We performed global metabolic profiling to discover significantly altered metabolites across the histological spectra of NAFLD and demonstrated that circulating BA levels were significantly altered correlating with the histological severity of NAFLD. Targeted BA analysis to identify the characteristic alterations in both BA concentration and composition was stratified by obesity status to assess the independent effect of NAFLD severity on BA metabolism. Moreover, we stringently adjusted BA alterations for various covariates, including age, gender, insulin resistance and genetic risk variants, to exclude their confounding effects on the association between NAFLD severity and BA alteration. Then, we found that circulating primary BAs significantly increased in nonobese subjects with NASH irrespective of conjugation status, which was confirmed in the independent validation cohort. In addition, individual histological subcomponents, such as ballooning and lobular inflammation, were also positively correlated with primary BA levels in nonobese NAFLD subjects, which was consistent with previous findings in Caucasian NAFLD subjects (mean BMI, 31.9 kg/m2). A positive correlation between circulating conjugated primary BA level and the histological severity of NAFLD was also found in the obese group; however, the correlation was not as prominent as that in the nonobese group, which corroborated the recent data based on lean Caucasian NAFLD subjects.
The current study not only replicated the positive correlation between primary BA level and insulin resistance in nonobese NAFLD subjects found in the previous studies[10,33,34] but also confirmed an increase in both conjugated and unconjugated primary BAs in nonobese NASH patients irrespective of diabetes status. Such an independent association between circulating BA alterations and NASH, even after adjusting for metabolic and genetic risk factors, provides robust evidence that NASH is associated with changes in BA metabolism independent of metabolic derangements such as obesity and insulin resistance. The alteration in circulating BA level according to the histological severity of NAFLD was more prominent in nonobese NAFLD patients based on the current study as well as the previous study. In nonobese NAFLD patients, genetic predisposing factors rather than metabolic risk factors have been known to be more relevant to the pathogenesis of NAFLD.[35,36] In addition, the alteration in BA metabolism may also be important to understand the pathogenesis of NASH in nonobese NAFLD patients.
In healthy liver, BA metabolism is tightly regulated through enterohepatic circulation of BAs, and BA homeostasis is maintained through multiple negative feedback mechanisms, such as farnesoid X receptor (FXR)/small heterodimer partner (SHP) and FGF19/FGF receptor-4 (FGFR4) signalling.[37,38] A growing body of literature has reported that the levels of total and primary BAs are elevated in the serum, plasma, stool and liver of patients with NAFLD, implying an association between BA metabolism and the pathogenesis of NASH.[1,7] Both conjugated and unconjugated primary BA levels have been reported to be higher in NASH patients than in NAFL patients among Caucasian NAFLD subjects.[3,6] In particular, the ratio of conjugated-to-unconjugated primary BAs increased in NASH patients compared to NAFL patients. The alteration in the primary-to-secondary BA ratio was also affected by NAFLD severity.[3,6,7] Taurine- and glycine-metabolizing bacteria were increased in the gut of patients with NASH, which was mainly attributed to alterations in the gut microbiome. Primary BAs arriving to the intestine are glycine- or taurine-conjugated, and their deconjugation by bile salt hydrolase (BSH) is required for downstream modifications by 7-alpha-dehydroxylase or 7-alpha-hydroxysteroid dehydrogenase to produce secondary BAs. However, the association between NAFLD severity and the primary-to-secondary BA ratio has not been consistent among studies, and further human and mechanistic studies are warranted to reach a consensus.[6,7]
The alteration in the primary BA concentration and composition may be derived from abnormalities in BA metabolism, including BA synthesis, conjugation, secretion into the intestine and conversion into secondary BAs from primary BAs, suggesting impaired regulation of BA metabolism. In the current study, hepatic BA synthesis-related genes, such as CYP7A1 and CYP8B1, were upregulated in nonobese NASH patients compared to nonobese NAFL patients, indicating that a significant elevation in the total BA concentration in nonobese NASH patients compared to nonobese NAFL patients could be attributed to increased BA synthesis. Moreover, hepatic cholesterol metabolism-related genes including HMGCR, NPC1L1, SCARB1 and LDLR were also upregulated in nonobese NASH patients, suggesting an increase in hepatic cholesterol pool through biosynthesis or uptake of cholesterol. On the other hand, hepatic expression of BA metabolism-regulatory genes, such as FXR and SHP, and serum FGF19 levels which indicate intestinal FXR activity did not show any significant changes in NASH patients compared to NAFL patients, which is in line with previous findings suggesting suppressed BA signalling despite elevated production of BAs in NAFLD patients. BA metabolism is closely associated with FXR signalling, and BAs play an important role as endogenous ligands of FXR. In the current study, the increase in hepatic BA synthesis without consequent FXR activation found in nonobese NASH patients might suggest the failure of the metabolic compensatory mechanism to protect from metabolic stress-induced liver injury.
BA compositions may be adapted to increase the proportion of hydrophilic forms to reduce hepatotoxicity during the progression of NAFL to NASH.[3,6,39] CA and CDCA are not toxic to humans, and CA and UDCA have been reported to improve insulin sensitivity. Moreover, UDCA is known to reduce the cytotoxicity of the circulating BA pool. In the current study, obese NASH patients had an altered composition of primary BAs, that is, an increased proportion of conjugated primary BAs, which are the relatively less toxic but hydrophilic BA forms.[37,40] In addition, in patients with nonobese NASH, the concentration of UDCA, the most hydrophilic BA, significantly increased (Table S4), which suggests that the BA pool in NASH patients may be adapted to be less hepatotoxic through the reduction in hydrophobicity regardless of obesity status.
Recently, alterations in serum BA level have been also reported in cholestatic liver diseases such as primary sclerosing cholangitis (PSC) and primary biliary cholangitis (PBC). Total BA concentrations and primary-to-secondary BA ratios were increased both in PSC and PBC[42,43] patients, which was similar to the pattern of changes in BA found in NAFLD subjects. In cholestatic liver diseases, excessive intrahepatic accumulation of BAs plays a pivotal role in hepatic injury even at the early stage without apparent jaundice. In particular, the risk of accompanying inflammatory bowel disease or future hepatic decompensation from PSC may be predicted by BA profiles and clinical markers.[41,45] Indeed, the selective FXR agonist, obeticholic acid can modulate BA metabolism and improve disease severity in PBC[46,47] as well as NASH,[16,48] which implicates that altered BA metabolism may play a pivotal role in chronic liver diseases.
The current study had several limitations. First, the current study was cross-sectional in nature, which prevents the determination of causality in the association between circulating BA alteration and NAFLD severity. However, extensive evaluation of circulating BA profiles and stringent adjustment for metabolic derangements associated with NAFLD enabled us to identify the independent changes in BA metabolism between NASH and NAFL patients. Second, the finding of no prominent alterations in serum BA levels according to NAFLD severity in obese NAFLD subjects might not be generalizable to other ethnic populations. Nevertheless, given that recent data based on Caucasian NAFLD subjects showed differences in serum BA levels between lean and nonlean NAFLD subjects, there might also be some differences in BA metabolism with regard to NASH development between the nonobese and obese populations. Further studies in diverse populations need to be carried out to clarify this.
Qualitative alterations in the BA pool composition as well as quantitative changes in the total BA concentration according to obesity status might modulate the activity of FXR/TGR5 and impact glucose/lipid metabolism differentially. In nonobese patients with NASH, hepatic BA synthesis might increase the total BA pool, especially at the unconjugated primary BA level, and the dysregulated BA metabolism in the liver or gut might reflect increasing levels of circulating conjugated primary BAs (Figure 6). Overall, altered BA metabolism in the liver or gut may be related to the progression of NAFL to NASH via a different mechanism depending on obesity status. In addition, diverse BAs with different binding affinities and potencies for FXR/TGR5 activation may disproportionately affect the pathogenesis of NAFLD. Therefore, our findings have brought to light that disparities in BA profiles between nonobese and obese subjects according to NAFLD severity may help us better understand the pathogenesis of individual phenotypes of NAFLD.
ABCG5/8, ATP-binding cassette sub-family G member 5/8; ALT, alanine aminotransferase; AST, aspartate aminotransferase; BA, bile acid; BACS, bile acid-coenzyme A synthase; BATT, bile acid-CoA: amino acid N-acyltransferase; BMI, body mass index; CA, cholate; CDCA, chenodeoxycholate; CYP7A1, cholesterol 7α-hydroxylase; CYP8B1, sterol 12α-hydroxylase; DCA, deoxycholate; DM, diabetes mellitus; FGF19, fibroblast growth factor 19; FGFR4, FGF receptor-4; FXR, farnesoid X receptor; GCA, glycocholate; GCDCA, glycochenodeoxycholate; GDCA, glycodeoxycholate; HMGCR, 3-hydroxy-3-methylglutaryl-CoA reductase; HOMA-IR, homeostasis model assessment of insulin resistance; hs-CRP, high-sensitivity C-reactive protein; LDLR, low-density lipoprotein receptor; NAFLD, nonalcoholic fatty liver disease; NAS, NAFLD activity score; NASH, nonalcoholic steatohepatitis; NPC1L1, Niemann-Pick C1 like 1; NTCP, Na+-taurocholate cotransport peptide; Q TOF-MS, quadrupole time-of-flight mass spectrometry; SCARB1, scavenger receptor class B type I; SHP, small heterodimer partner; TCA, taurocholate; TCDCA, taurochenodeoxycholate; TDCA, taurodeoxycholate; TLCA, taurolithocholate; TQ-MS, triple quadrupole mass spectrometry; TUDCA, tauroursodeoxycholate; UDCA, ursodeoxycholate; UPLC, ultra-performance liquid chromatography; VAT, visceral adipose tissue.
This work was supported by the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI) funded by the Ministry of Health & Welfare, Republic of Korea (H I17C0912 and HI21C0538); Hanmi Pharmaceutical Co., Ltd; the Korea Basic Science Institute (C060200); and National Research Foundation of Korea (NRF) grant funded by the Korean Government (NRF-2019M3A9D5A01102796, NRF-2020R1A2C2007835, NRF-2021R1A2C2005820, and NRF-2021M3A9E4021818). The funding organization played no role in the design and conduct of the study; in the collection, management, analysis and interpretation of data; or in the preparation, review or approval of the manuscript.
Data Availability Statement
All data that support the findings of this study are available from the corresponding authors upon reasonable request.
Liver International. 2021;41(12):2892-2902. © 2021 Blackwell Publishing