Lipidomics and body fat composition analysis characterises specic differences in cholesterol metabolism and steatosis between hepatitis C virus genotypes 1 and 3

Lipids play essential roles in the hepatitis C virus (HCV) life cycle and patients with chronic HCV infection display disordered lipid metabolism which resolves following successful anti-viral therapy. It has been proposed that HCV genotype 3 (HCV-G3) infection is an independent risk factor for hepatocellular carcinoma and evidence suggests lipogenic proteins are involved in hepatocarcinogenesis. We aimed to characterise variation in host lipid metabolism between participants chronically infected with HCV genotype 1 (HCV-G1) and HCV-G3 to identify likely genotype-specic differences in lipid metabolism. aminotransferase


Abstract Background
Lipids play essential roles in the hepatitis C virus (HCV) life cycle and patients with chronic HCV infection display disordered lipid metabolism which resolves following successful anti-viral therapy. It has been proposed that HCV genotype 3 (HCV-G3) infection is an independent risk factor for hepatocellular carcinoma and evidence suggests lipogenic proteins are involved in hepatocarcinogenesis.

Aims
We aimed to characterise variation in host lipid metabolism between participants chronically infected with HCV genotype 1 (HCV-G1) and HCV-G3 to identify likely genotype-speci c differences in lipid metabolism.

Methods
We combined several lipidomic approaches: analysis was performed between participants infected with HCV-G1 and HCV-G3, both in the fasting and non-fasting states, and after sustained virological response (SVR) to treatment. Sera were obtained from 112 fasting patients (25% with cirrhosis). Serum lipids were measured using standard enzymatic methods. Lathosterol and desmosterol were measured by gaschromatography mass spectrometry (MS). For further metabolic insight on lipid metabolism, ultraperformance liquid chromatography MS was performed on all samples. A subgroup of 13 participants had whole body fat distribution determined using in vivo magnetic resonance imaging and spectroscopy. A second cohort of (non-fasting) sera were obtained from HCV Research UK for comparative analyses: 150 treatment naïve patients and 100 non-viraemic patients post-SVR.

Results
HCV-G3 patients had signi cantly decreased serum apoB, non-HDL cholesterol concentrations, and more hepatic steatosis than those with HCV-G1. HCV-G3 patients also had signi cantly decreased serum levels of lathosterol, without signi cant reductions in desmosterol. Lipidomic analysis showed lipid species associated with reverse cholesterol transport pathway in HCV-G3.

Conclusions
We demonstrated that compared to HCV-G1, HCV-G3 infection is characterised by low LDL cholesterol levels, with preferential suppression of cholesterol synthesis via lathosterol, associated with increasing hepatic steatosis. The genotype-speci c lipid disturbances may shed light on genotypic variations in liver disease progression and promotion of hepatocellular cancer in HCV-G3.

Background And Aims
The life cycle of the hepatitis C virus (HCV) is interwoven with lipids at both the hepatocellular stages of virus entry, replication, and assembly (reviewed in (1)) and in the circulation with the formation of complex lipoviral particles (LVP) (reviewed in (2)). Chronic HCV infection (CHC) causes disordered lipid metabolism (3), and is associated both with lower serum concentrations of low-density lipoprotein (LDL) cholesterol and with hepatic steatosis (4,5) that resolves following successful anti-viral therapy (6, 7), particularly in those infected by HCV genotype-3 (HCV-G3). Additionally HCV-G3 has been found to be associated with more rapid liver brosis progression (8, 9) and an increased risk of developing hepatocellular cancer (HCC) (9, 10), compared to HCV genotype 1 (HCV-G1), independent of patients' age, diabetic status, body mass index, or antiviral treatment. A more recent Korean study of 1448 consecutive CHC patients has proposed HCV-G3 as an independent risk factor for HCC and disease progression (11).
Given the reliance of HCV on host lipid metabolism and clinical challenges posed by HCV-G3 infection, a detailed understanding of lipid perturbation in comparison with HCV-G1 may be relevant for understanding natural history of liver disease progression. Evidence is accumulating that lipogenic proteins are involved in hepatic carcinogenesis (12) and our previous work has suggested genotype differences in lipoprotein metabolism (13). In this study, we aimed to further characterise variation in host lipid metabolism between subjects chronically infected with HCV-G1 and HCV-G3.
We have combined several approaches to interrogation of lipid metabolism; initially analysis of the lipidome between subjects infected with HCV-G1and HCV-G3, both in the fasting and non-fasting states, and after sustained virological response (SVR) was performed to identify likely virally-mediated differences in lipid metabolism between the genotypes. Differences in the lipidomes have been correlated with detailed phenotyping of body fat distribution by in vivo magnetic resonance imaging and spectroscopy, including measurements of liver, adipose tissue and intramyocellular fat content in a subgroup of participants.
Additionally, for further mechanistic insight into virally-mediated lipid metabolism disturbances, measurement of non-cholesterol sterols in plasma was undertaken; this permitted evaluation of the relative contributions of endogenous cholesterol synthesis and dietary cholesterol absorption to whole body cholesterol homeostasis. Such non-cholesterol sterols are present in small quantities in plasma, distributed and transported with endogenously and exogenously derived cholesterol in all the lipoprotein classes. Lathosterol and desmosterol are late precursors in the endogenous cholesterol biosynthetic (mevalonate) pathway ( Figure 1). Absolute serum lathosterol and desmosterol concentrations and ratios to total serum cholesterol (i.e. lathosterol : total cholesterol ratio and desmosterol : total cholesterol ratio) are an index of endogenous cholesterol biosynthesis (14) and were also investigated in this study. Sitosterol is a plant sterol, derived exclusively from diet and is therefore an index of intestinal cholesterol absorption (15,16). Cholestanol is produced endogenously from cholesterol, excreted in bile and then reabsorbed. Serum concentrations of cholestanol re ect cholesterol absorption under physiologic conditions. Cholestanol is increased in cholestatic liver diseases due to decreased biliary secretion (17). Investigation of these pathways allowed greater insight into the disturbances of lipid metabolism caused by the different HCV genotypes and may allow further insight into the differing propensity for liver cancer development amongst differing viral genotypes.

Methods
Patients: Participants with chronic HCV infection were recruited at two centres: Newcastle-upon-Tyne and Imperial College London. All participants gave written, informed consent and the study had ethical approval (Northumberland REC 07/H0902/45 and Fife and Forth Valley REC 07/S0501/21). The research was performed in accordance with the relevant guidelines/regulations set out by the Northumberland and Fife and Forth Valley research ethics committees, and was performed in accordance with the Declaration of Helsinki of 1975.
All participants were age ≥ 18 years, HCV-RNA positive for >6 months, and not on a lipid modulating agent for 3-months prior to the study. Patients with hepatitis B, hepatitis delta, or HIV co-infection, or alcohol dependency were excluded. All participants attended following a >8 hour overnight fast for sample collection. The fasted cohort consisted of 112 fasting sera (39 G3, 73 G1); 25% had compensated cirrhosis evidenced by Fibroscan >12.5 KPa (Echosens, Paris, France). Baseline clinical and demographic data are shown in Table 1. Phenotyping of body fat distribution. A subgroup of 13 consecutively-attending participants from the fasted cohort (6 HCV-G1, 7 HCV-G3) at Imperial College London had additional detailed clinical phenotyping performed by determination of whole body fat distribution using in vivo magnetic resonance spectroscopy (MRS) to quantify intra-hepatocellular lipid (IHCL), intra-myocellular lipids in tibialis (T IMCL) and soleus muscles (S IMCL), and distribution of adipose tissue fat (% visceral and non-visceral fat) using magnetic resonance imaging, as previously described in detail by Thomas and colleagues (19).
Ultra Performance Liquid Chromatography Mass Spectroscopy (UPLC-MS) lipidomics. All samples were thawed at 4 C and prepared for UPLC-MS analysis by isopropanol protein precipitation by addition of 150µL of cold isopropanol to each 50 µL serum sample (ratio 3:1), as previously described by Sara an and colleagues in 2014 (20). Quality control (QC) samples were prepared by pooling equal volumes of all samples and injecting into the mass spectrometry system at regular intervals throughout the analytical runs, in order to de ne the system suitability, analytical stability, and sample repeatability. Serum lipid UPLC-MS pro ling was performed using an ACQUITY UPLC system (Waters Ltd., Elstree, UK), coupled to a Q-ToF Premier mass spectrometer (Waters MS Technologies Ltd, Manchester, UK) using an electrospray (ESI) ion source operated in both positive and negative electrospray ionisation modes (ESI+ and ESI-).
Liquid chromatography (LC) conditions have been previously described by Eliasson  Statistical analysis.
Where continuous data were normally distributed, two-sample t-tests were used to compare means between control groups. The Kruskal-Wallis test was used for comparison of non-parametric data.
Pearson's r correlation coe cient was used to determine relationships between continuous variables and Spearman's rank analysis for correlation between non-parametric variables. P < 0.05 was taken to indicate statistical signi cance. All statistical analyses were carried using Minitab version 16 (Minitab, State College, PA, USA).

Multivariate Statistical Analysis
The supervised and unsupervised multivariate models were generated using SIMCA (version 14.1, Umetrics, Umeå, Sweden). Principal component analysis (PCA) and orthogonal projections to latent structures discriminant analysis (OPLS-DA) were performed on all spectral data after pareto-scaling and log transformation for detection of patterns, trends and outliers; and construction of discriminant models were generated for classi cation and the discovery of potential biomarkers respectively.

Results
Clinical phenotype and fasting lipid pro les. Baseline clinical phenotype demonstrated no differences between HCV-G1 and HCV-G3 patients from the fasted cohort in terms of physical demographics of age, sex and BMI or severity of liver brosis (Table 1).
Fasting lipid pro les were signi cantly different in HCV-G3 compared to HCV-G1, manifesting as reductions in total cholesterol, non-HDL cholesterol and apoB (Table 1). Although there was no signi cant different in brosis assessment by liver stiffness, HCV-G3 participants had signi cantly increased liver enzymes: ALT and AST.
HCV-G3 decreases cholesterol synthesis via lathosterol rather than desmosterol intermediates. Noncholesterol sterol intermediates were analysed to understand potential pathways of low cholesterol pro les in HCV-G3. Lathosterol and desmosterol are both pre-cholesterol intermediates, and thus, serum   (Table 3). (lathosterol and desmosterol). IHFC showed a negative correlation with cholesterol synthesis via lathosterol in both HCV-G1 and HCV-G3, which was most marked in HCV-G1. However, we showed a positive correlation with desmosterol in both HCV-G1 and HCV-G3, which was most marked in HCV-G3 ( gure 2). Serum apoB concentration demonstrated weak positive correlations with IHFC in both HCV genotypes ( gure 3). There was a negative correlation between steatosis and fasting serum triglyceride (TG) levels in HCV-G1, with a non-signi cant positive correlation in HCV-G3. Collectively, this implies that steatosis in HCV-G3 is unrelated to decreased very low density lipoprotein (VLDL) particle export, but is more related to viral suppression of cholesterol synthesis via lathosterol, and relative sparing of desmosterol.
Non-targeted 'shotgun' lipidomics identi ed novel lipid species differentially regulated between HCV-G1 and HCV-G3. The UPLC-MS spectra from fasting sera of participants in the fasting CHC cohort were explored by PCA to detect clusters and outliers. Pairwise OPLS-DA established the lipids with the strongest contribution to genotypic separation. Figure 4 shows a PCA scores plot indicating clustering of HCV-G1 and HCV-G3, with close clustering of QC samples, indicating good platform stability.
Pairwise analysis using OPLS-DA established the lipid species with the greatest contribution to the genotype separation in positive ion mode (supplemental gure 1).
Using the S-plot from the OPLS-DA model, the in uence of individual lipid species (high signi cance and strong contribution to group separation) in the model were examined. Preliminary assignments were based on mass, fragmentation pattern and retention time of the identi ed lipid species that were upregulated in HCV Discriminating features of the lipidome between HCV genotypes 1 and 3 are not apparent following sustained virological response. To determine whether lipidomic differences resolve or persist after successful eradication of HCV following sustained virological response, further analysis was performed on a third cohort of non-viraemic post-SVR samples (SVR = sustained viral response following HCV antiviral treatment). Supplemental Figure 6 PCA demonstrates that there is no signi cant separation by previous HCV genotype exposure following SVR. This supports the notion that the observed genotypespeci c alterations in the lipidome in chronic HCV infection are due to the presence of active HCV infection and resolve with viral clearance.

Discussion
This study performed detailed characterisation of lipid metabolism in individuals chronically infected with HCV and demonstrated that there are distinct HCV genotype-speci c changes in lipid metabolism that change following SVR. This study is the most comprehensive description of altered lipid metabolism in subjects chronically infected with HCV-G3 to be reported. We have performed a combination of detailed lipid pro ling in fasting samples, including sterol markers of cholesterol synthesis and absorption, and additional quanti cation of liver, muscle and adipose tissue fat content by in vivo MRS and MRI in a small subgroup. We then performed UPLC-MS lipidomics analysis and made comparison between subjects with HCV-G1 and HCV-G3 in two independent viraemic cohorts, and a post-treatment SVR non-viraemic cohort. The ndings have demonstrated that individuals chronically infected with HCV-G3 have signi cantly decreased serum apoB, and non-HDL cholesterol concentrations, in conjunction with more hepatic steatosis than those with HCV-G1. This nding in itself is not new, but our observations challenge the widely held assumption that the steatosis in HCV-G3 is due to impaired hepatic VLDL export, potentially by inhibition of microsomal triglyceride transfer protein (MTP) (22). If this were the case, we would have expected negative correlations between liver fat with either serum apoB or TG concentrations in participants with HCV-G3. However, instead we observed a positive correlation between both apoB and TG with liver fat content in HCV-G3, and a negative correlation in individuals infected with HCV-G1. TG accumulation has been reported in HCC, compared to tumour adjacent tissue (23) and hepatic steatosis is known to be linked to HCC in CHC patients (24), so the mechanism(s) involved in promoting the differences between HCV-G3 and HCV-G1 could be relevant to hepatocarcinogenesis.
HCV-G3 subjects demonstrated an apparent divergence in decreasing markers of cholesterol synthesis, lathosterol and desmosterol. These two pre-cholesterol intermediates are on separate sides of the late cholesterol biosynthetic pathway. It appears that HCV-G3 preferentially decreases cholesterol synthesis via the lathosterol pathway ( gure 1). The observation of decreased serum lathosterol levels in HCV-G3, with relatively normal desmosterol levels implies that HCV-G3 selectively inhibits the lathosterol arm of endogenous cholesterol synthesis. Low lathosterol levels have been reported in another study of HCV-G2 and HCV-G3 infection, indicating that HCV-G3 selectively perturbs the late cholesterol synthesis pathway (25), and in HCV-G3 individuals with cirrhosis, low lathosterol was a predictor of virologic relapse following sofosbuvir and ribavirin treatment (26). We measured only lathosterol and desmosterol as synthesis markers, which provides information about the relative ux through the two pathways, but did not measure additional upstream pre-cholesterol intermediates. Desmosterol is produced from reduction of 7-dehydrodesmosterol by the enzyme ∆7-sterol reductase (DHCR7) in the Bloch pathway. It has been reported that HCV selectively perturbs the late stages of cholesterol biosynthesis in HCV-G2 and HCV-G3, where lathosterol and 7-dehydrocholesterol concentrations were low, but increased following viral clearance, and the proximal metabolite lanosterol was preserved (27). The present study adds to the literature by reporting low lathosterol concentrations in a larger number of HCV-G1 and HCV-G3 patients with chronic infection. This may contribute to the high prevalence of vitamin D de ciency among HCV patients (28). Of additional interest is the strong negative correlation between suppressed cholesterol synthesis via lathosterol and increased hepatic fat content. This implies that as HCV suppresses cholesterol synthesis, pathways of hepatic triglyceride accumulation are being activated without diminishing VLDL export, possibly by an up-regulation of compensatory pathways, such as reverse cholesterol transport and liver X receptors (LXR), which are potently activated by desmosterol.
The data from our untargeted lipidomic analyses of the same cohorts of subjects with HCV-G1 and HCV-G3 in the fasting state have identi ed additional lipid species differentially regulated between the genotypes, causing clear genotype speci c separation of fasting sera in the PCA scores plots. Amongst lipid species accounting for the separation, we observed increased phosphocholines in HCV-G1 and increased cholesteryl esters, including cholesteryl linoleate in HCV-G3. These changes in the lipidome were not apparent in patients who achieved SVR following treatment for previous HCV-G1 or HCV-G3 infection, implying that the lipidomic features are mediated by active HCV viraemia.
Experimental data generated from expression of HCV-G3a core protein in Huh-7 cells have previously reported increasing expression of cholesteryl esters, ceramides and glycosylceramides, but not triglycerides induced by the steatogenic HCV-G3 core protein and suggested that viral steatosis may be distinct from metabolic steatosis (29). In vivo lipoproteins undergo continuous remodelling during their transit in plasma and we have reported that HCV also undergoes remodelling and transfer on to very-low density lipoproteins after a fatty meal (30). Increased serum cholesteryl linoleate (CL) levels observed in fasting HCV-G3 participants in our study supports the concept that reverse cholesterol transport is also up-regulated in HCV-G3 infected participants. CL is a cholesteryl ester, which is not synthesised in the liver but produced in the reverse cholesterol transport pathway from peripheral tissues by lecithin-cholesterol acyl transferase (LCAT) activity on HDL. LCAT serves to maintain a cholesterol gradient between peripheral tissues and HDL. LCAT activity enriches HDL in CL as the predominant cholesterol ester. CL is subsequently redistributed amongst all apoB lipoprotein classes by cholesteryl ester transfer protein (CETP) activity, mediating CL transfer from HDL to apoB containing lipoproteins as well, which are subsequently tra cked back to the liver (31). CETP is increased in active HCV infection (32). Up-regulated reverse cholesterol transport to the liver may be a compensatory homeostatic response to decreased endogenous cholesterol synthesis in HCV-G3 infection.
Phosphatidylcholine (PC) is a highly abundant phospholipid, and functions as a major constituent of cell membranes. PC is a phospholipid with a typical structure of a choline head group and two fatty acids PCs (31). In vitro, a number of intermediates involved in PC synthesis have been shown to be elevated in HCV infected Huh-7.5 cells (33). In our in vivo lipidomics analysis, we have identi ed increased levels of long chain (C36 and C38) unsaturated PCs in fasting HCV-G1 participants. The functional importance of this is not known, but these variations in fatty acids may affect membrane uidity and utility of PC associated fatty acids as a source of liver triglycerides. In sera, PC is associated with all lipoprotein classes, including HDL and LDL. In HDL metabolism, nascent HDL particles produced by the liver contain lipid poor apoA1, which is then secreted from the liver and gathers excess cholesterol and phospholipids from peripheral tissues by ABCA1-or ABCG1-mediated e ux from peripheral tissues. As HDL particles acquire cholesterol from peripheral tissues they increase in size, hence also acquire additional PC from non-hepatic tissues to accommodate the increasing surface area of the HDL particles. PC associated with either HDL or LDL is subsequently e ciently taken up by hepatocytes. Although we have not shown signi cant quantitative changes in ApoA1, the PC composition of HDL may be altered by HCV-G1 infection. Studies in mice have indicated that PC is a major and under-recognised source of FA delivery to the liver, which can be a quantitatively important source of hepatic triglyceride. Up to one-third of HDL-PC delivered to the liver can be hydrolysed by PLC and subsequently re-esteri ed to form hepatic triglycerides. The HDL receptor in the liver is SR-B1, which is responsible for selective uptake of HDL-cholesteryl esters. SR-B1 is responsible for 50% of uptake of PC in isolated hepatocytes (34). HCV has been also demonstrated to utilise SR-B1 as a hepatocyte entry co-factor (35,36). Therefore, increased ux of PC through the reverse cholesterol transport / SR-B1 pathway into the liver may favour the HCV lifecycle by utilising entry pathways via SR-B1.
The strengths of this study are that the lipidomics analysis in cohort 1 was performed in fasting samples and demonstrated robust models in PCA of HCV genotype separation in the lipidome, allowing identi cation of several lipid species differentially regulated by HCV-G1 and HCV-G3, respectively. Although the trend was similar, the models were less robust in the second cohort, which could be explained by the fact that the post-prandial status of the serum samples in cohort 2 and 3 were unknown, and it is likely that donations to HCV Research UK included variable numbers of post-prandial samples. Therefore, the contribution of viraemia to alterations in the lipidome apparent in the fasting state could have been masked somewhat in the presence of varying degrees of postprandial lipaemia.
In summary, we have demonstrated that compared to HCV-G1, HCV-G3 infection is characterised by low LDL cholesterol levels, with preferential suppression of cholesterol synthesis via lathosterol, and preservation of desmosterol levels, associated with increasing hepatic steatosis. Lipidomics analysis revealed lipid species associated with reverse cholesterol transport speci cally increased in HCV-G3, which may imply genotype-speci c lipid mechanisms involved in liver disease progression and promotion of HCC (37).   Correlation between intra-hepatocellular lipid content (steatosis) and fasting serum mark-ers of endogenous cholesterol synthesis lathosterol and desmosterol in HCV genotypes 1 (N=6) and 3 (N=7) Figure 3