Large-scale studies have estimated that around 30–40% HIV-infected individuals, with or without hepatitis C virus (HCV) coinfection, suffer from fatty liver disease (FLD) [1–4]. Because of this and taking into account that FLD can lead to advanced liver disease, including cirrhosis and hepatocellular carcinoma , the identification of conditions underlying FLD in the setting of HIV infection is a major challenge.
The heritability of nonalcoholic fatty liver disease (NAFLD) measured by computed tomography scan in individuals without HIV infection has been estimated at 25–30% . This data indicate the existence of an important genetic basis for this disorder in that population. In order to discover single nucleotide polymorphisms (SNPs) associated with NAFLD, several genome-wide association studies (GWAS) have been carried out in HIV-uninfected individuals of several ethnicities [6–11]. These studies have identified several SNPs independently associated with NAFLD located within or close to different loci including PNPLA3, COL13A1, EFAB4B, EHBP1L1, FDFT1, GCKR, LYPLAL1, NCAN, PPP1R3B, GC, LCP1, LPPR4, SLC38A8, PARVB, and SAMM50 genes.
The causes that induce FLD, including nonalcoholic steatohepatitis, in the HIV-infected population are not completely understood. The possible role of antiretroviral drugs and of HCV coinfection in the development and progression of FLD has been studied with contradictory results [12–17]. In a recent study, the BMI was found to be the major factor associated with FLD in HIV-infected patients . Likewise, as in HIV-uninfected patients, genetic factors should also have a role on the risk of FLD development in this population. However, those SNPs related to NAFLD in the general population by GWAS have been scarcely studied in the setting of HIV infection .
The aim of this study was to evaluate the association of those SNPs previously found to be related to NAFLD by GWAS performed in the general population with the presence of FLD and steatohepatitis in HIV-infected individuals.
Patients and methods
Patients and study design
Caucasian HIV-infected individuals who consecutively attended outpatient clinics in two university hospitals in Spain between November 2011 and July 2013 were included in this study. Individuals were selected if they were 18 years or older and gave their written informed consent to participate in the study. Pregnant women as well as those individuals who showed family relationship with other selected patients were excluded from the study.
All patients underwent a hepatic transient elastography examination for determining controlled attenuation parameter (CAP) during a single visit previously scheduled as a routine follow-up visit. Data for this study were collected during the same visit. In all patients, a whole blood sample was collected for routine laboratory and genetic determinations.
Transient elastography examination
Liver stiffness with CAP measurements was performed by an experienced operator at each participating center using a commercially available elastography device (FibroScan 502; Echosens, Paris, France). Elastography examinations were conducted as stated elsewhere [1,18]. CAP is an estimate of the total ultrasonic attenuation and it is expressed in dB/m. According to previously reported standards, examinations with 10 successful shots, an interquartile range for liver stiffness less than 30% of the median value and a success rate of at least 60%, were considered as reliable [1,18].
Polymorphism selection, genotyping, and genotyping quality control
SNPs were selected from those reported as associated with NAFLD by GWAS performed in the general population [6–11] (Table 1).
DNA was extracted from frozen whole blood using the Quick Pure Blood DNA extraction Kit (Macherey-Nagel, Düren, Germany). The SNPs were genotyped using a custom Golden Gate protocol (Illumina, San Diego, California, USA). SNP genotyping services were provided by the Spanish ‘Centro Nacional de Genotipado’ (CGEN-ISCIII) (www.cegen.org).
All SNPs and genotyped samples were subjected to quality control filters using the PLINK software (http://pngu.mgh.harvard.edu/∼purcell/plink/) . SNPs that were not successfully genotyped in at least 80% of individuals as well as those with a P value for Hardy–Weinberg equilibrium lesser than 0.05 were discarded (Table 1). In addition, individuals with more than 20% of missing genotype data were excluded.
A cutoff of 238 dB/m was selected to define the presence of FLD, that is, steatosis involving at least 10% of hepatocytes. This cut-off has been reported to be useful to discriminate hepatic steatosis affecting at least 10% hepatocytes [1,20]. In addition, it has shown a negative predictive value of 0.81 to detect this level of steatosis in a prospective study .
A composite variable, including the presence of FLD and elevated alanine aminotransferase (ALT) levels, was considered as a surrogate marker of steatohepatitis. ALT levels above 19 IU/l in women and 30 IU/l in men were considered elevated .
Two primary variables were used in this study as follows: presence of FLD and presence of steatohepatitis.
Plink software  was used to carry out standard case/control allelic association (one degree of freedom) studies. As a quality control, only those SNPs showing univariate allelic association (P < 0.05) with the studied phenotype in the same direction than that had been formerly reported in the respective GWAS were considered in subsequent analyses.
The online resource at the Institute for Human Genetics, Munich, Germany (http://ihg.gsf.de) was used for testing comparison of genotypic frequencies between groups to determine P values, odds ratios (ORs) and 95% confidence intervals (95% CIs) as well as different genetic models.
Haploview software (version 4.1) (https://www.broad.harvard.edu/haploview/haploview) was employed to calculate linkage disequilibrium D’ values between genetic markers.
To compare the categorical variables in two different groups of individuals, we used the Pearson's chi-square test. Quantitative variables among patient groups were compared by means of the Student's t-test (data normally distributed) or Mann–Whitney U test (data not normally distributed). Logistic regression models were elaborated including variables with a univariate P value < 0.15, as well as age and sex, to obtain adjusted P and OR values and to detect factors independently associated with the primary variables. All these calculations were carried out using the SPSS software 22.0 (IBM Corporation, Somers, New York, New York, USA).
The estimation of power to detect risk alleles of FLD or steatohepatitis was performed by the Episheet software available at http://members.aol.com/krothman.
This study was in compliance with the national legislation and it was performed according to the ethical guidelines of the Declaration of Helsinki. The study was approved by the Ethics Committee of the Hospital Universitario de Valme. Written consent was obtained from all individuals before sampling.
Features of the study population
A total of 447 patients fulfilled the inclusion criteria for this study. However, 16 individuals were excluded because of low rate of sample genotyping (three individuals) or low success in liver stiffness examination (13 individuals). Overall, 431 consecutive patients were enrolled in this study. Among them, 47 (10.9%) had been previously diagnosed with type 2 diabetes mellitus. More than 96% of them were treated with antiretroviral therapy and showed undetectable plasma HIV-RNA. A total of 350 (81.2%) patients were tested positive for serum HCV antibodies, of which 26 (7.4%) had spontaneously cleared the virus and 48 (13.7%) had achieved sustained virological response. Therefore, 276 (64.0%) individuals carried detectable plasma HCV-RNA. In the study population, FLD was detected in 179 (41.5%) individuals, whereas steatohepatitis was found in 122 (28.3%) individuals. The main characteristics of the study population are summarized in Supplementary Table 1, http://links.lww.com/QAD/A746.
Genotypic quality control analysis
Among the 19 SNPs selected for this study, located within or close to 15 known loci, two (10.5%) were discarded by the quality control analysis due either to deviation of the Hardy–Weinberg equilibrium (rs7324845, within LCP1 gene), or to low genotyping call rate (rs3761472, within SAMM50 gene) (Table 1). The 17 SNPs available for association studies reached a total genotyping rate of 99.7% in the entire population.
Association of genetic factors with fatty liver disease
A standard case/control allelic association analysis was carried out to detect those markers associated with FLD (Supplementary Table 2, http://links.lww.com/QAD/A746). Two SNPs were related to this disorder, rs738491, within SAMM50 gene, and rs12743824, linked to LPPR4 gene. These associations were in the same direction that had been previously reported (Table 1). Specifically, the SAMM50 rs738491-T allele was associated with a higher frequency of FLD (P = 0.016; OR = 1.44; 95% CI = 1.06–1.95), whereas the LPPR4 rs12743824-A allele was associated to a lower frequency of this condition (P = 0.018; OR = 0.71; 95% CI = 0.54–0.94) (Supplementary Table 2, http://links.lww.com/QAD/A746).
The genotypic distribution of theses polymorphisms is displayed in Table 2. In a more detailed study of the SAMM50 rs738491 polymorphism, we observed that among 38 individuals carrying the TT genotype, 23 (60.5%) harbored FLD, whereas the proportion of individuals with FLD among those bearing CT or CC was 39.6% (156 of 393 individuals) (P = 0.012; OR = 2.33; 95% CI = 1.18–4.60). Regarding LPPR4 rs12743824, only 111 of 293 (37.8%) individuals carrying the AC or AA genotypes showed FLD and 70 of 138 (49.2%) harboring the CC genotype were diagnosed with FLD (P = 0.025; OR = 0.62; 95% CI = 0.41–0.94) (Table 2).
In addition to these SNPs, higher fasting plasma glucose levels and BMI, current treatment with nucleoside reverse transcriptase inhibitors, and absence of treatment with CCR5 antagonist were also associated with the presence of FLD in the univariate study (Table 2). The results of the multivariate analysis are summarized in Table 3. Only the LPPR4 rs12743824 marker and higher BMI were identified as independent risk factors associated with FLD.
We performed a further sensitivity analysis limited to HCV-coinfected participants, and the results of the multivariate analysis were qualitatively unchanged (data not shown).
Association of genetic factors with steatohepatitis
In the standard case/control allelic association analysis, the rs4240624 genetic marker, located close to PPP1R3B, and the SAMM50 rs738491 polymorphism, were associated with this disorder and in the same directions that had been previously reported (Table 1). Thus, the PPP1R3B rs4240624-G allele was associated with a lower frequency of steatohepatitis (P = 0.024; OR = 0.40; 95% CI = 0.18–0.91), whereas the SAMM50 rs738491-T allele was associated with a higher frequency of this condition. (P = 0.034; OR = 1.41; 95% CI = 1.02–1.96) (Supplementary Table 2, http://links.lww.com/QAD/A746). The LPPR4 rs12743824 marker, (P = 0.059), as well as the rs2143571 SNP (P = 0.051) trended to be associated with steatohepatitis (Supplementary Table 2, http://links.lww.com/QAD/A746). This last SNP is also located within SAMM50 gene and in linkage disequilibrium with the SAMM50 rs738491 marker (D’ = 64) (Fig. 1).
The genotypic distributions of PPP1R3B rs4240624 and SAMM50 rs738491 polymorphisms in patients with and without steatohepatitis are displayed in Table 2. For PPP1R3B rs4240624, only seven of 43 individuals (15.2%) carrying the protective G allele (GG or GA genotypes) showed steatohepatitis vs. 115 of 385 (29.8%) AA individuals (P = 0.037; OR = 0.42; 95% CI = 0.18–0.97). Regarding SAMM50 rs738491, among 38 individuals with the TT genotype, 20 (52.6%) showed steatohepatitis, whereas 102 of 393 (25.9%) individuals with genotype TC or CC showed this disorder (P = 4.9 × 10−4; OR = 3.17; 95% CI = 1.61–5.46) (Table 2).
Other univariate associations with steatohepatitis were observed for higher BMI, higher liver stiffness, higher fasting plasma glucose levels and current treatment with nucleoside reverse transcriptase inhibitors (Table 2). In multivariate analysis, only BMI and the SAMM50 rs738491 marker remained as independently associated factors with steatohepatitis (Table 3).
When only HCV-coinfected participants were selected, the results of the multivariate analysis were qualitatively unchanged (data not shown).
In this study, we found that LPPR4 rs12743824 and SAMM50 rs738491 were independently associated with CAP-determined FLD and with a surrogate marker of steatohepatitis, respectively.
We have described a dominant genetic model for the protective effect of the LPPR4 rs12743824-A allele on FLD risk. Moreover, we have reported a recessive model for the effect of SAMM50 rs738491-T allele on the risk of steatohepatitis development. In the original works in which these genetics variants were associated with NAFLD, additive genetic models were analyzed as is usual in GWAS [7,11] (Table 1). However, when the genotype data reported by Kitamoto et al. were analyzed, it was possible to test that the genetic model proposed herein for SAMM50 rs738491 had a higher effect size than that observed for the additive model. Regarding LPPR4 rs12743824, no genotype data were reported in the original work  and, for that reason, it was not possible to perform a comparison between different genetic models. In spite of this when an additive model was applied in our population, similar results than those reported in the original work were obtained. Nevertheless, the effect size observed for the additive model was lower than that obtained for the genetic model proposed herein.
LPPR4 is a member of the lipid phosphatase family involved in the dephosphorylating of a number of bioactive mediators that regulate a variety of cell functions . Its expression is restricted to the neurons, mainly in the hypothalamus. For that reason, the possible role of this gene in the development of FLD cannot be easily inferred. However, it is known that the HIV can infect the central nervous system including the hypothalamus. This fact could explain the disturbances in neuroendocrine functioning observed in some HIV-infected patients . Because of the role that the hypothalamus plays in the control of lipid metabolism , the association between the LPPR4 variant and FLD in the HIV-infected population reported herein could be pointing out a specific link between the HIV-associated damage of the central nervous system and the metabolic alterations in the HIV infection setting. If our data are confirmed, it will be necessary to explore this possibility.
SAMM50 is a component of the sorting and assembly machinery necessary for the assembly of beta-barrel proteins in the outer mitochondrial membrane. Recently, it has been reported that SAMM50 is involved in the structural integrity of the mitochondrial cristae and in the maintenance of mitochondrial DNA. Thus, it was observed that long-term depletion of SAMM50 influences the amount of proteins from all large respiratory complexes that contain the mitochondria . Accumulating evidence indicate that hepatic mitochondria play a critical role in the development and pathogenesis of FLD . Moreover, mitochondrial structural defects have been observed in patients with nonalcoholic steatohepatitis . Interestingly, our results indicated that SAMM50 rs738491-T allele is associated with steatohepatitis. Similar results were obtained by the authors who firstly reported the independent association of this gene with FLD in Japanese population . Altogether, these data suggest the existence of a genetic susceptibility to mitochondrial damage, which could be more evident in some stress conditions such as those induced by HIV infection. According to this hypothesis, it has been reported that the liver mitochondrial DNA content seems to be affected by HIV infection . Therefore, our results support the idea that the liver mitochondrial damage could have an important role in the development of steatohepatitis in the HIV-infected population.
Other polymorphisms located in the proximity of SAMM50 rs738491 marker within or close to SAMM50 and PARVB genes, which had been previously associated with FLD in a Japanese population , were also included in this study. However, our study did not confirm such relationship, although one of them (SAMM50 rs2143571) showed a trend to an association with both end points analyzed here (Supplementary Table 2, http://links.lww.com/QAD/A746). Those discrepancies are probably due to a higher linkage disequilibrium between these markers in Japanese as compared with Caucasian populations  (Fig. 1).
The PNPLA3 rs738409 marker has been reported to be associated with FLD in several studies conducted in the general population of different ethnicities [6,9–11,30,31]. However, the magnitude and strength of the effect of the variant on this disease widely varied among populations of similar ethnic background . In addition, some studies failed to confirm such association [8,14]. In a recent analysis carried out in the Multicenter AIDS Cohort Study (MACS), no association was found between FLD and the PNPLA3 rs738409 marker in HIV-uninfected individuals . However, in the subset of HIV-infected individuals included in that analysis of the MACS, the PNPLA3 rs738409 variation was associated with FLD . Our study, which recruited a population of HIV-infected patients with FLD almost three times larger than that of MACS, failed to replicate that finding. It has been postulated that this maker is important as a secondary risk factor for FLD when a stressor such as obesity is present [32,33]. Moreover, it has been observed that the effect of this SNP may be influenced by sex, with a stronger association among women . These findings could explain why we did not observe such association in our sample wherein the 81% of the individuals were men with low BMI.
This study has some limitations. First, hepatic elastography with CAP does not measure hepatic inflammation. To solve this problem, we used CAP-determined FLD with elevated ALT as a surrogate marker of steatohepatitis as similarly described . However, elevated ALT is a nonspecific marker of liver inflammation [34,35]. This could have resulted in an overestimation of cases with steatohepatitis. Second, we did not apply a multiple testing correction in our statistical analysis, for that reason our results might be taken with caution. However, as a quality control, we established as valid associations those that showed the same effect direction that was reported in the original work using standard allelic cases/controls association methods, and that were independently associated with FLD or steatosis in multivariate analysis including factors strongly related to these end points, as BMI. We think that the consistence of SNP effects among different studies could be an interesting alternative to validate SNPs related to FLD susceptibility. Otherwise, the application of multiple testing correction would have eliminated those SNPs that have been successfully replicated in our population. Third, taking into account the sample size analyzed, the case/control ratio and a minor allele frequency of 0.25, our study had a 80 or 72% power to detect OR = 1.8 in the study of FLD or steatohepatitis, respectively. Therefore, this study could be underpowered to detect associations with those SNPs showing a lesser minor allele frequency (Supplementary Table 2, http://links.lww.com/QAD/A746). Finally, the studied population had a great proportion of HCV/HIV coinfected patients. Because of this, our results are mainly applicable to this particular subpopulation. Therefore, although the LPPR4 and SAMM50 alleles were independent factors associated with FLD and steatohepatitis, respectively, in our sample, as proven in the multivariate analysis, specific studies in HIV-monoinfected patients may be required to confirm the associations reported herein.
In summary, we have carried out one of the largest studies performed on this topic and the first one, to our knowledge, that explored the association between numerous of NAFLD-related SNPs and FLD, including steatohepatitis, in the setting of HIV infection. This work reports for the first time an association between the LPPR4 rs12743824 marker and FLD, as well as an association between the SAMM50 rs738491 variation and steatohepatitis in HIV-infected individuals. These findings suggest a role of the central nervous system in the FLD development and support the hypothesis that the liver mitochondria damage has an important role on the progression of FLD in the HIV-infected population. Although our results need to be confirmed in other series, they could open new clues for understanding the cause of FLD in this setting. Finally, these genetic markers could help clinicians identify HIV-individuals at risk of developing FLD.
Source of funding: This study was supported by the grants from the Consejería de Salud de la Junta de Andalucía (PI-0118–2013, PI-0481–2012 and AC-0095–2013), Gilead (GLDL13–00145), the Ministerio de Salud (EC11–2086), the ISCIII-Subdirección General de Evaluación and the Fondo Europeo de Desarrollo (ISCIII-RETIC RD06/006 and RD12/0017). JAP has received a research extension grant from the Programa de Intensificación de la Actividad de Investigación del Servicio Nacional de Salud Carlos III (I3SNS).
Authors’ contributions: Study design: J.M., A.R., J.A.P., F.T., D.M., and L.M.R. Clinical characterization, database construction and clinical research oversight: J.M., A.R., F.T., D.M., N.M., and J.A.P. DNA management and bioinformatics: L.M.R., A.R.-J., M.F., and K.N. Statistical analysis: L.M.R., A.R.-J., M.F., J.M., and K.N. Writers: L.M.R., J.M., and J.A.P. with contribution of all authors.
J.M. is the recipient of a grant from the Servicio Andaluz de Salud de la Junta de Andalucía (B-0037). K.N. is the recipient of a research contract from the ISCIII (CP13/00187). A.R.-J. is the recipient of a Post-Doctoral Research Extension Grant from the Fundación Progreso y Salud, Consejería de Salud, Innovación y Ciencia de la Junta de Andalucia (RH-0024–2013 CSIPS-POSTDOCTORALES Y GRUPOS).
Conflicts of interest
There are no conflicts of interest.
1. Macias J, Gonzalez J, Tural C, Ortega-Gonzalez E, Pulido F, Rubio R, et al. Prevalence and factors associated with liver steatosis as measured by transient elastography with controlled attenuation parameter in HIV-infected patients
2. Crum-Cianflone N, Dilay A, Collins G, Asher D, Campin R, Medina S, et al. Nonalcoholic fatty liver disease among HIV-infected persons
. J Acquir Immune Defic Syndr
3. Guaraldi G, Squillace N, Stentarelli C, Orlando G, D’Amico R, Ligabue G, et al. Nonalcoholic fatty liver disease in HIV-infected patients referred to a metabolic clinic: prevalence, characteristics, and predictors
. Clin Infect Dis
4. Rivero-Juarez A, Camacho A, Merchante N, Perez-Camacho I, Macias J, Ortiz-Garcia C, et al. Incidence of liver damage of uncertain origin in HIV patients not co-infected with HCV/HBV
. PLoS One
5. Ascha MS, Hanouneh IA, Lopez R, Tamimi TA, Feldstein AF, Zein NN. The incidence and risk factors of hepatocellular carcinoma in patients with nonalcoholic steatohepatitis
6. Speliotes EK, Yerges-Armstrong LM, Wu J, Hernaez R, Kim LJ, Palmer CD, et al. Genome-wide association analysis identifies variants associated with nonalcoholic fatty liver disease that have distinct effects on metabolic traits
. PLoS Genet
7. Adams LA, White SW, Marsh JA, Lye SJ, Connor KL, Maganga R, et al. Association between liver-specific gene polymorphisms and their expression levels with nonalcoholic fatty liver disease
8. Chalasani N, Guo X, Loomba R, Goodarzi MO, Haritunians T, Kwon S, et al. Genome-wide association study identifies variants associated with histologic features of nonalcoholic fatty liver disease
2010; 139:1567–1576.1576 e1561-1566.
9. Kawaguchi T, Sumida Y, Umemura A, Matsuo K, Takahashi M, Takamura T, et al. Genetic polymorphisms of the human PNPLA3 gene are strongly associated with severity of nonalcoholic fatty liver disease in Japanese
. PLoS One
10. Romeo S, Kozlitina J, Xing C, Pertsemlidis A, Cox D, Pennacchio LA, et al. Genetic variation in PNPLA3 confers susceptibility to nonalcoholic fatty liver disease
. Nat Genet
11. Kitamoto T, Kitamoto A, Yoneda M, Hyogo H, Ochi H, Nakamura T, et al. Genome-wide scan revealed that polymorphisms in the PNPLA3, SAMM50, and PARVB genes are associated with development and progression of nonalcoholic fatty liver disease in Japan
. Hum Genet
12. Macias J, Berenguer J, Japon MA, Giron-Gonzalez JA, Rivero A, Lopez-Cortes LF, et al. Hepatic steatosis and steatohepatitis in human immunodeficiency virus/hepatitis C virus-coinfected patients
13. Martinez V, Ta TD, Mokhtari Z, Guiguet M, Miailhes P, Valantin MA, et al. Hepatic steatosis in HIV-HCV coinfected patients receiving antiretroviral therapy is associated with HCV-related factors but not antiretrovirals
. BMC Res Notes
14. Price JC, Seaberg EC, Latanich R, Budoff MJ, Kingsley LA, Palella FJ Jr, et al. Risk factors for fatty liver in the Multicenter AIDS Cohort Study
. Am J Gastroenterol
15. Sulkowski MS, Mehta SH, Torbenson M, Afdhal NH, Mirel L, Moore RD, et al. Hepatic steatosis and antiretroviral drug use among adults coinfected with HIV and hepatitis C virus
16. Woreta TA, Sutcliffe CG, Mehta SH, Brown TT, Higgins Y, Thomas DL, et al. Incidence and risk factors for steatosis progression in adults coinfected with HIV and hepatitis C virus
17. Machado MV, Oliveira AG, Cortez-Pinto H. Hepatic steatosis in patients coinfected with human immunodeficiency virus/hepatitis C virus: a meta-analysis of the risk factors
18. Vergara S, Macias J, Rivero A, Gutierrez-Valencia A, Gonzalez-Serrano M, Merino D, et al. The use of transient elastometry for assessing liver fibrosis in patients with HIV and hepatitis C virus coinfection
. Clin Infect Dis
19. Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MA, Bender D, et al. PLINK: a tool set for whole-genome association and population-based linkage analyses
. Am J Hum Genet
20. Myers RP, Pollett A, Kirsch R, Pomier-Layrargues G, Beaton M, Levstik M, et al. Controlled Attenuation Parameter (CAP): a noninvasive method for the detection of hepatic steatosis based on transient elastography
. Liver Int
21. Friedrich-Rust M, Romen D, Vermehren J, Kriener S, Sadet D, Herrmann E, et al. Acoustic radiation force impulse-imaging and transient elastography for noninvasive assessment of liver fibrosis and steatosis in NAFLD
. Eur J Radiol
22. Prati D, Taioli E, Zanella A, Della Torre E, Butelli S, Del Vecchio E, et al. Updated definitions of healthy ranges for serum alanine aminotransferase levels
. Ann Intern Med
23. Sciorra VA, Morris AJ. Roles for lipid phosphate phosphatases in regulation of cellular signaling
. Biochim Biophys Acta
24. Langford D, Baron D, Joy J, Del Valle L, Shack J. Contributions of HIV infection in the hypothalamus and substance abuse/use to HPT dysregulation
25. Nogueiras R, Lopez M, Dieguez C. Regulation of lipid metabolism by energy availability: a role for the central nervous system
. Obes Rev
26. Ott C, Ross K, Straub S, Thiede B, Gotz M, Goosmann C, et al. Sam50 functions in mitochondrial intermembrane space bridging and biogenesis of respiratory complexes
. Mol Cell Biol
27. Nassir F, Ibdah JA. Role of mitochondria in nonalcoholic fatty liver disease
. Int J Mol Sci
28. Sanyal AJ, Campbell-Sargent C, Mirshahi F, Rizzo WB, Contos MJ, Sterling RK, et al. Nonalcoholic steatohepatitis: association of insulin resistance and mitochondrial abnormalities
29. Bauerle J, Laguno M, Mauss S, Mallolas J, Murillas J, Miquel R, et al. Mitochondrial DNA depletion in liver tissue of patients infected with hepatitis C virus: contributing effect of HIV infection?
. HIV Med
30. Hernaez R, McLean J, Lazo M, Brancati FL, Hirschhorn JN, Borecki IB, et al. Association between variants in or near PNPLA3, GCKR, and PPP1R3B with ultrasound-defined steatosis based on data from the third National Health and Nutrition Examination Survey
. Clin Gastroenterol Hepatol
31. Palmer ND, Musani SK, Yerges-Armstrong LM, Feitosa MF, Bielak LF, Hernaez R, et al. Characterization of European ancestry nonalcoholic fatty liver disease-associated variants in individuals of African and Hispanic descent
32. Sookoian S, Pirola CJ. Meta-analysis of the influence of I148 M variant of patatin-like phospholipase domain containing 3 gene (PNPLA3) on the susceptibility and histological severity of nonalcoholic fatty liver disease
33. Graff M, North KE, Franceschini N, Reiner AP, Feitosa M, Carr JJ, et al. PNPLA3 gene-by-visceral adipose tissue volume interaction and the pathogenesis of fatty liver disease: the NHLBI family heart study
. Int J Obes (Lond)
34. Schindhelm RK, Diamant M, Dekker JM, Tushuizen ME, Teerlink T, Heine RJ. Alanine aminotransferase as a marker of nonalcoholic fatty liver disease in relation to type 2 diabetes mellitus and cardiovascular disease
. Diabetes Metab Res Rev
35. Yamada J, Tomiyama H, Yambe M, Koji Y, Motobe K, Shiina K, et al. Elevated serum levels of alanine aminotransferase and gamma glutamyltransferase are markers of inflammation and oxidative stress independent of the metabolic syndrome