g) and conjugated and unconjugated α-MCA (h) bile acid species, as determined in warm-housed and cold-housed mice that were treated without or with ezetimibe (EZ) (for a–d,f–h, n = 8 mice per group; for e, warm-housed: no EZ (warm), n = 6 mice; +EZ, n = 7 mice; cold-housed: no EZ (cold), n = 8 mice; +EZ, n = 8 mice). (i–n) MDS plots of weighted UniFrac-distance-based 16S rRNA-encoding sequence analysis of fecal samples from warm-housed and cold-housed mice that were treated without or with EZ (warm-housed: no EZ (warm), n = 8 mice; +EZ, n = 10 mice; cold-housed: no EZ (cold), n = 7 mice; +EZ, n = 9 mice) (i), warm-housed and cold-housed mice that were fed a chow diet supplemented without or with cholesterol (warm-housed, no added Chol (warm) n = 8 mice; +Chol, n = 8 mice; cold-housed: no added Chol (cold), n = 7 mice; +Chol, n = 8 mice) (j), warm-housed and cold-housed WT control and Cyp7b1−/− mice (WT:, n = 5 mice per group; Cyp7b1−/−, n = 4 mice per group) (k), mice that were infected with either AAV-GFP or AAV-Cyp7b1 and housed at 30 °C (AAV-GFP, n = 7 mice; AAV-Cyp7b1, n = 5 mice) (l) or 22 °C (AAV-GFP, n = 7 mice per group) (m), and warm-housed and cold-housed Mdr2−/− mice (warm-housed, n = 7 mice; cold-housed, n = 8 mice) (n). Throughout, data are mean ± s.e.m. *P < 0.05, **P < 0.01, ***P < 0.001; by two-way ANOVA (a–h)

ARTICLES o Mock●Ez 2,: 心 Housing temperature (C) Housing temperature (C) h Cyp7b1- WT·cvp7b °A° oWT●c7b7 O AAV-GFP· AAV-Cyp7b1 O AAV-GFP O AAV-GFP 6000 9401●AVcp 38 Housing temperature (C) z Figure 6 CYP7B1-derived bile acids promote adaptive thermogenesis. (a, b)Body temperature in Wt(n=8 per group)(a)and dbldb(30"C, s with cholesterol (n=8 mice per group).(d) Representative photographs (top)and histology images(bottom)of BaT from mice that were housed at cC(n=2 mice pet ar lel with n=5 images per mouse). Scale bars, 1 cm(top)and 100 um(bottom).(e, f) Relative expression of genes involved in the housekeeping protein AKT)(right)of UCPl expression in BAT and inguinal WAT from WT(n= 9)and Cyp 7b1--(n=8)mice housed at 6C One representative out of three technical replicates of nine(Wt, n=9 mice)or eight (Cyp 7b1-ki n=8 mice)biological replicates is shown. Uncropped estern blot images are shown in Supplementary Figure 15.(h, i)Oxygen consumption profiles(h)and quantification (i)in WT and Cyp- mice that ere housed at 6C and monitored during a 12-h light and a 12-h dark phase(n=9 mice per group).i) Body temperature in WT and Cyp7b1--mice, measured at the indicated ambient temperatures (n=9 mice per group).( k, I )Profiles(k)and quantification (I)of oxygen consumption in mice that wer infected with AAV-GFP or AAV-Cyp7bl and housed at 22C, as monitored during a 12-h light and a 12-h dark phase (n= 6 mice per group).(m)Body mperature in mice that were infected with AAv-GFP (n=7 mice)or AAV-Cyp7bl(n= 4 mice), as measured at the indicated ambient temperature Throughout, data are mean ts.e. m.P<0.05, *P<0.01, **P<0.00l; by unpaired two tailed Students t-test (a-c, e-m). intestinal microbiome are dependent on the synthesis and biliary thermogenesis, we observed by indirect calorimetry that Cyp7b1-- excretion of bile acids by the host mice consumed significantly less O2, which is indicative of dimin- Exogenous bile acids are known to stimulate BAT in mice and temperatures( Fig. response to gradually decreased ambient CYP7Bl-derived bile acids promote thermogenesis body tem gy expenditure,(Fig. 6h, i), and we observed a lower humans24. 25. To study the thermo-metabolic consequences of higher Moreover, AAV-mediated CyP7bl overexpression in the liver evels of endogenous bile acids after cold exposure, we investigated Supplementary Fig. 13a)caused slightly higher expression of genes whether dietary cholesterol uptake and its subsequent conversion via involved in thermogenesis in BAT and inguinal WAT, as compared EZ treatment in cold-housed mice resulted in lower plasma bile acid Fig. 13b, c). This. expressing the control AAV( Supplementary levels as compared to that in untreated controls(Supplementary (Fig. 6k, I), greater tail heat loss(Supplementary Fig. 13d)and higher WT and dbldb mice that were cold-housed(Fig 6a, b). Conversely, that cholesterol conversion to bile acids via CYP7B1 modulates adap dietary cholesterol supplementation led to higher plasma levels of bile tive thermogenesis by brown and beige adipocytes acids(Supplementary Fig 12c, d) and body temperature(Fig. 6c) To study the alternative bile acid pathway in this context, we ana- DISCUSSION lyzed thermogenic parameters in cold-exposed WT and Cyp7b1-l- Obesity and HFDs are known to be related to shifts in the gut micro- miceRelative to WT controls, CyP7b1-- mice showed higher BAT biota. These in turn have an effect on the development and progres- lipid content(Fig. 6d), lower thermogenic gene expression(Fig. 6e, f) sion of chronic metabolic diseases through alterations in circulating andlower amountsof mitochondrial uncoupling protein 1(UCP1), which metabolites and hormones2-6. Recent studies indicate that cold is essential for heat production via adaptive thermogenesis, ( Fig. 6g) exposure leading to the thermogenic activation of beige and brown in BAT and inguinal WAT. Consistent with impaired adaptive adipocytes is associated with changes in the gut microbiome DNLINE PUBLICATION

© 2017 Nature America, Inc., part of Springer Nature. All rights reserved. a r t i c l e s nature medicine advance online publication  intestinal microbiome are dependent on the synthesis and biliary excretion of bile acids by the host. CYP7B1-derived bile acids promote thermogenesis Exogenous bile acids are known to stimulate BAT in mice and humans24,25. To study the thermo-metabolic consequences of higher levels of endogenous bile acids after cold exposure, we investigated whether dietary cholesterol uptake and its subsequent conversion via CYP7B1 would affect thermogenesis. Consistent with this concept, EZ treatment in cold-housed mice resulted in lower plasma bile acid levels as compared to that in untreated controls (Supplementary Fig. 12a,b), which was associated with lower body temperatures in WT and db/db mice that were cold-housed (Fig. 6a,b). Conversely, dietary cholesterol supplementation led to higher plasma levels of bile acids (Supplementary Fig. 12c,d) and body temperature (Fig. 6c). To study the alternative bile acid pathway in this context, we ana￾lyzed thermogenic parameters in cold-exposed WT and Cyp7b1−/− mice. Relative to WT controls, Cyp7b1−/− mice showed higher BAT lipid content (Fig. 6d), lower thermogenic gene expression (Fig. 6e,f) and lower amounts of mitochondrial uncoupling protein 1 (UCP1), which is essential for heat production via adaptive thermogenesis, (Fig. 6g) in BAT and inguinal WAT. Consistent with impaired adaptive thermogenesis, we observed by indirect calorimetry that Cyp7b1−/− mice consumed significantly less O2, which is indicative of dimin￾ished energy expenditure, (Fig. 6h,i), and we observed a lower body temperature in response to gradually decreased ambient temperatures (Fig. 6j). Moreover, AAV-mediated Cyp7b1 overexpression in the liver (Supplementary Fig. 13a) caused slightly higher expression of genes involved in thermogenesis in BAT and inguinal WAT, as compared to that in liver overexpressing the control AAV (Supplementary Fig. 13b,c). This was accompanied by higher O2 consumption (Fig. 6k,l), greater tail heat loss (Supplementary Fig. 13d) and higher body temperature (Fig. 6m). In conclusion, these data demonstrate that cholesterol conversion to bile acids via CYP7B1 modulates adap￾tive thermogenesis by brown and beige adipocytes. DISCUSSION Obesity and HFDs are known to be related to shifts in the gut micro￾biota. These in turn have an effect on the development and progres￾sion of chronic metabolic diseases through alterations in circulating metabolites and hormones2–6. Recent studies indicate that cold exposure leading to the thermogenic activation of beige and brown adipocytes is associated with changes in the gut microbiome that are 40 Mock * *** *** ** * * EZ 39 38 37 36 40 * * * Chow WT Cyp7b1–/– WT Cyp7b1–/– WT Cyp7b1–/– WT Cyp7b1–/– WT Cyp7b1–/– WT Cyp7b1–/– WT Cyp7b1 WT Cyp7b1 –/– –/– Chow + Chol 39 2.5 ** *** ** * *** Gene expression (fold) 2.0 1.5 1.0 0.5 0.0 2.5 UCP1 AKT UCP1 AKT BAT ingWAT BAT ** ** UCP1/AKT (a.u.)1.5 6,000 4,000 VO2 (ml/h/kg) VO2 (ml/h/kg) 2,000 0 6,000 AAV-GFP AAV-Cyp7b1 AAV-GFP AAV-Cyp7b1 AAV-GFP AAV-Cyp7b1 4,000 VO2 (ml/h/kg) VO2 (ml/h/kg) 2,000 0 6,000 8,000 4,000 2,000 0 6,000 4,000 2,000 Light phase Dark phase Light phase Dark phase * * 0 1.0 0.5 0.0 ing WAT Gene expression (fold) 2.0 1.5 1.0 0.5 0.0 38 37 36 40 db/db 30 °C db/db 16 °C db/db 30 °C EZ db/db 16 °C EZ 38 36 34 32 30 22 16 6 30 22 16 Ucp1 Dio2 Elovl3 Ppargc1a Prdm16 Ucp1 Dio2 Elovl3 Ppargc1a Prdm16 6 30 22 16 6 Rectal temperature (°C) 40 ** 39 * 38 37 36 Rectal temperature (°C) 30 22 16 40 39 38 37 36 Rectal temperature (°C) Rectal temperature (°C) Rectal temperature (°C) a b c d e f g h i j k l m Housing temperature (°C) Housing temperature (°C) Housing temperature (°C) Housing temperature (°C) Figure 6 CYP7B1-derived bile acids promote adaptive thermogenesis. (a,b) Body temperature in WT (n = 8 per group) (a) and db/db (30 °C, n = 7; 16 °C, n = 8; 30 °C +EZ, n = 7; 16 °C +EZ, n = 6) (b) mice that were treated without or with EZ and housed at the indicated ambient temperatures. (c) Body temperatures in mice that were housed at the indicated ambient temperatures and fed a chow diet supplemented without or with cholesterol (n = 8 mice per group). (d) Representative photographs (top) and histology images (bottom) of BAT from mice that were housed at 6 °C (n = 2 mice per group, with n = 5 images per mouse). Scale bars, 1 cm (top) and 100 µm (bottom). (e,f) Relative expression of genes involved in thermogenesis in BAT (e) and inguinal WAT (f (WT, n = 9 mice; Cyp7b1−/−, n = 8 mice)). (g) Western blot analysis (left) and quantification (relative to the housekeeping protein AKT) (right) of UCP1 expression in BAT and inguinal WAT from WT (n = 9) and Cyp7b1−/− (n = 8) mice housed at 6 °C. One representative out of three technical replicates of nine (WT, n = 9 mice) or eight (Cyp7b1−/−; n = 8 mice) biological replicates is shown. Uncropped western blot images are shown in Supplementary Figure 15. (h,i) Oxygen consumption profiles (h) and quantification (i) in WT and Cyp7b1−/− mice that were housed at 6 °C and monitored during a 12-h light and a 12-h dark phase (n = 9 mice per group). (j) Body temperature in WT and Cyp7b1−/− mice, measured at the indicated ambient temperatures (n = 9 mice per group). (k,l) Profiles (k) and quantification (l) of oxygen consumption in mice that were infected with AAV-GFP or AAV-Cyp7b1 and housed at 22 °C, as monitored during a 12-h light and a 12-h dark phase (n = 6 mice per group). (m) Body temperature in mice that were infected with AAV-GFP (n = 7 mice) or AAV-Cyp7b1 (n = 4 mice), as measured at the indicated ambient temperatures. Throughout, data are mean ± s.e.m. *P < 0.05, **P < 0.01, ***P < 0.001; by unpaired two-tailed Student’s t-test (a–c,e–m)

ARTICLES linked to a beneficial metabolic outcome. Here we demonstrate Cyp7b1 induction as the main driver for a cold-induced increase in that accelerated metabolism of endogenous and dietary cholesterol bile acid synthesis, which increased energy expenditure. This proc in response to cold exposure leads to increased bile acid synthesis ess of enhanced cholesterol catabolism was triggered by accelerated via the alternative pathway. This results in profoundly elevated bile lipoprotein processing in cold-activated BAT, which stimulated acid excretion and distinct changes in the gut bacterial composition, cholesterol flux toward the liver. Furthermore, we demonstrated that providing a mechanism for how cold-induced increase in energy unlike CYP7Al, the rate-limiting enzyme of the classical pathway l8, expenditure alters the gut microbiome. Moreover, cholesterol intake CYP7Bl was not suppressed by pharmacological FXR agonism. Thus, and its hepatic conversion to bile acids are found to activate brown CYP7B1 is not subject to classical negative feedback control regulated and beige adipocytes, suggesting a role in diet-induced thermogen- by elevated bile acid levels. This is of special note, as it enables the esis. The findings of our study are summarized in a model describ- body to maintain cholesterol homeostasis despite massively increased g how BAT activation stimulates cholesterol flux and conversion to cholesterol conversion to bile acids in the cold. Although the exact bile acids, a process that shapes the gut microbiome and promotes mechanism underlying the induction of Cyp7bl in cold-exposed adaptive thermogenesis (Supplementary Fig 14) mice remains to be elucidated, this is the first report demonstrat- The recent discovery that a substantial number of adult humans ing a physiological regulation of the alternative bile acid synthesis retain functional BAT31-35 has introduced adaptive thermogenesis pathway. Notably, CYP7B1 expression was reduced in subjects with a promising approach to enhance energy expenditurel4, 36. In acti- type 2 diabetes, suggesting a role of the alternative pathway for meta- vated BAT, dietary glucose and triglycerides serve as fuels to meet bolic homeostasis in humans. However, previous studies indicated the high-energy demands during heat production 5, 16.37. This in turn higher levels of CYP8B1-derived bile acids in the plasma of humans 8 causes increased appetite and food intake, which leads to the uptake with type 2 diabetes#. 45. It will be of interest to dissect the bile 4 of potentially harmful food components. Cholesterol, which is an acid synthesis pathways in intervention studies, including BAT abundant ingredient of typical energy-rich diets, in particular can- activation by cold, and compare them in healthy individuals and E not be used for heat production and could, in principle, accumulate those with diabetes in arteries, and hence promote cardiovascular disease. However, we Taken together, our findings imply that increased bile acid o previously showed that sustained activation of BAT in mice resulted levels contribute to beneficial effects of beige and brown adipocytes 3 in reduced atherosclerosis 7. Our present study indicates that excess on obesity-associated comorbidities. Mechanistically, elevated z dietary cholesterol is partially converted to bile acids, which are subse- bile acid levels regulate thermogenic responses, which have been s quently removed by fecal excretion, thereby contributing to systemic described to occur in the BAT of mice and humans, likely via TGR5 a have shown that cold exposure induces alterations in the gut micro- the thermogenic activity of skeletal muscle, which may explain the biota, which are associated with improved metabolic parameters 0, I. increased glucose uptake observed in both the BAT and muscle of sub- However, the trigger for cold-induced microbiota remodeling remains jects with type 2 diabetes after cold acclimation 6. Future studies are unknown. In this context, it is of note that bile acids are not only warranted to address the therapeutically relevant question of wheth involved in energy harvesting by facilitating intestinal lipid uptake beneficial cold-mediated metabolic effects are mediated directly by but they also interact with intestinal bacteria26. Notably, we found bile acids acting on TGR5 and other receptors or indirectly by com- g that cold-triggered elevated fecal bile acid levels and alterations in positional changes in the gut microbior gut microbiota could be attenuated by inhibiting the intestinal trans porter NPCILI using ezetimibe, a cholesterol-lowering drug with METHODS tablished anti-atherosclerotic properties. This demonstrates that Methods, including statements of data availability and any associated the enhanced catabolism of cholesterol, resulting in increased fecal accession codes and references, are available in the online version of bile acids, is important for the observed changes in gut microbiota the paper. in cold-exposed mice. Mechanistically, in addition to direct bacte riostatic effects27, increased bile acid levels may indirectly alter the Note: Any Supplementary Information and Source Data files are available in the online version of the paper. gut microbiome via taurine, which we found to be increased as a nsequence of enhanced bile acid deconjugation. In support of this ACKNOWLEDGMENTS idea, it was recently shown that taurine triggers intestinal secretion We thank S.Ehret,BHenkel,AKuhl and E.-M. Azizi for excellent technical of defensins in an inflammasome-dependent manner. In any event, polyclonal antibody, and J. Nedergard and B Cannon (Wenner-Gren Institute, this mechanism would also depend on increased bile acid synthesis Stockholm University)for the UCPl-spcific polyclonal antibody. This work t in the liver. From a general perspective, we provided evidence for the supported by grants funded by the Deutsche Forschungsgemeinschaft (SFB841 und consequences"(LH and of the gut microbiome ported by Merck Sharp Dohme(MSD)(H ) the I project RESOLVE Recently, an induction of Cyp 7bl and other genes related to bile acid FP7-HEALTH-2012-305707 (.H. ) a University Medical Center Hamburg metabolism was described in WT mice in response to cold exposure. Eppendorf MD/PhD fellowship(CS )and the US National Institutes of Health The induction of Cyp7bl after cold exposure or after B3-adrenergi receptor agonism observed in our study suggest a BAT-liver crosstalk AUTHOR CONTRIBUTIONS that regulates hepatic gene expression and thermogenic capacity 0, 4. Aw,C. L.S. and J.H. designed the study, were involved in all aspects of In addition, because germ-free mice have higher hepatic Cyp7b1 lev- the experiments and wrote the manuscript; M CRE-AH and A.E. were elst24s, factors released by the gut microbiota may either directly, were involved in the metabolic studies: M.E. M.D. A.F.C. K und pw.s or indirectly via BAT activation, regulate Cyp7b1 expression. Here, e involved in study design; and all authors read and commented on using Cyp7b1-- mice, we demonstrated the relevance of hepatic the manuscript

© 2017 Nature America, Inc., part of Springer Nature. All rights reserved. Ar t i c l e s 10 advance online publication nature medicine linked to a beneficial metabolic outcome10,11. Here we demonstrate that accelerated metabolism of endogenous and dietary cholesterol in response to cold exposure leads to increased bile acid synthesis via the alternative pathway. This results in profoundly elevated bile acid excretion and distinct changes in the gut bacterial composition, providing a mechanism for how cold-induced increase in energy expenditure alters the gut microbiome. Moreover, cholesterol intake and its hepatic conversion to bile acids are found to activate brown and beige adipocytes, suggesting a role in diet-induced thermogen￾esis. The findings of our study are summarized in a model describ￾ing how BAT activation stimulates cholesterol flux and conversion to bile acids, a process that shapes the gut microbiome and promotes adaptive thermogenesis. (Supplementary Fig. 14). The recent discovery that a substantial number of adult humans retain functional BAT31–35 has introduced adaptive thermogenesis as a promising approach to enhance energy expenditure14,36. In acti￾vated BAT, dietary glucose and triglycerides serve as fuels to meet the high-energy demands during heat production15,16,37. This in turn causes increased appetite and food intake, which leads to the uptake of potentially harmful food components. Cholesterol, which is an abundant ingredient of typical energy-rich diets, in particular can￾not be used for heat production and could, in principle, accumulate in arteries, and hence promote cardiovascular disease. However, we previously showed that sustained activation of BAT in mice resulted in reduced atherosclerosis17. Our present study indicates that excess dietary cholesterol is partially converted to bile acids, which are subse￾quently removed by fecal excretion, thereby contributing to systemic cholesterol homeostasis in cold-exposed mice. Notably, recent studies have shown that cold exposure induces alterations in the gut micro￾biota, which are associated with improved metabolic parameters10,11. However, the trigger for cold-induced microbiota remodeling remains unknown. In this context, it is of note that bile acids are not only involved in energy harvesting by facilitating intestinal lipid uptake but they also interact with intestinal bacteria26. Notably, we found that cold-triggered elevated fecal bile acid levels and alterations in gut microbiota could be attenuated by inhibiting the intestinal trans￾porter NPC1L1 using ezetimibe, a cholesterol-lowering drug with established anti-atherosclerotic properties38. This demonstrates that the enhanced catabolism of cholesterol, resulting in increased fecal bile acids, is important for the observed changes in gut microbiota in cold-exposed mice. Mechanistically, in addition to direct bacte￾riostatic effects27, increased bile acid levels may indirectly alter the gut microbiome via taurine, which we found to be increased as a consequence of enhanced bile acid deconjugation. In support of this idea, it was recently shown that taurine triggers intestinal secretion of defensins in an inflammasome-dependent manner39. In any event, this mechanism would also depend on increased bile acid synthesis in the liver. From a general perspective, we provided evidence for the concept that the processing of dietary ingredients, such as choles￾terol, by the host but not the diet per se determines the composition of the gut microbiome. Recently, an induction of Cyp7b1 and other genes related to bile acid metabolism was described in WT mice in response to cold exposure11. The induction of Cyp7b1 after cold exposure or after β3-adrenergic receptor agonism observed in our study suggest a BAT–liver crosstalk that regulates hepatic gene expression and thermogenic capacity40,41. In addition, because germ-free mice have higher hepatic Cyp7b1 lev￾els42,43, factors released by the gut microbiota may either directly, or indirectly via BAT activation, regulate Cyp7b1 expression. Here, using Cyp7b1−/− mice, we demonstrated the relevance of hepatic Cyp7b1 induction as the main driver for a cold-induced increase in bile acid synthesis, which increased energy expenditure. This proc￾ess of enhanced cholesterol catabolism was triggered by accelerated lipoprotein processing in cold-activated BAT, which stimulated cholesterol flux toward the liver. Furthermore, we demonstrated that unlike CYP7A1, the rate-limiting enzyme of the classical pathway18, CYP7B1 was not suppressed by pharmacological FXR agonism. Thus, CYP7B1 is not subject to classical negative feedback control regulated by elevated bile acid levels. This is of special note, as it enables the body to maintain cholesterol homeostasis despite massively increased cholesterol conversion to bile acids in the cold. Although the exact mechanism underlying the induction of Cyp7b1 in cold-exposed mice remains to be elucidated, this is the first report demonstrat￾ing a physiological regulation of the alternative bile acid synthesis pathway. Notably, CYP7B1 expression was reduced in subjects with type 2 diabetes, suggesting a role of the alternative pathway for meta￾bolic homeostasis in humans. However, previous studies indicated higher levels of CYP8B1-derived bile acids in the plasma of humans with type 2 diabetes44,45. It will be of interest to dissect the bile acid synthesis pathways in intervention studies, including BAT activation by cold, and compare them in healthy individuals and those with diabetes. Taken together, our findings imply that increased bile acid levels contribute to beneficial effects of beige and brown adipocytes on obesity-associated comorbidities. Mechanistically, elevated bile acid levels regulate thermogenic responses, which have been described to occur in the BAT of mice and humans, likely via TGR5 (refs. 11,22,24,25). Furthermore, bile acids can also directly promote the thermogenic activity of skeletal muscle24, which may explain the increased glucose uptake observed in both the BAT and muscle of sub￾jects with type 2 diabetes after cold acclimation46. Future studies are warranted to address the therapeutically relevant question of whether beneficial cold-mediated metabolic effects are mediated directly by bile acids acting on TGR5 and other receptors or indirectly by com￾positional changes in the gut microbiome. Methods Methods, including statements of data availability and any associated accession codes and references, are available in the online version of the paper. Note: Any Supplementary Information and Source Data files are available in the online version of the paper. Acknowledgments We thank S. Ehret, B. Henkel, A. Kuhl and E.-M. Azizi for excellent technical assistance, P. Dawson (Emory University School of Medicine) for the ASBT-specific polyclonal antibody, and J. Nedergaard and B. Cannon (Wenner-Gren Institute, Stockholm University) for the UCP1-spcific polyclonal antibody. This work was supported by grants funded by the Deutsche Forschungsgemeinschaft (SFB841, “Liver inflammation: infection, immune regulation und consequences” (J.H. and M.D.); KFO306, “Primary sclerosing cholangitis (J.H. and to A.F.)), a Heisenberg Professorship (HE3645/7-1 (J.H.) and DA1063/3-2 (M.D.)), an EFSD award supported by Merck Sharp Dohme (MSD) (J.H.), the EU FP7 project RESOLVE FP7-HEALTH-2012-305707 (J.H.), a University Medical Center Hamburg– Eppendorf MD/PhD fellowship (C.S.) and the US National Institutes of Health grant HL087564 (P.W.S.). AUTHOR CONTRIBUTIONS A.W., C.J., L.S. and J.H. designed the study, were involved in all aspects of the experiments and wrote the manuscript; M.C.R., F.-A.H. and A.F. were responsible for the microbiome analysis; M.B., N.S., M.H., I.E., C.S. and C.M. were involved in the metabolic studies; M.F., M.D., A.F., C.K. und P.W.S. were involved in study design; and all authors read and commented on the manuscript