If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Division of Nutritional Sciences, University of Illinois Urbana-Champaign, Urbana, IL, USACarle Illinois College of Medicine, University of Illinois Urbana-Champaign, Urbana, IL, USA
Beckman Institute for Advanced Science and Technology, University of Illinois Urbana-Champaign, Urbana, IL, USACarle Illinois College of Medicine, University of Illinois Urbana-Champaign, Urbana, IL, USADepartment of Bioengineering, University of Illinois Urbana-Champaign, Urbana, IL, USADepartment of Electrical and Computer Engineering, University of Illinois Urbana-Champaign, Urbana, IL, USACancer Center at Illinois, University of Illinois Urbana-Champaign, Urbana, IL, USA
Corresponding author. , Department of Molecular and Integrative Physiology, University of Illinois Urbana-Champaign, Urbana, IL 61801. Tel: 217 300 7905; Fax: 217 244 5858.
Department of Molecular and Integrative Physiology, University of Illinois Urbana-Champaign, Urbana, IL, USABeckman Institute for Advanced Science and Technology, University of Illinois Urbana-Champaign, Urbana, IL, USADivision of Nutritional Sciences, University of Illinois Urbana-Champaign, Urbana, IL, USACancer Center at Illinois, University of Illinois Urbana-Champaign, Urbana, IL, USA
Cholestasis impairs brown fat mitochondrial function and thermogenesis.
•
Under thermoneutral housing, control brown fat mimics the cholestatic brown fat.
•
UCP1 activation restores expression of thermogenic genes reduced by bile acid excess.
Abstract
Background & Aims
Although fat loss is observed in patients with cholestasis, how chronically elevated bile acids (BAs) impact white and brown fat depots remains obscure.
Methods
To determine the direct effect of pathological levels of BAs on lipid accumulation and mitochondrial function, primary white and brown adipocyte cultures along with fat depots from two separate mouse models of cholestatic liver diseases - (i) genetic deletion of farnesoid X receptor; small heterodimer (DKO) and (ii) injury by 3,5-diethoxycarbonyl-1,4-dihydrocollidine (DDC) were used.
Results
As expected, cholestatic mice accumulate high systemic BA levels and exhibit fat loss. Here, we demonstrate that chronic exposure to pathological BA levels results in mitochondrial dysfunction and defective thermogenesis. Consistently, both DKO and DDC-fed mice exhibit lower body temperature. Importantly, thermoneutral (30 °C) housing of the cholestatic DKO mice rescues the decrease in brown fat mass, and the expression of genes responsible for lipogenesis and regulation of mitochondrial function. To overcome systemic effects, primary adipocyte cultures were treated with pathological BA concentrations. Mitochondrial permeability and respiration analysis revealed that BA overload is sufficient to reduce mitochondrial function in primary adipocytes, which is not due to cytotoxicity. Instead, we found robust reduction in Ucp1, Prdm16, and Dio2 transcript in brown adipocytes upon treatment with CDCA while TCA led to the suppression of Dio2 transcript. This BA-mediated decrease in transcripts was alleviated by pharmacological activation of uncoupling protein 1 (Ucp1).
Conclusions
High concentrations of BAs cause defective thermogenesis by reducing the expression of crucial regulators of mitochondrial function, including UCP1, which may explain the clinical features of hypothermia and fat loss observed in patients with cholestatic liver diseases.
Lay summary
We uncover a detrimental effect of chronic bile acid (BA) overload on adipose mitochondrial function. Pathological concentration of different BAs reduces expression of distinct genes involved in energy expenditure, which is alleviated by pharmacological UCP1 activation.
S.A.B. reports consulting fees and stock from LiveBx, LLC. M.B. reports Diversity Travel Award from the Obesity Society to attend Obesity Week 2022. The other authors have declared that no conflict of interest exists.
Financial support
This work was supported by start-up funds from the University of Illinois Urbana-Champaign (S.A.), R01 DK113080 from NIDDK (S.A.), R01 DK130317 (S.A.), and Cancer Center at Illinois planning grant and seed grant (S.A), and NIH/NCI R01CA241618 (to institution (S.A.B)).
Authors contributions
W.Z. P.V. and S.A. conceived and designed research; B.M.R. M.T.G. and M.B. supplied human adipose samples; W.Z. P.V. C.Z. Y.L. R.R. and S.A. performed experiments; C.Z. and S.A.B. were responsible for Raman spectroscopy analysis, W.Z. P.V. C.Z. Y.L. R.R. analyzed data; W.Z. P.V. C.Z. Y.L. R.R. and S.A. interpreted data; W.Z. and S.A. drafted the manuscript; all authors were involved in editing and revising the manuscript, and had final approval of the submitted and published versions.
Data availability
All data are in the main text and the supplementary information.
Introduction
Patients with liver diseases display increased systemic bile acid (BA) levels up to several hundred micromolar (∼300 μM) than the normal concentration.
display loss of body weight and fat mass. This indicates that serum BAs are linked with energy metabolism. Previous studies reveal that treating with low BA concentrations can promote heat production in the brown adipose tissue (BAT) and increase energy expenditure by activating membrane G protein-coupled receptor TGR5.
in a dose-dependent manner, we investigated if chronically elevated levels of BAs impact mitochondrial function in the adipose tissue. We examined both brown adipose tissue (BAT) which is linked to fat burning and is rich in mitochondria, and the white adipose tissue (WAT) that is primarily responsible for fat storage.
Next, we examined the adipose histology and mitochondrial structure with electron microscopy in the genetic mouse model for juvenile onset cholestasis (farnesoid X receptor (Fxr); small heterodimer (Shp) double knockout (DKO)).
Small heterodimer partner deletion prevents hepatic steatosis and when combined with farnesoid X receptor loss protects against type 2 diabetes in mice.
Then, we determined if we could overcome the brown fat defect by housing the mice in thermoneutral conditions. To validate this finding, we investigated if another model of cholestasis, which is caused by chronic exposure of wild-type (WT) mice to 3,5-diethoxycarbonyl-1,4-dihydrocollidine (DDC) diet
also led to compromised adipose mitochondrial function. Adipose depots from both models of cholestasis were analyzed for mitochondrial respiration, enzyme activity, mitochondrial DNA copy number, and gene expression. We also challenged the WT and DKO mice with a high fat diet to test if elevated BAs contribute to fat loss as well as thermogenic defect during obesity. We then investigated a cell autonomous role for BAs in the adipose milieu using differentiated primary adipocyte cultures from both fat depots in vitro, and analyzed mitochondrial membrane potential, respiration, and gene expression in the presence and absence of BA overload. Finally, we tested if a pharmacological activator of UCP1 can alleviate thermogenic gene expression pattern in the adipocytes.
Materials and methods
Human adipose tissue
Human perigastric white adipose tissue samples were obtained from tissue normally discarded from obese patients undergoing Roux-en-Y gastric bypass surgery.
All procedures were approved by the Institutional Review Board (IRB) at Carle Foundation Hospital and the University of Illinois Urban-Champaign (IRB protocol # 14092). Informed consent was obtained from the subjects and the privacy rights of the subjects were observed.
Animals
To induce short-term BA overload, 8- to 10-week-old male C57BL/6 WT mice were fed with 1% cholic acid (CA)-supplemented (Envigo) or normal chow (Envigo) diet for 5 days.
Male DKO and WT mice at 8- to 10-week-old were used. These mice were bred and maintained on a 12:12 light/dark cycle with ad libitum access to tap water and a normal chow diet in a climate‐controlled (23 °C) animal facility at the University of Illinois Urbana‐Champaign. At 8-10 weeks, mice were fed a normal chow or 45% high-fat diet (Envigo) for 8 weeks to mimic obesogenic conditions and housed either at room temperature (RT, 23 °C) or at thermoneutrality (TN, 30 °C) to blunt brown fat thermogenic activity.
To induce chronic cholestatic liver disease, 8- to 10-week-old male WT mice were fed with 0.1% DDC-supplemented (Envigo) or normal chow diet for 6 weeks.
DKO and WT mice were weighed weekly. After 8-week chow or 45% high-fat diet feeding, a subset of the mice was used for monitoring core body temperature fluctuation using a Comprehensive Laboratory Animal Monitoring System (CLAMS) (Oxymax, Columbus Instruments). Briefly, animals were surgically implanted with a transmitter in the abdominal cavity and acclimated to the CLAMS cages. Fluctuations in the body temperature were recorded over the subsequent 24-hour period. The rest of mice were sacrificed at the end of the experimental regimen. Interscapular brown, inguinal and gonadal white adipose tissues were collected for primary preadipocyte culture and analysis of histology, gene expression, protein levels, the degree of lipid unsaturation, mitochondrial respiratory enzyme activity, and isolated mitochondrial respiration.
DDC and chow-fed WT mice were weighed weekly. After 6 weeks, a subset of mice was used for indirect calorimetry using the CLAMS as previously described.
Small heterodimer partner deletion prevents hepatic steatosis and when combined with farnesoid X receptor loss protects against type 2 diabetes in mice.
The rest of mice were sacrificed at the end of the experimental regimen. Serum, liver, interscapular brown and gonadal white adipose tissues were collected for analysis of total BA levels, histology, gene expression, and isolated mitochondria respiration assay. All experiments were performed following the National Institutes of Health guidelines for the care and use of laboratory animals, and all procedures were approved by the Institutional Animal Care and Use Committee at the University of Illinois Urbana‐Champaign.
Statistical analysis
Data were expressed as means ± SEM. Statistical analyses were performed using GraphPad Prism 9 software. Differences between two groups were analyzed using Student’s t test, and multiple group comparisons were analyzed using a one-way or two-way ANOVA with a Fisher’s LSD post hoc test. P < 0.05 was considered statistically significant.
For further details regarding the materials and methods used, please refer to the CTAT table and supplementary information.
Results
Chronic BA excess in DKO mice results in compromised brown adipose function
Previously, a 0.5% CA diet had been shown to promote the expression of thermogenic genes Pgc1a and Dio2, indicating the beneficial effect of low dose of BAs.
and found that we were able to recapitulate this increase in Dio2 and Pgc1a in brown adipose tissue (Supplementary Fig. 1). We then examined the DKO mouse model of juvenile onset cholestasis
Small heterodimer partner deletion prevents hepatic steatosis and when combined with farnesoid X receptor loss protects against type 2 diabetes in mice.
to study the effect of chronic BA overload on the adipose tissue. As previously shown, DKO mice display excessive BA concentrations in the serum (Supplementary Fig. 2A),
Small heterodimer partner deletion prevents hepatic steatosis and when combined with farnesoid X receptor loss protects against type 2 diabetes in mice.
Even under chow diet, these mice displayed reduced white and brown fat mass (Figure 1A) with smaller adipocyte size (Figure 1B and Supplementary Fig. 2B) compared to WT mice. These findings correlated well with reduced expression of lipogenic genes in the WAT (C/Ebpa, Dgat2) (Figure 1C) and BAT (C/Ebpa, Srebp1c, Dgat1, Dgat2) (Figure 1D), suggesting reduced lipogenesis in the DKO adipose tissues. Although we did find increased mRNA levels of SCD1 (Figure 1C), a rate-limiting enzyme for the synthesis of unsaturated fatty acids,
in the DKO WAT, no alteration was observed in the degrees of unsaturation as measured with Raman spectroscopy compared to WT mice (Supplementary Fig. 2C). Further, we investigated and found that the response to β adrenergic stimulation with isoproterenol was dampened in DKO compared to WT mice (Figure 1E). We also found lower transcript levels of Hsl and Atgl in the WAT and BAT, respectively (Figure 1F and G), which corroborates lower stimulated lipolysis in DKO mice.
Figure 1Cholestatic DKO mice display decreased fat accumulation. DKO and WT mice were fed with chow and housed at RT (A-D) Fat mass/body weight ratios (n=8 mice) (A), representative images of H&E-stained WAT and BAT sections (scale bar: 200 μm; n=6-7 mice) (B), and mRNA levels of genes related to lipogenesis in the WAT (C) and BAT (D) from DKO and WT mice (n=5-7 mice) (E-G) Levels of serum free fatty acids (FFAs) under basal and isoproterenol-stimulated conditions (n=5-6 mice) (E), and mRNA levels of lipolytic genes in the WAT (F) and BAT (G) from DKO and WT mice (n=5-7 mice). Data are represented as mean ± SEM. Differences between two groups were analyzed using Student’s t test, and multiple group comparisons were analyzed using a two-way ANOVA with a Fisher’s LSD post hoc test. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.
we examined gene expression and mitochondrial activity, and continually monitored the body temperature of DKO and WT mice. Intriguingly, DKO mice exhibited decreased expression of thermogenic genes Prdm16 and Ucp1 (Figure 2A). We validated the reduction in UCP1 protein levels (Figure 2B) and activities of citrate synthase as well as oxidative phosphorylation (OXPHOS) enzyme complex III and IV albeit an increase in complex I (Figure 2C). Consistently, DKO BAT displayed inhibited mitochondrial respiration including state III, coupled, and uncoupled/leak respiration (Figure 2D and E). Furthermore, the mtDNA copy number, a surrogate indicator of mitochondrial numbers was comparable between DKO and WT mice (Figure 2F), suggesting that the DKO mice do not have less mitochondria. We also tested if BA overload in DKO caused cell death in the brown fat depots by TUNEL staining and found no positive cells (Supplementary Fig. 2D), indicating the absence of apoptosis. Instead, coherent with poor mitochondrial function, DKO mice exhibited lower body temperature than WT mice during the daytime; however, the activity-induced increase in body temperature at night was unaffected (Figure 2G).
Figure 2Cholestatic DKO mice exhibit BAT mitochondrial dysfunction. DKO and WT mice were fed with chow or HFD for 8 weeks and housed at RT (A-E) mRNA levels of thermogenic genes (n=7 mice) (A), protein levels of UCP1 (n=5 mice) (B), and the activities of oxidative phosphorylation (OXPHOS) enzyme complexes (n=3 mice) (C), traces (D) and quantification (E) of oxygen consumption rate (OCR) of isolated mitochondria of the BAT from WT and DKO mice upon chow. OCR at state III and in the presence of oligomycin (Oligo) or rotenone & antimycin A (Rot/AA) (n=9 wells per group isolated from BAT of 5 mice) (F) Mitochondrial DNA (mtDNA) copy number of the BAT from WT and DKO mice upon chow (n=5 mice) (G) Body temperature of DKO and WT mice for 24 h upon chow (n=7 mice) (H) Representative electron microscopy images (magnification: 10,000x) from DKO and WT mice upon HFD (n=3-4 mice) (I–K) mRNA levels of thermogenic genes (n=6-8 mice) (I), protein levels of UCP1 (n=5 mice) (J), and mtDNA copy number (n=5 mice) (K) of BAT from DKO and WT mice at RT upon HFD. Data are represented as mean ± SEM (A, C–F, I and K) or mean (G). Differences between two groups were analyzed using Student’s t test, and multiple group comparisons were analyzed using a two-way ANOVA with a Fisher’s LSD post hoc test. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.
Next, we examined these adipose depots after challenging DKO and WT mice for 8 weeks with a 45% high fat diet (HFD). DKO mice maintained lower fat mass (Supplementary Fig. 2E) and smaller adipocyte size (Supplementary Fig. 2F and G) in both the fat depots compared to WT mice upon HFD challenge. This result is in line with significant reduction in the expression of key lipogenic genes Pparg, C/Ebpa, Srebp1c, Acc1, Fasn, Scd1, Dgat1, and Dgat2 in the DKO BAT (Supplementary Fig. 2I). However, lipogenic gene expression profile was not different between the two genotypes in the WAT except for lower levels of Dgat2 (Supplementary Fig. 2H). DKO WAT displayed increase, while DKO BAT showed decrease in transcript levels of Atgl compared to WT mice (Supplementary Fig. 2J). Although we did not find overt difference in mitochondrial ultrastructure using electron microscopy (Figure 2H), we discovered lower expression of thermogenic genes Prdm16, Pgc1a, Dio2 and Ucp1 (Figure 2I) and a reduction in UCP1 protein levels (Figure 2J) in the DKO versus WT mice under HFD. Also, mtDNA copy number was lower in the DKO BAT than WT BAT (Figure 2K). We also confirmed these effects in HFD are not due to cell death as we observed negligible TUNEL staining of BAT and WAT (Supplementary Fig. 2K). These results suggest that BA excess can impair expression of lipogenic genes and decrease thermogenic function in the adipocytes.
Thermoneutral housing is sufficient to increase brown fat mass during cholestasis
To confirm the brown adipose mitochondrial dysfunction in cholestatic DKO mice, we housed both WT and DKO mice at thermoneutrality (TN, 30 °C), which excludes the thermogenic effect of the BAT.
The decrease in mRNA levels of Prdm16 and Ucp1 noted in the DKO BAT than the WT BAT at room temperature (RT) (Figure 2A) was abolished at TN (Figure 3A). Also, the reduction in UCP1 protein levels were pronounced at RT compare to TN in DKO BAT (Figure 2B and 3B). Despite a slightly smaller adipocyte size (Figure 3D and Supplementary Fig. 3A), the brown fat mass at TN was comparable between WT and DKO mice (Figure 3C), which correlated well with the similar transcription profile of key lipogenic and lipase genes (Figure 3F and G). These results indicate that the ineffective brown fat function may in part protect the DKO mice against obesity, which is in line with previous studies that adipose-specific mitochondrial dysregulation causes fat loss.
Adipose tissue mitochondrial dysfunction triggers a lipodystrophic syndrome with insulin resistance, hepatosteatosis, and cardiovascular complications.
Although, thermoneutral housing enhanced C/Ebpa transcript expression in DKO WAT to that of WT and altered other genes in the lipogenic pathways (Figure 1C and 3E), DKO mice still exhibited reduced white fat mass (Figure 3C) and smaller adipocyte size (Figure 3D and Supplementary Fig. 3A).
Figure 3Thermoneutral housing replicates the brown fat phenotype of cholestatic DKO in the WT mice. DKO and WT mice were fed with chow or HFD for 8 weeks and housed at TN (A-B) mRNA levels of thermogenic genes (n=7-10 mice) (A) and protein levels of UCP1 (n=5 mice) (B) of the BAT from WT and DKO mice upon chow (C–H) Fat mass/body weight ratios (n=7-11 mice) (C), representative images of H&E-stained WAT and BAT sections (scale bar: 200 μm; n=7-11) (D), and mRNA levels of lipogenic (E-F) and lipolytic (G-H) genes in the WAT and BAT from DKO and WT mice upon chow (n=7-11 mice) (I-J) Representative electron microscopy images (I) (abnormal mitochondria with irregular shape (asterisks), loss of cristae (empty arrowheads) or myelin figures (arrowheads), magnification: 10,000x), and quantification of the structure of BAT (J) from DKO and WT mice upon HFD (n=4-6 mice). Data are represented as mean ± SEM. Student’s t test. *P < 0.05, **P < 0.01, ****P < 0.0001.
Next, we challenged the DKO, and WT mice housed at 30 °C with a HFD for 8 weeks. This double hit, led to a dramatic decrease in the ratio of normal mitochondria in the DKO BAT with the majority of them revealing abnormal ultrastructure with either irregular shape, loss of cristae, or presence of myelin figures (Figure 3H and I), all of which indicate mitochondrial damage or degeneration.
Under thermoneutral conditions, HFD-fed DKO mice displayed similar levels of Prdm16 and Pgc1a but lower expression of Dio2 and Ucp1 in the BAT compared to HFD-fed WT mice when housed at TN (Figure 3J). Nonetheless, at TN, DKO mice gained a similar percentage of body weight as WT mice in response to HFD (Supplementary Fig. 3B), unlike poor weight gain seen in DKO mice under normal housing condition. Notably, DKO and WT mice exhibited comparable BAT mass at TN, but WAT mass was still lower in DKO than the WT mice (Supplementary Fig. 3C). The decrease in WAT adipocyte size was prominent and maintained in the HFD-fed DKO mice housed at TN (Supplementary Fig. 3D and E), despite induced expression of lipogenic genes in the DKO compared to the WT mice (Supplementary Fig. 3F). On the other hand, thermoneutral housing was sufficient to increase brown adipocyte size (Supplementary Fig. 3D and E) and led to comparable expression of lipogenic genes Pparg, C/Ebpa, Srebp1c, and Dgat1 in the BAT of HFD-fed DKO vs HFD-fed WT mice (Supplementary Fig. 3G). DKO WAT at TN showed similar expression of lipases as seen at RT (Supplementary Fig. 2J and 3H), while the reductions in Atgl and Hsl mRNA levels in DKO BAT were abrogated at TN (Supplementary Fig. 2J and 3H). These findings indicate that WT and DKO BAT are comparable to each other under TN condition.
DDC-fed mice mimic the brown adipose dysfunction phenotypes in DKO mice
FXR and SHP are implicated in lipid metabolism, and particularly, FXR has been shown to regulate adipose tissue lipid accumulation.
The farnesoid X receptor regulates adipocyte differentiation and function by promoting peroxisome proliferator-activated receptor-gamma and interfering with the Wnt/beta-catenin pathways.
Therefore, to overcome the caveat of Fxr and Shp deletion in DKO mice, we examined a chronic DDC-induced model of cholestasis in WT mice with intact FXR and SHP expression. As expected, DDC diet led to elevated levels of circulating BAs (Supplementary Fig. 4A), liver injury indicated by histological alterations (Supplementary Fig. 4B) and raised serum concentrations of liver enzymes aspartate transaminase (AST) and alanine transaminase (ALT) (Supplementary Fig. 4C) in WT mice. We also observed that DDC challenge led to a lower body weight (Figure 4A) as anticipated, which was accompanied by the loss of WAT mass (Figure 4B), and a decrease in brown fat mass (Figure 4B) and adipocyte size (Figure 4C and Supplementary Fig. 4D). The weight loss with DDC is drastic at 2 weeks and then stabilizes, therefore we examined daily food intake using a Comprehensive Laboratory Animal Monitoring System (CLAMS) on week 6 of the diet. Food intake was comparable between DDC- and chow-fed mice (Supplementary Fig. 4E). Energy expenditure during the day remained unaltered, but the active nighttime energy expenditure was increased despite reduced physical activity in DDC-fed cholestatic mice (Supplementary Fig. 4F-G).
Figure 4Cholestatic DDC-fed mice show attenuated BAT mitochondrial function and fat loss (A-G) Body weight (n=11-12 mice) (A), white and brown fat mass (n=11-12 mice) (B), representative images of H&E-stained BAT sections (scale bar: 200 μm) (n=8-10 mice) (C), transcript levels of lipogenic (D), lipolytic (E), and thermogenic (F) genes in the BAT of DDC- and chow-fed WT mice (n=7-9 mice) (G-H) Traces (G) and quantification (H) of oxygen consumption rate (OCR) of BAT isolated mitochondria from DDC- and chow-fed WT mice. OCR at state III and in the presence of oligomycin (Oligo) or rotenone & antimycin A (Rot/AA) (n=25 wells per group isolated from BAT of 9 mice) (I) Relative mtDNA content of BAT from DDC- and chow-fed WT mice (n=5 mice) (J) Body surface temperature of DDC- and chow-fed WT mice during daytime (n=11-12 mice). Data are represented as mean ± SEM. Differences between two groups were analyzed using Student’s t test, and multiple group comparisons were analyzed using a two-way ANOVA with a Fisher’s LSD post hoc test. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.
Upon further analysis, we found that brown fat from DDC-fed mice phenocopied many aspects of the BAT from DKO mice, including the reduction in lipogenic genes C/Ebpa and Dgat2 along with Pparg, Fasn, and Scd1, which was unique to the DDC diet (Figure 1D and 4D). Remarkably, brown fat from DDC-fed WT mice also recapitulated the reductions in thermogenic genes Prdm16 and Ucp1 (Figure 2A and 4F). We also found reduced mitochondrial respiration, including state III and uncoupled/leak respiration (Figure 4G and H), and lower mtDNA copy number (Figure 4I) in DDC-fed WT BAT. In line with these changes, DDC-fed mice exhibited lower body temperature during the light phase like that of the cholestatic DKO mice (Figure 2G and 4J). Additionally, we confirmed that the loss in thermogenic genes and mitochondrial defect is not secondary to cell death as measured by TUNEL staining (Supplementary Fig. 4H). These results demonstrate that chronic cholestasis causes defects in brown adipose function leading to poor thermogenesis and fat loss.
Pathological BA concentrations are sufficient to cause mitochondrial dysfunction in adipocytes
To determine the cell autonomous effect of BAs, we treated differentiated primary adipocytes derived from either WAT or BAT with pathological concentrations of primary BAs chenodeoxycholic acid (CDCA), taurochenodeoxycholic acid (TCDCA) or taurocholic acid TCA.
Excess CDCA, TCDCA, and TCA resulted in a general decline in mitochondrial membrane potential (Figure 5A and B) and respiration (Figure 5C and D, Supplementary Fig. 5A and B) without affecting adipocyte viability (Supplementary Fig. 5C). This result indicates that BAs regulate adipocyte mitochondrial function in a concentration-dependent manner. Elevated BAs lowered FCCP-induced maximal respiration compared to vehicle-treated adipocytes (Figure 5C and D, Supplementary Fig. 5A and B). Thus, BAs can reduce the spare respiratory capacity of the mitochondria, which is important to keep up with ATP demands of a cell. Further, TCDCA or TCA treatment resulted in lower basal and uncoupled/leak respiration (Figure 5D, Supplementary Fig. 5A and B), indicative of defective uncoupling in brown adipocytes.
Figure 5Pathological levels of BAs result in adipocyte mitochondrial dysfunction in vitro. Primary adipocyte cultures were isolated from WAT and BAT of 5-6 mice and n=4-5 wells per treatment was used for analysis (A-B) Mitochondrial membrane potential of white (A) and brown (B) adipocytes upon different concentrations of CDCA, TCDCA, or TCA treatment for 24 h (C-D) Traces of oxygen consumption rate (OCR) of CDCA/TCDCA/TCA-treated white (C) and brown (D) adipocytes. OCR at basal and in the presence of oligomycin (Oligo), FCCP, or rotenone & antimycin A (Rot/AA). Data are represented as mean ± SEM. Differences between multiple groups were analyzed using a one-way or two-way ANOVA with a Fisher’s LSD post hoc test. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.
To probe how BAs were altering mitochondria in adipocytes, we examined the literature to identify if there are putative genes that could transport BAs into adipocytes. We identified a subset of four possible transporters namely Slco1a6, Slco1b2, Slc51a, and Slc51b that may transport BAs from the systemic circulation into the adipocytes. We validated their transcript expression in differentiated and undifferentiated primary adipocytes derived from either brown or white adipose depots. In white adipocytes, expression profile of these transporters was maintained irrespective of the differentiation status except for a decrease in Slc51a post differentiation (Supplementary Fig. 6A). We also confirmed the expression of human orthologs of SLC51A and SLC51B in human white fat depot (Supplementary Fig. 6C). Intriguingly however, the expression of Slco1a6, Slco1b2, and Slc51b were all induced upon differentiation specifically in brown adipocyte cultures (Supplementary Fig. 6B), denoting the possibility of BA transport into mature brown adipocytes.
BA-induced suppression of thermogenic genes in brown adipocytes can be rescued by the activation of UCP1
In addition to mitochondrial activity, we also examined the expression of genes regulating thermogenesis and found that CDCA overload drastically reduced the expression of Ucp1, Prdm16, and Dio2 transcripts in brown adipocytes (Figure 6B), whereas in white adipocytes we saw reductions in Pgc1a, Dio2, and Cpt1a transcript levels (Supplementary Fig. 7). Since BAs are endogenous ligands for FXR and TGR5, we examined if pharmacological activation of FXR
mimic the BA-mediated reduction of thermogenic genes. We found FXR agonist GW4064 suppressed Prdm16 but induced Dio2 (Supplementary Fig. 8). On the other hand, TGR5 activation induced the expression of Ucp1, Prdm16, Pgc1a, and Dio2 as expected (Supplementary Fig. 8). Finally, GW4064 in combination with INT-777 was able to dampen the induction of many of the thermogenic genes (Supplementary Fig. 8). Overall, neither GW4064 nor INT-777 was able to fully recapitulate the CDCA-mediated reduction of all the analyzed thermogenic genes.
Figure 6UCP1 activation rescues BA overload-induced reductions in the expression of thermogenic genes in brown adipocytes. Primary adipocyte cultures were isolated from WAT and BAT of 6 mice and n=3 wells per treatment was used for analysis (A) Ucp1 mRNA levels in brown adipocytes upon TTNPB (1 nM) treatment for 48 h (B–C) mRNA levels of genes responsible for mitochondrial function in differentiated brown adipocytes with or without TTNPB (1 nM) administration for 24 h followed by another 24-hour treatment of CDCA (100 μM) (B) or TCA (300 μM) (C). Data are represented as mean ± SEM. Differences between two groups were analyzed using Student’s t test, and multiple group comparisons were analyzed using a one-way ANOVA with a Fisher’s LSD post hoc test. *P < 0.05, ***P < 0.001, ****P < 0.0001.
Nonetheless, reduction of these genes was also noted in the cholestatic animal models-DKO and DDC-fed mice, highlighting the relevance of CDCA in mediating these defects. However, TCA overload reduced Dio2 gene expression only in brown adipocytes (Figure 6C and Supplementary Fig. 7). The varied responses to different BA species in adipocytes from both fat depots suggest that the composition of the BA pool can also influence the outcome of mitochondrial function distinctly. Finally we investigated if pharmacological activator of UCP1, TTNPB
can alleviate some of these defects in the brown adipocytes. As previously shown, TTNPB dramatically upregulated Ucp1 mRNA levels (Figure 6A). Of note, the TTNPB treatment was sufficient to alleviate CDCA-mediated decreases in the transcript levels of Prdm16 and Dio2, and TCA-induced reductions in the expression of Dio2 (Figure 6B and C). This result suggests that maintaining Ucp1 mRNA during cholestasis is beneficial to maintain the thermogenic gene profile of brown adipocyte function.
Discussion
Several liver diseases result in weight and fat loss,
In this study, we investigated the impact of chronic cholestatic disease on the adipose tissue function using two mouse models. We demonstrate that pathological BA concentrations are sufficient to cause mitochondrial dysfunction, poor thermoregulation, and fat loss.
Although a caveat of the DKO model of cholestasis is that it is a global knockout of Fxr and Shp, it captures an increase in BA mixture rather than an increase in a single type of BAs and mimics pediatric cholestasis.
To overcome this caveat and to tease apart the BA effect, we also performed in vivo studies using DDC-induced cholestasis mouse model and in vitro studies using primary adipocyte cultures from white and brown adipose with intact FXR and SHP signaling.
Consistent with previous findings that BAs can promote brown adipose activity by activating TGR5-cAMP-DIO2 signaling pathway,
and we observed that short-term administration of BAs induced Dio2 and Pgc1a transcript expression (Supplementary Fig. 1). But, during chronic cholestasis, BA concentrations can elevate to hundreds of micromolar, which in turn impairs BAT mitochondrial function leading to poor thermoregulation as observed in both DKO and DDC-fed cholestatic mice. This is surprising since we had previously found enhanced mitochondrial function in DKO skeletal muscle.
Small heterodimer partner deletion prevents hepatic steatosis and when combined with farnesoid X receptor loss protects against type 2 diabetes in mice.
These results demonstrate tissue-specific effects of BA overload such that skeletal muscle activity is increased but the brown fat mitochondrial function is compromised. Neither Fxr nor Shp deletion alters the basal BAT thermogenic gene expression,
indicating that this effect is secondary to BA overload. We also examined for cell death in the BAT and did not find the evidence for it (Supplementary Fig. 2D, 2K, and 4H). Importantly, heat generated by brown fat
is hampered in both cholestatic mouse models, and they display lower body temperature (Figure 2G and 4J).
To evaluate the direct consequence of BAs on adipocytes, we examined primary adipocyte cultures obtained from white or brown fat depot. Adipocytes treated with pathological BA concentrations as observed in clinical cholestasis
were viable (Supplementary Fig. 5C) but revealed poor mitochondrial function (Figure 5A-D, 6B–C, Supplementary Fig. 5A-B, and 7). Such impairments in mitochondrial respiration, increased mitochondrial permeability, and cellular injury have also been noted in hepatocytes
Agonistic activation of FXR or TGR5 did not recapitulate the suppression of thermogenic genes subsequent to high concentrations of CDCA (Supplementary Fig. 8). TGR5 activation induced them whereas FXR agonist reduced only Prdm16 transcript in brown adipocytes. However, cotreatment led to inhibition of TGR5 effects when FXR was activated, indicating a complex interplay of these two signaling maybe involved in brown adipocytes upon pathological BA exposure.
Of note, UCP1 suppression in cholestasis was conserved in mice and adipocyte cultures. Intriguingly, Ucp1 deficiency or mitochondrial dysfunction has been linked to fat loss, and resistant to diet-induced obesity.
Cold tolerance of UCP1-ablated mice: a skeletal muscle mitochondria switch toward lipid oxidation with marked UCP3 up-regulation not associated with increased basal, fatty acid- or ROS-induced uncoupling or enhanced GDP effects.
Uncoupling Protein 1 and Sarcolipin Are Required to Maintain Optimal Thermogenesis, and Loss of Both Systems Compromises Survival of Mice under Cold Stress.
when housed at 30 °C. Our findings reveal that DKO and DDC-fed cholestatic mice exhibit overlapping phenotypes with Ucp1 knockout mice, including reduced UCP1 levels, fat loss, impaired BAT mitochondrial function, and dysregulated thermogenesis (Table 1). Notably, thermoneutral housing reversed the reductions in body weight gain and brown fat mass of DKO mice (Fig. 3C, Supplementary Fig. 3B and C), highlighting the reduction of UCP1-mediated brown fat thermogenesis in cholestasis. More importantly, activating UCP1 is sufficient to recover the expression of thermogenic genes Prdm16 and/or Dio2 (Figure 6B and C). Our findings uncover that BA excess can lead to mitochondrial defect in the BAT and may explain these clinical presentations of fat loss
Small heterodimer partner deletion prevents hepatic steatosis and when combined with farnesoid X receptor loss protects against type 2 diabetes in mice.
Cold tolerance of UCP1-ablated mice: a skeletal muscle mitochondria switch toward lipid oxidation with marked UCP3 up-regulation not associated with increased basal, fatty acid- or ROS-induced uncoupling or enhanced GDP effects.
Uncoupling Protein 1 and Sarcolipin Are Required to Maintain Optimal Thermogenesis, and Loss of Both Systems Compromises Survival of Mice under Cold Stress.
Small heterodimer partner deletion prevents hepatic steatosis and when combined with farnesoid X receptor loss protects against type 2 diabetes in mice.
Cold tolerance of UCP1-ablated mice: a skeletal muscle mitochondria switch toward lipid oxidation with marked UCP3 up-regulation not associated with increased basal, fatty acid- or ROS-induced uncoupling or enhanced GDP effects.
The authors are grateful to Dr. Oludemilade Akinrotimi for her assistance with obtaining transmission electron microscopy samples for analysis, Mr. Shawn D’Souza for quantitative real-time PCR, and Dr. Lee-Jun Wong’s laboratory at Baylor College of Medicine for mitochondrial complex activity analysis. We also thank Ms. Lou Ann Miller from the Biological Electron Microscopy core at Materials Research Laboratory, the University of Illinois Urbana-Champaign for performing transmission electron microscopy. We acknowledge the Microscopy Suite at the Beckman Institute, the University of Illinois Urbana-Champaign for the access to the Raman micro-spectrometer used in the study.
Appendix A. Supplementary data
The following is/are the supplementary data to this article:
Small heterodimer partner deletion prevents hepatic steatosis and when combined with farnesoid X receptor loss protects against type 2 diabetes in mice.
Adipose tissue mitochondrial dysfunction triggers a lipodystrophic syndrome with insulin resistance, hepatosteatosis, and cardiovascular complications.
The farnesoid X receptor regulates adipocyte differentiation and function by promoting peroxisome proliferator-activated receptor-gamma and interfering with the Wnt/beta-catenin pathways.
Cold tolerance of UCP1-ablated mice: a skeletal muscle mitochondria switch toward lipid oxidation with marked UCP3 up-regulation not associated with increased basal, fatty acid- or ROS-induced uncoupling or enhanced GDP effects.
Uncoupling Protein 1 and Sarcolipin Are Required to Maintain Optimal Thermogenesis, and Loss of Both Systems Compromises Survival of Mice under Cold Stress.
00045-9&id=ga1.jpg){kind=link}
00045-9&id=gr1.jpg){kind=link}
00045-9&id=gr2.jpg){kind=link}
00045-9&id=gr3.jpg){kind=link}
00045-9&id=gr4.jpg){kind=link}
00045-9&id=gr5.jpg){kind=link}
00045-9&id=gr6.jpg){kind=link}