Dig Dis Sci (2014) 59:2967–2974 DOI 10.1007/s10620-014-3279-z
ORIGINAL ARTICLE
Role of the PNPLA3 I148M Polymorphism in Nonalcoholic Fatty Liver Disease and Fibrosis in Korea Sang Soo Lee • Young-Sang Byoun • Sook-Hyang Jeong • Byung Hyun Woo • Eun Sun Jang • Jin-Wook Kim • Hyun Young Kim
Received: 17 January 2014 / Accepted: 2 July 2014 / Published online: 29 July 2014 Ó Springer Science+Business Media New York 2014
Abstract Background The role of the patatin-like phospholipase domain-containing 3 (PNPLA3) single-nucleotide polymorphism (SNP), rs738409, in the development and progression of nonalcoholic fatty liver disease (NAFLD) has not been studied in the Korean population. Aims The aim of the study was to investigate the genotype frequency and allele distribution of PNPLA3 rs738409
Sang Soo Lee and Young-Sang Byoun have equally contributed to this work. S. S. Lee Y.-S. Byoun S.-H. Jeong (&) B. H. Woo E. S. Jang J.-W. Kim Department of Internal Medicine, Seoul National University College of Medicine, Seoul National University Bundang Hospital, 82 Gumi-ro, 173 Beon-gil, Bundang-gu, Seongnam-si 463 707, Gyeonggi-do, Republic of Korea e-mail: [email protected] S. S. Lee e-mail: [email protected] Y.-S. Byoun e-mail: [email protected] B. H. Woo e-mail: [email protected] E. S. Jang e-mail: [email protected] J.-W. Kim e-mail: [email protected] H. Y. Kim Health Promotion Center, Seoul National University College of Medicine, Seoul National University Bundang Hospital, 82 Gumi-ro, 173 Beon-gil, Bundang-gu, Seongnam-si, Gyeonggi-do, Republic of Korea e-mail: [email protected]
and the association between the SNP and development of NAFLD and liver fibrosis. Methods A total of 339 Korean adults (155 NAFLD patients and 184 healthy controls) were enrolled. PNPLA3 SNP genotyping was carried out using a TaqMan allelic discrimination assay. Liver fibrosis severity was evaluated by NAFLD fibrosis score (NFS) and BARD score. Results The frequencies of the PNPLA3 rs738409 genotypes, CC, CG, and GG in the healthy control group were 29.9, 50.0, and 20.1 %, respectively, and those in NAFLD patients were 20.0, 48.4, and 31.6 %, respectively, showing a higher frequency of the risk allele (G allele) (p = 0.006). Among the NAFLD patients, the CG?GG genotype frequency was significantly higher in patients with advanced fibrosis, defined as NFS C -1.455 or BARD score C2, than in patients with mildto-moderate fibrosis (p = 0.012 and p = 0.046, respectively). In multivariate analysis, the CG?GG genotype was an independent factor for NAFLD development (odds ratio 2.568, 95 % CI 1.109–5.945, p = 0.028) and for advanced liver fibrosis according to the criteria of NFS C -1.455 (odds ratio 18.573, 95 % CI 2.035–169.526, p = 0.010) or a BARD score C2 (odds ratio 4.040, 95 % CI 1.084–15.048, p = 0.037). Conclusion The PNPLA3 rs738409 polymorphism is common and may confer a significant risk of NAFLD and advanced liver fibrosis in the Korean population. Keywords Nonalcoholic fatty liver disease Patatin-like phospholipase domain-containing 3 Single-nucleotide polymorphism Fibrosis Korea
Introduction Nonalcoholic fatty liver disease (NAFLD) includes a wide spectrum of liver disease, ranging from simple steatosis to
123
2968
nonalcoholic steatohepatitis (NASH), NASH-associated liver cirrhosis, and hepatocellular carcinoma (HCC) [1]. Although hepatitis B virus (HBV) infection has been the major cause of liver disease in many Asian countries, NAFLD began to emerge as the major liver disease with rise in global obesity and the diabetes epidemic [2]. The recent prevalence of NAFLD is 20–30 % in the Western adult population and 10–25 % in the Asian adult population [3–5]. Prevalence of NASH in the general population has been reported to be 3–5 %, and it can progress to cirrhosis in 15–25 % of cases [6, 7]. In addition to the environmental factors related to obesity and insulin resistance, genetic factors are important influences on the progression of NAFLD. A genome-wide association study revealed that the rs738409 single-nucleotide polymorphism (SNP) in patatin-like phospholipase domain-containing 3 (PNPLA3) is strongly associated with hepatic fat content [8]. PNPLA3, alternatively referred to as adiponutrin, encodes a 481-amino acid protein that is highly expressed in the liver and plays a role in the hydrolysis of triglycerides [9, 10]. The rs738409 SNP in the PNPLA3 gene is a missense variation, in particular an isoleucine-to-methionine substitution at amino acid 148 (I148M) that results in the abolishment of triglyceride lipase activity, which interferes with hepatic triglyceride hydrolysis, leading to decreased incorporation into very low-density lipoprotein (VLDL) and increased intracellular triglyceride (TG) content [11, 12]. Recently, Kumari et al. [13] showed that PNPLA3 I148M substitution increases TG synthesis by converting lysophosphatidic acid (LPA) into phosphatidic acid, leading to hepatic steatosis. However, the mechanism by which the I148M variant of PNPLA3 affects hepatic steatosis remains unclear. The risk allele (G allele) frequencies of the rs738409 SNP are concordant with the prevalence of NAFLD in different ethnic groups, such as Hispanics, Europeans, African Americans, and Asians, including Japanese and Taiwanese populations. However, there has been no report on the genotype frequency and effects of rs738409 in Korean NAFLD patients, for whom a rapid change to a Westernized lifestyle has occurred. The genetic epidemiology of NAFLD is of particular interest, considering that interaction between genetic factors and environmental factors might be a key to intervening in or preventing NAFLD-related mortality in certain high-risk populations. This study investigated the genotype frequency of rs738409 in PNPLA3 in a healthy Korean population of health check examinees (control group) and in a group of NAFLD patients to evaluate the role of the rs738409 SNP in PNPLA3 in the development of NAFLD and NAFLDrelated liver fibrosis.
123
Dig Dis Sci (2014) 59:2967–2974
Methods Study Populations All subjects were ethnic Koreans. The study protocol was approved by the Institutional Review Board of the Seoul National University Bundang Hospital and is in accordance with the principles of the Declaration of Helsinki 1975. A total of 389 Korean adults, including 155 NAFLD patients with a higher alanine aminotransferase (ALT) level than the upper normal limit and 184 healthy controls, who were consecutively enrolled in the health promotion center of the same hospital during 2004–2005, were included in the study. The clinical diagnosis of NAFLD was based on the following criteria: (1) the absence of serologic or clinical evidence of hepatitis B and hepatitis C viral infection; (2) alcohol consumption \20 g/day; (3) exclusion of other liver diseases, such as autoimmune hepatitis, drug-induced hepatitis, hemochromatosis, Wilson disease, intestinal bypass surgery, primary biliary cirrhosis, and primary sclerosing cholangitis; and (4) radiological findings of fatty liver, including hepatomegaly, diffuse increase in echogenicity of the liver parenchyma, and vascular blunting on ultrasound. The 184 healthy control subjects were selected from a total of 739 health check examinees enrolled during the same period based on the following criteria: (1) no fatty liver on ultrasound examination, (2) normal results in biochemical liver function tests, (3) alcohol consumption \20 g/day, and (4) no components of metabolic syndrome such as diabetes, hyperlipidemia, and hypertension. Blood samples were stored at -70 °C until analyzed. The clinical data for the analysis included age, sex, height, body weight, the presence of hypertension or diabetes, laboratory results on complete blood count, HBV surface antigen (HBsAg), anti-HBV surface antibody (antiHBs), anti-hepatitis C virus antibody (anti-HCV), iron, ferritin, TG, total cholesterol, low-density lipoprotein cholesterol (LDL), high-density lipoprotein cholesterol (HDL), total protein, serum albumin, ALT, aspartate aminotransferase (AST), c-glutamyl transpeptidase (GGT), alkaline phosphatase (ALP), platelet count, and ultrasonographic findings of the abdomen. Evaluation of Liver Fibrosis For the evaluation of liver fibrosis, the NAFLD fibrosis score (NFS), BARD score, and AST/platelet ratio index (APRI) were calculated [14–17]. The NFS formula is -1.675 ? 0.037 9 age (years) ? 0.094 9 body mass index (BMI, kg/m2) ? 1.13 9 impaired fasting glucose/ diabetes (yes, 1; no, 0) ? 0.99 9 ALT/AST ratio– 0.013 9 platelets (109/L)–0.66 9 albumin (g/dl). Among
Dig Dis Sci (2014) 59:2967–2974
the 155 NAFLD patients, 11 patients did not have complete data required for the calculation of NFS. Therefore, 144 NAFLD patients were divided into two groups according to the probability of advanced liver fibrosis. Patients with NFS \ -1.455 were classified as ‘‘low NFS group’’ (low probability of advanced liver fibrosis) and those with NFS C -1.455 were classified as ‘‘intermediate or high NFS group’’ (intermediate or high probability of advanced liver fibrosis) [14]. The BARD score is composed of 3 variables: BMI C 28 kg/m2 is assigned 1 point, AST/ALT ratio C0.8 is assigned 2 points, and type 2 diabetes mellitus is assigned 1 point. According to Harrison et al. [15], a BARD score of 2–4 points is associated with advanced fibrosis. To distinguish between NAFLD patients with and without advanced liver fibrosis, we divided the 155 NAFLD patients into ‘‘high BARD score group’’ (BARD score C 2) and ‘‘low BARD score group’’ (BARD score \ 2). APRI is calculated by using the formula: (AST/ upper limit of normal 9 100)/platelet count [16]. PNPLA3 rs738409 Genotyping Whole-blood samples collected in an EDTA bottle were used for genomic DNA extraction. PNPLA3 rs738409 genotyping was performed with 20 ng of DNA using the TaqMan polymerase chain reaction (PCR) method (Applied Biosystems, Foster City, CA, USA). The thermal cycling conditions were 95 °C for 10 min, 40 cycles for 15 s at 95 °C, and 1 min at 60 °C. To ensure optimal analysis, internal controls with known genotype and negative controls with water were used. The success rate for genotyping was 100 %. The genotype frequency was consistent with Hardy–Weinberg equilibrium for both, the controls (p = 0.896) and NAFLD patients (p = 0.813). The genotype analysis was based on either an additive model (CC, CG, and GG type) or dominant model (CC vs. CG?GG type) of inheritance. Statistical Analyses Continuous data are presented as the mean ± standard deviation (SD) and were compared using Mann–Whitney U test and Kruskal–Wallis test, as appropriate. Categorical variables were compared using Fisher’s exact test and a chi-square test. Genotype frequency in cases and controls was analyzed using the Cochran–Armitage trend test. The determined genotype frequencies were tested for consistency with Hardy–Weinberg equilibrium in both, the controls and cases, respectively. The association between PNPLA3 genotype and the presence of NAFLD or liver fibrosis was evaluated by multivariate logistic regression analysis after adjusting for potential confounding variables. Risk analysis was performed by calculating the odds ratio
2969
(OR) and the 95 % confidence interval (CI). All tests were two-tailed, with a significant p value defined as \0.05. All statistical operations were performed using PASW software version 18 (SPSS, Chicago, IL, USA).
Results PNPLA3 rs738409 Genotype Frequency in Healthy Controls and NAFLD Patients The demographic and clinical features of healthy controls and NAFLD patients are summarized in Table 1. The proportion of men among the NAFLD patients was higher than that in the healthy controls (p \ 0.001). However, there was no significant difference in the genotypes between males and females in either NAFLD patients (p = 0.924) or healthy controls (p = 0.573) (data not shown). The PNPLA3 rs738409 genotype frequency in healthy controls was significantly different from that in NAFLD patients according to both, the additive and the dominant models. Moreover, the frequency of the risk allele, G allele, was significantly higher in NAFLD patients than in healthy controls (OR 1.534, 95 % CI 1.129–2.085, p = 0.006) (Table 2). To confirm the association between the PNPLA3 rs738409 genotype and development of NAFLD, univariate and multivariate analyses were performed. The genotype analysis was based on either the additive model (CC, CG, and GG type) or dominant model (CC vs. CG?GG type) of inheritance. The results revealed that the PNPLA3 SNP was an independent factor associated with NAFLD, along with BMI [ 25 kg/m2, male gender and serum TG [ 130 mg/dL (Table 3). To investigate the clinical variables that might be associated with the genotypes of rs738409, we compared age; gender; BMI; frequency of hypertension, diabetes, and hyperlipidemia; and serum ALT levels among the different genotypes in the NAFLD cases and controls. However, there was no significant factor associated with the genotype (data not shown). Association Between PNPLA3 rs738409 Genotype and Liver Fibrosis in NAFLD Patients To evaluate the association of PNPLA3 genotype and liver fibrosis, the genotype frequency was compared between the low NFS group (NFS \ -1.455, n = 104, 72.2 %) and the intermediate or high NFS group (NFS C -1.455, n = 40, 27.8 %) (Fig. 1a). The intermediate or high NFS group revealed a significantly higher frequency of CG?GG genotypes than the low NFS group (p = 0.012) in the dominant model of inheritance, suggesting that CG?GG
123
2970
Dig Dis Sci (2014) 59:2967–2974
Table 1 Baseline characteristics of the healthy controls and the patients with nonalcoholic fatty liver disease (n = 339)
Age, mean ± SD (years) Gender, male BMI, mean ± SD (kg/m2) Hypertension Diabetes Cholesterol, mean ± SD (mg/dL) TG, mean ± SD (mg/dL) Uric acid, mean ± SD (mg/dL) Platelet, mean ± SD (9103/uL) ALP, mean ± SD (IU/L) AST, mean ± SD (IU/L) ALT, mean ± SD (IU/L) GGT, mean ± SD (IU/L)
Healthy controls (n = 184)
NAFLD patients (n = 155)
p value
45.3 ± 10.6 60 (32.6) 22.1 ± 2.6 0 (0) 0 (0) 190.1 ± 28.8 89.8 ± 54.7 4.8 ± 1.3 241.0 ± 52.0 58.9 ± 14.1 21.1 ± 5.4 18.1 ± 7.3 21.6 ± 18.3
44.1 ± 15.5 101 (65.2) 27.5 ± 3.7 46 (29.7) 42 (27.1) 206.0 ± 45.1 179.9 ± 86.3 6.2 ± 1.8 226.9 ± 69.6 85.1 ± 39.0 55.9 ± 33.1 93.0 ± 63.2 70.9 ± 45.9
0.361 \0.001 \0.001 \0.001 \0.001 \0.001 \0.001 \0.001 0.068 \0.001 \0.001 \0.001 \0.001
Data are presented as the mean ± SD for continuous data and percentages for categorical data The p values were obtained by comparing the quantitative phenotype between the two groups (Mann–Whitney U test). Gender, diabetes, and hypertension were analyzed using the chi-square test NAFLD nonalcoholic fatty liver disease, BMI body mass index, TG triglyceride, ALP alkaline phosphatase, AST aspartate transaminase, ALT alanine transaminase, GGT c-glutamyl transpeptidase
Table 2 Genotype and allele frequencies of PNPLA3 I148M polymorphism in healthy controls and nonalcoholic fatty liver disease patients Genotype
Healthy controls (n = 184)
NAFLD patients (n = 155)
p value
Genotypes CC/CG/GG (%) Dominant model CC/CG?GG (%) Alleles C/G (%)
55/92/37 (29.9/50.0/20.1) 55/129 (29.9/70.1) 202/166 (54.9/45.1)
31/75/49 (20.0/48.4/31.6) 31/124 (20.0/80.0) 137/173 (44.2/55.8)
0.006a 0.037b 0.006c
NAFLD nonalcoholic fatty liver disease a Comparisons for genotypes performed using the Cochran–Armitage trend test b Comparisons were performed using the chi-square test for categorical variables c Comparisons were performed using the additive model Table 3 Univariate and multivariate analyses of risk factors associated with nonalcoholic fatty liver disease (n = 339) Variables
Univariate OR
Multivariate p
Dominant modela OR (95 % CI)
PNPLA3 rs738409 genotype CG or GG genotype 1.705 CC/CG/GG genotype 1.534 Age, [46 years 1.136 Gender, male 3.865 18.545 BMI, [25 (kg/m2) TG [ 130 (mg/dL) 14.668
0.038 0.006 0.564 \0.001 \0.001 \0.001
2.568 (1.109–5.945) 1.427 2.159 11.065 11.365
(0.680–2.993) (1.027–4.537) (5.359–22.847) (5.443–23.732)
CI confidence interval, OR odds ratio, BMI body mass index, TG triglyceride a CG or GG genotype for rs738409 b Odds ratio per number of variant 148 M alleles
123
Additive modelb p
OR (95 % CI)
p
1.985 1.386 2.249 10.741 11.349
0.008 0.389 0.033 \0.001 \0.001
0.028 0.347 0.042 \0.001 \0.001
(1.200–3.283) (0.659–2.912) (1.068-4.734) (5.194–22.215) (5.417–23.777)
Dig Dis Sci (2014) 59:2967–2974
2971
Discussion
Fig. 1 Comparison of the genotype frequency of PNPLA3 rs738409 between the advanced fibrosis group and less advanced fibrosis group according to the nonalcoholic fatty liver disease fibrosis score (NFS) or BARD score in nonalcoholic fatty liver disease patients. a In 144 nonalcoholic fatty liver patients, the intermediate or high NFS group (NFS C -1.455) displayed a higher frequency of CG?GG genotypes than the low NFS group (NFS \ -1.455, p = 0.012). b In 155 nonalcoholic fatty liver patients, the high BARD score group (BARD score C 2) displayed a higher frequency of CG?GG genotypes than the low BARD score group (BARD score \ 2, p = 0.046). The p value is calculated using a chi-square test. NFS nonalcoholic fatty liver disease fibrosis score
genotype is associated with advanced fibrosis. Moreover, the high BARD score group (BARD score C2, n = 48) displayed a significantly higher frequency of CG?GG genotypes than the low BARD score group (BARD score \2, n = 107, p = 0.046) (Fig. 1b). To confirm this finding, univariate and multivariate logistic regression analyses were performed. The multivariable analysis indicated that the presence of the CG?GG genotype of rs738409 was an independent factor associated with advanced liver fibrosis in NAFLD, regardless of using the criteria of NFS or BARD score (Tables 4, 5). However, the significance was observed only in the dominant model and not in the additive model. On the other hand, no significant associations between APRI and genotype of rs738409 were observed (data not shown).
In this study, the frequency of the PNPLA3 rs738409 G allele was 0.45 in the controls and 0.56 in NAFLD patients in the Korean study population, which suggests that there was a significant association between this SNP and the development of NAFLD (p = 0.006). In comparison with the CC genotype, the CG?GG genotype under a dominant model raised the risk of NAFLD 2.6-fold after adjustment for age, gender, BMI, and TG level (95 % CI 1.109–5.945, p = 0.028), whereas, under an additive model, the OR for NAFLD was 1.9 (95 % CI 1.200–3.283, p = 0.008). Moreover, using the indirect serum markers of fibrosis, CG?GG genotype was an independent factor for advanced liver fibrosis in NAFLD patients. Therefore, the PNPLA3 SNP is significantly related to the development of NAFLD as well as to the development of liver fibrosis in Korean NAFLD patients. A recent meta-analysis revealed that the rs738409 I148M variant of PNPLA3 is a strong modifier of the natural history of NAFLD in different ethnic populations around the world [18]. However, only 2 studies from the Asian region, one from Taiwan [19] (pediatric population) and the other from Japan [20], were included in the metaanalysis. Moreover, there was no report on the genotype frequency or effect of the SNP in the development of NAFLD-associated fibrosis in Korean population. In the population without NAFLD, the reported G-allele frequency of rs738409 PNPLA3 is 0.17 in African Americans, 0.23 in Europeans, 0.40 in the Taiwanese, 0.44 in Japanese, and 0.49 in the Hispanics [8, 20, 21]. The G-allele frequency of 0.45 in the Korean healthy controls of this study is comparable to that of examined Japanese populations, demonstrating the high-frequency range among various ethnic groups. The prevalence of NAFLD is concordant with the frequencies of the PNPLA3 G allele in American, Argentinean, and German populations [22–24]. Moreover, this SNP was identified as a risk factor of NAFLD in normoglycemic Taiwanese and obese pediatric populations [19, 21], and in a Japanese population [20]. In the present study, the G-allele frequency in NAFLD patients was significantly higher than in healthy controls, and the PNPLA3 genotype based on both, the dominant and additive model, was an independent factor for NAFLD according to the multivariate analysis. Therefore, the carriers of the PNPLA3 rs738409 G allele were susceptible to NAFLD in the Korean population, which is concordant with the results of studies of various ethnic groups. A previous meta-analysis revealed the association of rs738409 with increased ALT levels in NAFLD, but there was no significant association with insulin resistance phenotype, such as HOMR-IR, fasting insulin level, or fasting
123
2972
Dig Dis Sci (2014) 59:2967–2974
Table 4 Univariate and multivariate analyses of predictors of advanced liver fibrosis (nonalcoholic fatty liver disease fibrosis score C -1.455) in patients with nonalcoholic fatty liver disease according to nonalcoholic fatty liver disease fibrosis score Variables
Univariate OR
Multivariate Dominant modela
p
OR (95 % CI)
Additive modelb p
OR (95 % CI)
p
PNPLA3 rs738409 genotype CG or GG genotype
5.700
0.014
CC/CG/GG genotype
1.656
0.070
Age, [46 years
9.000
\0.001
0.405
0.020
Gender, male 2
18.573 (2.035–169.526)
0.010
31.569 (5.974–166.831)
\0.001
1.765 (0.823–3.785) 1.596 (0.422–6.041)
0.491
26.526 (5.570–126.334)
0.144 \0.001
1.858 (0.496–6.966)
0.358
BMI, [25 (kg/m )
2.191
0.131
3.810 (0.991–14.650)
0.052
3.092 (0.857–11.148)
0.085
Hypertension
5.133
\0.001
1.900 (0.643–83.271)
0.246
1.779 (0.615–5.145)
0.288
Diabetes
9.889
\0.001
21.301 (5.449–83.271)
\0.001
15.788 (4.644–53.669)
\0.001
Available data in 144 patients NAFLD nonalcoholic fatty liver disease, CI confidence interval, OR odds ratio, BMI body mass index, TG triglyceride a CG or GG genotype for rs738409 b
Odds ratio per number of variant 148 M alleles
Table 5 Univariate and multivariate analyses of predictors of advanced liver fibrosis (BARD score C 2) in patients with nonalcoholic fatty liver disease according to BARD score (n = 155) Variables
Univariate OR
Multivariate Dominant modela
p
OR (95 % CI)
Additive modelb p
OR (95 % CI)
p
PNPLA3 rs738409 genotype CG or GG genotype
2.760
0.052
CC/CG/GG genotype
1.483
0.118
Age, [ 46 years
5.214
Gender, male
0.263 2
4.040 (1.084–15.048)
0.037 1.426 (0.774–2.628)
0.256
\0.001
2.536 (0.825–7.792)
0.104
2.726 (0.913–8.142)
0.072
\0.001
0.245 (0.075–0.799)
0.020
0.284 (0.091–0.890)
0.031
BMI, [ 25 (kg/m )
1.859
0.163
3.151 (1.001–9.921)
0.050
2.790 (0.904–8.610)
0.074
Hypertension
3.864
\0.001
1.715 (0.664–4.428)
0.265
1.633 (0.639–4.171)
0.305
Diabetes
7.886
\0.001
8.779 (3.364–22.915)
\0.001
8.502 (3.259–22.175)
\0.001
NAFLD nonalcoholic fatty liver disease, CI confidence interval, OR odds ratio, BMI body mass index, TG triglyceride a b
CG or GG genotype for rs738409 odds ratio per number of variant 148 M alleles
glucose level [18]. However, there were no differences in age, gender, BMI, diabetes, hypertension, and ALT level among the 3 genotypes in both, the controls and NAFLD patients, in our study; this was most likely because of the small size or different subject profiles. Recently, several noninvasive scoring systems have been proposed to predict advanced liver fibrosis without invasive liver biopsy, such as the NFS, BARD score, and APRI. The NFS is composed of 6 easily available variables and can accurately predict the presence or absence of advanced fibrosis in NAFLD patients. According to the previous results [14], the low cutoff score (-1.455) had an 88–93 % negative predictive value to exclude advanced
123
fibrosis, whereas the high cutoff (0.676) had an 82–90 % positive predictive value in identifying the presence of advanced fibrosis. A recent meta-analysis showed that NFS has been independently validated in different ethnicities including American Caucasians, blacks, and Hispanics, Asians, and Europeans [25]. In addition, Chang et al. [26] showed that NAFLD and its severity using NFS was associated with increased risk of diabetes in a healthy Korean population. The BARD score is another simple noninvasive scoring system for assessing the severity of fibrosis in NAFLD. A BARD score of 0 or 1 point had a 96 % negative predictive value (96 %) to identify patients without advanced
Dig Dis Sci (2014) 59:2967–2974
liver fibrosis, whereas the positive predictive value of a high BARD score (2–4 point) was not very good (43 %) [15]. In a study by Ruffillo et al. [27], both scoring systems had higher specificity than sensitivity. By using the NFS and BARD score in the present study, we demonstrated that there was a significant association between the CG?GG genotype of rs738409 and fibrosis of NAFLD, and the CG?GG genotype was an independent predictor of advanced liver fibrosis in NAFLD in the multivariate analysis. This finding suggests that the influence of the I148M variant on the susceptibility to advanced liver fibrosis may follow a dominant model. Consistent with our study, a previous meta-analysis revealed that the influence of PNPLA3 genotypes on NAFLD followed an additive genetic model, except for liver fibrosis, which followed a dominant genetic model [18]. APRI has been proposed as a noninvasive tool for the assessment of liver fibrosis in chronic liver disease [16]; however, we could not find a significant association between the PNPLA3 genotype and APRI score. In general, approximately 15–25 % of patients with NASH progress to cirrhosis [7]. The G allele of rs738409 in PNPLA3 is associated not only with histologic steatosis but also with fibrosis [28–30]. The association of this SNP with advanced liver fibrosis and steatosis was observed in previous Caucasian, Argentinean, Caucasian, and Asian populations [23, 28, 29, 31]. Moreover, several new studies in the Asian showed that the PNPLA3 rs738409 variant is strongly associated with the severity of NAFLD [32–34]. A recent meta-analysis revealed that PNPLA3 is associated with an increased risk of both advanced fibrosis and HCC among patients with liver disease [35]. On the other hand, the PNPLA3 rs738409 variant was strongly associated with alcoholic cirrhosis, which shares many similar features with NAFLD-related cirrhosis [36, 37], and with HCVrelated cirrhosis [38]. The I148M PNPLA3 variant influences the liver damage of steatohepatitis in chronic liver disease with metabolic, alcoholic, and viral risk factors [39]. This is the first study to elucidate a significant role of PNPLA3 rs738409 in the development of NAFLD and fibrosis in NAFLD in the Korean population. However, it has several limitations. First, this was a single-center, cross-sectional study with a relatively small sample size. Second, only 10 patients with NAFLD underwent liver biopsy, and we did not measure insulin resistance or quantitative intrahepatic fat content in the subject, so that the relationship between the SNP and those indices could not be investigated. Moreover, only 1 SNP of PNPLA3 (rs738409) was evaluated in this study, although 4 new SNPs were reported in recent papers [40]. Lastly, the study population were enrolled about 10 years age, so that some
2973
clinical features such BMI may not exactly reflect current status. Conclusion In conclusion, the G-allele frequency of the PNPLA3 SNP rs738409 in the healthy Korean population was 0.45, and the carriers of the PNPLA3 rs738409 G allele were susceptible to NAFLD. The CG?GG genotype may confer a significant risk of fibrosis in Korean patients with NAFLD. Therefore, NAFLD patients carrying a risk genotype may be a priority target group for intensive lifestyle modifications and therapeutic trials. Acknowledgments This study was supported by a research grant from an intramural fund of Seoul National University Bundang Hospital (No. 02-2012-046). Conflict of interest
Nothing to declare for all authors.
References 1. Angulo P. Nonalcoholic fatty liver disease. N Engl J Med. 2002;346:1221–1231. 2. Chae HB, Kim JH, Kim JK, Yim HJ. Current status of liver diseases in Korea: hepatitis B. Korean J Hepatol. 2009;15:S13– S24. 3. Park SH, Jeon WK, Kim SH, et al. Prevalence and risk factors of non-alcoholic fatty liver disease among Korean adults. J Gastroenterol Hepatol. 2006;21:138–143. 4. Park SH. Current status of liver disease in Korea: nonalcoholic fatty liver disease. Korean J Hepatol. 2009;15:S34–S39. 5. Chitturi S, Wong VW, Farrell G. Nonalcoholic fatty liver in Asia: firmly entrenched and rapidly gaining ground. J Gastroenterol Hepatol. 2011;26:163–172. 6. Vernon G, Baranova A, Younossi ZM. Systematic review: the epidemiology and natural history of non-alcoholic fatty liver disease and non-alcoholic steatohepatitis in adults. Aliment Pharmacol Ther. 2011;34:274–285. 7. Sanyal AJ. American gastroenterological A. AGA technical review on nonalcoholic fatty liver disease. Gastroenterology. 2002;123:1705–1725. 8. Romeo S, Kozlitina J, Xing C, et al. Genetic variation in PNPLA3 confers susceptibility to nonalcoholic fatty liver disease. Nat Genet. 2008;40:1461–1465. 9. Baulande S, Lasnier F, Lucas M, Pairault J. Adiponutrin, a transmembrane protein corresponding to a novel dietary- and obesity-linked mRNA specifically expressed in the adipose lineage. J Biol Chem. 2001;276:33336–33344. 10. Lake AC, Sun Y, Li JL, et al. Expression, regulation, and triglyceride hydrolase activity of Adiponutrin family members. J Lipid Res. 2005;46:2477–2487. 11. He S, McPhaul C, Li JZ, et al. A sequence variation (I148M) in PNPLA3 associated with nonalcoholic fatty liver disease disrupts triglyceride hydrolysis. J Biol Chem. 2010;285:6706–6715. 12. Pirazzi C, Adiels M, Burza MA, et al. Patatin-like phospholipase domain-containing 3 (PNPLA3) I148M (rs738409) affects hepatic VLDL secretion in humans and in vitro. J Hepatol. 2012;57:1276–1282.
123
2974 13. Kumari M, Schoiswohl G, Chitraju C, et al. Adiponutrin functions as a nutritionally regulated lysophosphatidic acid acyltransferase. Cell Metabol. 2012;15:691–702. 14. Angulo P, Hui JM, Marchesini G, et al. The NAFLD fibrosis score: a noninvasive system that identifies liver fibrosis in patients with NAFLD. Hepatol. 2007;45:846–854. 15. Harrison SA, Oliver D, Arnold HL, Gogia S, Neuschwander-Tetri BA. Development and validation of a simple NAFLD clinical scoring system for identifying patients without advanced disease. Gut. 2008;57:1441–1447. 16. Wai CT, Greenson JK, Fontana RJ, et al. A simple noninvasive index can predict both significant fibrosis and cirrhosis in patients with chronic hepatitis C. Hepatology. 2003;38:518–526. 17. Angulo P, Bugianesi E, Bjornsson ES, et al. Simple noninvasive systems predict long-term outcomes of patients with nonalcoholic fatty liver disease. Gastroenterology. 2013;145:e784. 18. Sookoian S, Pirola CJ. Meta-analysis of the influence of I148M variant of patatin-like phospholipase domain containing 3 gene (PNPLA3) on the susceptibility and histological severity of nonalcoholic fatty liver disease. Hepatology. 2011;53:1883–1894. 19. Lin YC, Chang PF, Hu FC, Yang WS, Chang MH, Ni YH. A common variant in the PNPLA3 gene is a risk factor for nonalcoholic fatty liver disease in obese Taiwanese children. J Pediatrics. 2011;158:740–744. 20. Hotta K, Yoneda M, Hyogo H, et al. Association of the rs738409 polymorphism in PNPLA3 with liver damage and the development of nonalcoholic fatty liver disease. BMC Med Genet. 2010;11:172. 21. Wang CW, Lin HY, Shin SJ, et al. The PNPLA3 I148M polymorphism is associated with insulin resistance and nonalcoholic fatty liver disease in a normoglycaemic population. Liver Int. 2011;31:1326–1331. 22. Kotronen A, Johansson LE, Johansson LM, et al. A common variant in PNPLA3, which encodes adiponutrin, is associated with liver fat content in humans. Diabetologia. 2009;52:1056–1060. 23. Sookoian S, Castano GO, Burgueno AL, Gianotti TF, Rosselli MS, Pirola CJ. A nonsynonymous gene variant in the adiponutrin gene is associated with nonalcoholic fatty liver disease severity. J Lipid Res. 2009;50:2111–2116. 24. Kantartzis K, Peter A, Machicao F, et al. Dissociation between fatty liver and insulin resistance in humans carrying a variant of the patatin-like phospholipase 3 gene. Diabetes. 2009;58:2616–2623. 25. Musso G, Gambino R, Cassader M, Pagano G. Meta-analysis: natural history of non-alcoholic fatty liver disease (NAFLD) and diagnostic accuracy of non-invasive tests for liver disease severity. Ann Med. 2011;43:617–649. 26. Chang Y, Jung HS, Yun KE, Cho J, Cho YK, Ryu S. Cohort study of non-alcoholic fatty liver disease, NAFLD fibrosis score, and the risk of incident diabetes in a Korean population. Am J Gastroenterol. 2013;108:1861–1868.
123
Dig Dis Sci (2014) 59:2967–2974 27. Ruffillo G, Fassio E, Alvarez E, et al. Comparison of NAFLD fibrosis score and BARD score in predicting fibrosis in nonalcoholic fatty liver disease. J Hepatol. 2011;54:160–163. 28. Valenti L, Al-Serri A, Daly AK, et al. Homozygosity for the patatin-like phospholipase-3/adiponutrin I148M polymorphism influences liver fibrosis in patients with nonalcoholic fatty liver disease. Hepatology. 2010;51:1209–1217. 29. Rotman Y, Koh C, Zmuda JM, Kleiner DE, Liang TJ, Nash CRN. The association of genetic variability in patatin-like phospholipase domain-containing protein 3 (PNPLA3) with histological severity of nonalcoholic fatty liver disease. Hepatology. 2010;52:894–903. 30. Speliotes EK, Butler JL, Palmer CD, et al. PNPLA3 variants specifically confer increased risk for histologic nonalcoholic fatty liver disease but not metabolic disease. Hepatology. 2010;52:904–912. 31. Li Y, Xing C, Cohen JC, Hobbs HH. Genetic variant in PNPLA3 is associated with nonalcoholic fatty liver disease in China. Hepatology. 2012;55:327–328. 32. Kitamoto T, Kitamoto A, Yoneda M, 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. 2013;132:783–792. 33. Zain SM, Mohamed R, Mahadeva S, et al. A multi-ethnic study of a PNPLA3 gene variant and its association with disease severity in non-alcoholic fatty liver disease. Hum Genet. 2012;131:1145–1152. 34. Kawaguchi T, Sumida Y, Umemura A, et al. Genetic polymorphisms of the human PNPLA3 gene are strongly associated with severity of non-alcoholic fatty liver disease in Japanese. PloS One. 2012;7:e38322. 35. Singal AG, Manjunath H, Yopp AC, et al. The effect of PNPLA3 on fibrosis progression and development of hepatocellular carcinoma: a meta-analysis. Am J Gastroenterol. 2014;109:325–334. 36. Tian C, Stokowski RP, Kershenobich D, Ballinger DG, Hinds DA. Variant in PNPLA3 is associated with alcoholic liver disease. Nat Genet. 2010;42:21–23. 37. Stickel F, Buch S, Lau K, et al. Genetic variation in the PNPLA3 gene is associated with alcoholic liver injury in caucasians. Hepatology. 2011;53:86–95. 38. Valenti L, Fargion S. Patatin-like phospholipase domain containing-3 Ile148 Met and fibrosis progression after liver transplantation. Hepatology. 2011;54:1484. 39. Valenti L, Alisi A, Nobili V. I148M PNPLA3 variant and progressive liver disease: A new paradigm in hepatology. Hepatology 2012. 40. Speliotes EK, Yerges-Armstrong LM, Wu J, et al. Genome-wide association analysis identifies variants associated with nonalcoholic fatty liver disease that have distinct effects on metabolic traits. PLoS Genet. 2011;7:e1001324.
ORIGINAL ARTICLE
Role of the PNPLA3 I148M Polymorphism in Nonalcoholic Fatty Liver Disease and Fibrosis in Korea Sang Soo Lee • Young-Sang Byoun • Sook-Hyang Jeong • Byung Hyun Woo • Eun Sun Jang • Jin-Wook Kim • Hyun Young Kim
Received: 17 January 2014 / Accepted: 2 July 2014 / Published online: 29 July 2014 Ó Springer Science+Business Media New York 2014
Abstract Background The role of the patatin-like phospholipase domain-containing 3 (PNPLA3) single-nucleotide polymorphism (SNP), rs738409, in the development and progression of nonalcoholic fatty liver disease (NAFLD) has not been studied in the Korean population. Aims The aim of the study was to investigate the genotype frequency and allele distribution of PNPLA3 rs738409
Sang Soo Lee and Young-Sang Byoun have equally contributed to this work. S. S. Lee Y.-S. Byoun S.-H. Jeong (&) B. H. Woo E. S. Jang J.-W. Kim Department of Internal Medicine, Seoul National University College of Medicine, Seoul National University Bundang Hospital, 82 Gumi-ro, 173 Beon-gil, Bundang-gu, Seongnam-si 463 707, Gyeonggi-do, Republic of Korea e-mail: [email protected] S. S. Lee e-mail: [email protected] Y.-S. Byoun e-mail: [email protected] B. H. Woo e-mail: [email protected] E. S. Jang e-mail: [email protected] J.-W. Kim e-mail: [email protected] H. Y. Kim Health Promotion Center, Seoul National University College of Medicine, Seoul National University Bundang Hospital, 82 Gumi-ro, 173 Beon-gil, Bundang-gu, Seongnam-si, Gyeonggi-do, Republic of Korea e-mail: [email protected]
and the association between the SNP and development of NAFLD and liver fibrosis. Methods A total of 339 Korean adults (155 NAFLD patients and 184 healthy controls) were enrolled. PNPLA3 SNP genotyping was carried out using a TaqMan allelic discrimination assay. Liver fibrosis severity was evaluated by NAFLD fibrosis score (NFS) and BARD score. Results The frequencies of the PNPLA3 rs738409 genotypes, CC, CG, and GG in the healthy control group were 29.9, 50.0, and 20.1 %, respectively, and those in NAFLD patients were 20.0, 48.4, and 31.6 %, respectively, showing a higher frequency of the risk allele (G allele) (p = 0.006). Among the NAFLD patients, the CG?GG genotype frequency was significantly higher in patients with advanced fibrosis, defined as NFS C -1.455 or BARD score C2, than in patients with mildto-moderate fibrosis (p = 0.012 and p = 0.046, respectively). In multivariate analysis, the CG?GG genotype was an independent factor for NAFLD development (odds ratio 2.568, 95 % CI 1.109–5.945, p = 0.028) and for advanced liver fibrosis according to the criteria of NFS C -1.455 (odds ratio 18.573, 95 % CI 2.035–169.526, p = 0.010) or a BARD score C2 (odds ratio 4.040, 95 % CI 1.084–15.048, p = 0.037). Conclusion The PNPLA3 rs738409 polymorphism is common and may confer a significant risk of NAFLD and advanced liver fibrosis in the Korean population. Keywords Nonalcoholic fatty liver disease Patatin-like phospholipase domain-containing 3 Single-nucleotide polymorphism Fibrosis Korea
Introduction Nonalcoholic fatty liver disease (NAFLD) includes a wide spectrum of liver disease, ranging from simple steatosis to
123
2968
nonalcoholic steatohepatitis (NASH), NASH-associated liver cirrhosis, and hepatocellular carcinoma (HCC) [1]. Although hepatitis B virus (HBV) infection has been the major cause of liver disease in many Asian countries, NAFLD began to emerge as the major liver disease with rise in global obesity and the diabetes epidemic [2]. The recent prevalence of NAFLD is 20–30 % in the Western adult population and 10–25 % in the Asian adult population [3–5]. Prevalence of NASH in the general population has been reported to be 3–5 %, and it can progress to cirrhosis in 15–25 % of cases [6, 7]. In addition to the environmental factors related to obesity and insulin resistance, genetic factors are important influences on the progression of NAFLD. A genome-wide association study revealed that the rs738409 single-nucleotide polymorphism (SNP) in patatin-like phospholipase domain-containing 3 (PNPLA3) is strongly associated with hepatic fat content [8]. PNPLA3, alternatively referred to as adiponutrin, encodes a 481-amino acid protein that is highly expressed in the liver and plays a role in the hydrolysis of triglycerides [9, 10]. The rs738409 SNP in the PNPLA3 gene is a missense variation, in particular an isoleucine-to-methionine substitution at amino acid 148 (I148M) that results in the abolishment of triglyceride lipase activity, which interferes with hepatic triglyceride hydrolysis, leading to decreased incorporation into very low-density lipoprotein (VLDL) and increased intracellular triglyceride (TG) content [11, 12]. Recently, Kumari et al. [13] showed that PNPLA3 I148M substitution increases TG synthesis by converting lysophosphatidic acid (LPA) into phosphatidic acid, leading to hepatic steatosis. However, the mechanism by which the I148M variant of PNPLA3 affects hepatic steatosis remains unclear. The risk allele (G allele) frequencies of the rs738409 SNP are concordant with the prevalence of NAFLD in different ethnic groups, such as Hispanics, Europeans, African Americans, and Asians, including Japanese and Taiwanese populations. However, there has been no report on the genotype frequency and effects of rs738409 in Korean NAFLD patients, for whom a rapid change to a Westernized lifestyle has occurred. The genetic epidemiology of NAFLD is of particular interest, considering that interaction between genetic factors and environmental factors might be a key to intervening in or preventing NAFLD-related mortality in certain high-risk populations. This study investigated the genotype frequency of rs738409 in PNPLA3 in a healthy Korean population of health check examinees (control group) and in a group of NAFLD patients to evaluate the role of the rs738409 SNP in PNPLA3 in the development of NAFLD and NAFLDrelated liver fibrosis.
123
Dig Dis Sci (2014) 59:2967–2974
Methods Study Populations All subjects were ethnic Koreans. The study protocol was approved by the Institutional Review Board of the Seoul National University Bundang Hospital and is in accordance with the principles of the Declaration of Helsinki 1975. A total of 389 Korean adults, including 155 NAFLD patients with a higher alanine aminotransferase (ALT) level than the upper normal limit and 184 healthy controls, who were consecutively enrolled in the health promotion center of the same hospital during 2004–2005, were included in the study. The clinical diagnosis of NAFLD was based on the following criteria: (1) the absence of serologic or clinical evidence of hepatitis B and hepatitis C viral infection; (2) alcohol consumption \20 g/day; (3) exclusion of other liver diseases, such as autoimmune hepatitis, drug-induced hepatitis, hemochromatosis, Wilson disease, intestinal bypass surgery, primary biliary cirrhosis, and primary sclerosing cholangitis; and (4) radiological findings of fatty liver, including hepatomegaly, diffuse increase in echogenicity of the liver parenchyma, and vascular blunting on ultrasound. The 184 healthy control subjects were selected from a total of 739 health check examinees enrolled during the same period based on the following criteria: (1) no fatty liver on ultrasound examination, (2) normal results in biochemical liver function tests, (3) alcohol consumption \20 g/day, and (4) no components of metabolic syndrome such as diabetes, hyperlipidemia, and hypertension. Blood samples were stored at -70 °C until analyzed. The clinical data for the analysis included age, sex, height, body weight, the presence of hypertension or diabetes, laboratory results on complete blood count, HBV surface antigen (HBsAg), anti-HBV surface antibody (antiHBs), anti-hepatitis C virus antibody (anti-HCV), iron, ferritin, TG, total cholesterol, low-density lipoprotein cholesterol (LDL), high-density lipoprotein cholesterol (HDL), total protein, serum albumin, ALT, aspartate aminotransferase (AST), c-glutamyl transpeptidase (GGT), alkaline phosphatase (ALP), platelet count, and ultrasonographic findings of the abdomen. Evaluation of Liver Fibrosis For the evaluation of liver fibrosis, the NAFLD fibrosis score (NFS), BARD score, and AST/platelet ratio index (APRI) were calculated [14–17]. The NFS formula is -1.675 ? 0.037 9 age (years) ? 0.094 9 body mass index (BMI, kg/m2) ? 1.13 9 impaired fasting glucose/ diabetes (yes, 1; no, 0) ? 0.99 9 ALT/AST ratio– 0.013 9 platelets (109/L)–0.66 9 albumin (g/dl). Among
Dig Dis Sci (2014) 59:2967–2974
the 155 NAFLD patients, 11 patients did not have complete data required for the calculation of NFS. Therefore, 144 NAFLD patients were divided into two groups according to the probability of advanced liver fibrosis. Patients with NFS \ -1.455 were classified as ‘‘low NFS group’’ (low probability of advanced liver fibrosis) and those with NFS C -1.455 were classified as ‘‘intermediate or high NFS group’’ (intermediate or high probability of advanced liver fibrosis) [14]. The BARD score is composed of 3 variables: BMI C 28 kg/m2 is assigned 1 point, AST/ALT ratio C0.8 is assigned 2 points, and type 2 diabetes mellitus is assigned 1 point. According to Harrison et al. [15], a BARD score of 2–4 points is associated with advanced fibrosis. To distinguish between NAFLD patients with and without advanced liver fibrosis, we divided the 155 NAFLD patients into ‘‘high BARD score group’’ (BARD score C 2) and ‘‘low BARD score group’’ (BARD score \ 2). APRI is calculated by using the formula: (AST/ upper limit of normal 9 100)/platelet count [16]. PNPLA3 rs738409 Genotyping Whole-blood samples collected in an EDTA bottle were used for genomic DNA extraction. PNPLA3 rs738409 genotyping was performed with 20 ng of DNA using the TaqMan polymerase chain reaction (PCR) method (Applied Biosystems, Foster City, CA, USA). The thermal cycling conditions were 95 °C for 10 min, 40 cycles for 15 s at 95 °C, and 1 min at 60 °C. To ensure optimal analysis, internal controls with known genotype and negative controls with water were used. The success rate for genotyping was 100 %. The genotype frequency was consistent with Hardy–Weinberg equilibrium for both, the controls (p = 0.896) and NAFLD patients (p = 0.813). The genotype analysis was based on either an additive model (CC, CG, and GG type) or dominant model (CC vs. CG?GG type) of inheritance. Statistical Analyses Continuous data are presented as the mean ± standard deviation (SD) and were compared using Mann–Whitney U test and Kruskal–Wallis test, as appropriate. Categorical variables were compared using Fisher’s exact test and a chi-square test. Genotype frequency in cases and controls was analyzed using the Cochran–Armitage trend test. The determined genotype frequencies were tested for consistency with Hardy–Weinberg equilibrium in both, the controls and cases, respectively. The association between PNPLA3 genotype and the presence of NAFLD or liver fibrosis was evaluated by multivariate logistic regression analysis after adjusting for potential confounding variables. Risk analysis was performed by calculating the odds ratio
2969
(OR) and the 95 % confidence interval (CI). All tests were two-tailed, with a significant p value defined as \0.05. All statistical operations were performed using PASW software version 18 (SPSS, Chicago, IL, USA).
Results PNPLA3 rs738409 Genotype Frequency in Healthy Controls and NAFLD Patients The demographic and clinical features of healthy controls and NAFLD patients are summarized in Table 1. The proportion of men among the NAFLD patients was higher than that in the healthy controls (p \ 0.001). However, there was no significant difference in the genotypes between males and females in either NAFLD patients (p = 0.924) or healthy controls (p = 0.573) (data not shown). The PNPLA3 rs738409 genotype frequency in healthy controls was significantly different from that in NAFLD patients according to both, the additive and the dominant models. Moreover, the frequency of the risk allele, G allele, was significantly higher in NAFLD patients than in healthy controls (OR 1.534, 95 % CI 1.129–2.085, p = 0.006) (Table 2). To confirm the association between the PNPLA3 rs738409 genotype and development of NAFLD, univariate and multivariate analyses were performed. The genotype analysis was based on either the additive model (CC, CG, and GG type) or dominant model (CC vs. CG?GG type) of inheritance. The results revealed that the PNPLA3 SNP was an independent factor associated with NAFLD, along with BMI [ 25 kg/m2, male gender and serum TG [ 130 mg/dL (Table 3). To investigate the clinical variables that might be associated with the genotypes of rs738409, we compared age; gender; BMI; frequency of hypertension, diabetes, and hyperlipidemia; and serum ALT levels among the different genotypes in the NAFLD cases and controls. However, there was no significant factor associated with the genotype (data not shown). Association Between PNPLA3 rs738409 Genotype and Liver Fibrosis in NAFLD Patients To evaluate the association of PNPLA3 genotype and liver fibrosis, the genotype frequency was compared between the low NFS group (NFS \ -1.455, n = 104, 72.2 %) and the intermediate or high NFS group (NFS C -1.455, n = 40, 27.8 %) (Fig. 1a). The intermediate or high NFS group revealed a significantly higher frequency of CG?GG genotypes than the low NFS group (p = 0.012) in the dominant model of inheritance, suggesting that CG?GG
123
2970
Dig Dis Sci (2014) 59:2967–2974
Table 1 Baseline characteristics of the healthy controls and the patients with nonalcoholic fatty liver disease (n = 339)
Age, mean ± SD (years) Gender, male BMI, mean ± SD (kg/m2) Hypertension Diabetes Cholesterol, mean ± SD (mg/dL) TG, mean ± SD (mg/dL) Uric acid, mean ± SD (mg/dL) Platelet, mean ± SD (9103/uL) ALP, mean ± SD (IU/L) AST, mean ± SD (IU/L) ALT, mean ± SD (IU/L) GGT, mean ± SD (IU/L)
Healthy controls (n = 184)
NAFLD patients (n = 155)
p value
45.3 ± 10.6 60 (32.6) 22.1 ± 2.6 0 (0) 0 (0) 190.1 ± 28.8 89.8 ± 54.7 4.8 ± 1.3 241.0 ± 52.0 58.9 ± 14.1 21.1 ± 5.4 18.1 ± 7.3 21.6 ± 18.3
44.1 ± 15.5 101 (65.2) 27.5 ± 3.7 46 (29.7) 42 (27.1) 206.0 ± 45.1 179.9 ± 86.3 6.2 ± 1.8 226.9 ± 69.6 85.1 ± 39.0 55.9 ± 33.1 93.0 ± 63.2 70.9 ± 45.9
0.361 \0.001 \0.001 \0.001 \0.001 \0.001 \0.001 \0.001 0.068 \0.001 \0.001 \0.001 \0.001
Data are presented as the mean ± SD for continuous data and percentages for categorical data The p values were obtained by comparing the quantitative phenotype between the two groups (Mann–Whitney U test). Gender, diabetes, and hypertension were analyzed using the chi-square test NAFLD nonalcoholic fatty liver disease, BMI body mass index, TG triglyceride, ALP alkaline phosphatase, AST aspartate transaminase, ALT alanine transaminase, GGT c-glutamyl transpeptidase
Table 2 Genotype and allele frequencies of PNPLA3 I148M polymorphism in healthy controls and nonalcoholic fatty liver disease patients Genotype
Healthy controls (n = 184)
NAFLD patients (n = 155)
p value
Genotypes CC/CG/GG (%) Dominant model CC/CG?GG (%) Alleles C/G (%)
55/92/37 (29.9/50.0/20.1) 55/129 (29.9/70.1) 202/166 (54.9/45.1)
31/75/49 (20.0/48.4/31.6) 31/124 (20.0/80.0) 137/173 (44.2/55.8)
0.006a 0.037b 0.006c
NAFLD nonalcoholic fatty liver disease a Comparisons for genotypes performed using the Cochran–Armitage trend test b Comparisons were performed using the chi-square test for categorical variables c Comparisons were performed using the additive model Table 3 Univariate and multivariate analyses of risk factors associated with nonalcoholic fatty liver disease (n = 339) Variables
Univariate OR
Multivariate p
Dominant modela OR (95 % CI)
PNPLA3 rs738409 genotype CG or GG genotype 1.705 CC/CG/GG genotype 1.534 Age, [46 years 1.136 Gender, male 3.865 18.545 BMI, [25 (kg/m2) TG [ 130 (mg/dL) 14.668
0.038 0.006 0.564 \0.001 \0.001 \0.001
2.568 (1.109–5.945) 1.427 2.159 11.065 11.365
(0.680–2.993) (1.027–4.537) (5.359–22.847) (5.443–23.732)
CI confidence interval, OR odds ratio, BMI body mass index, TG triglyceride a CG or GG genotype for rs738409 b Odds ratio per number of variant 148 M alleles
123
Additive modelb p
OR (95 % CI)
p
1.985 1.386 2.249 10.741 11.349
0.008 0.389 0.033 \0.001 \0.001
0.028 0.347 0.042 \0.001 \0.001
(1.200–3.283) (0.659–2.912) (1.068-4.734) (5.194–22.215) (5.417–23.777)
Dig Dis Sci (2014) 59:2967–2974
2971
Discussion
Fig. 1 Comparison of the genotype frequency of PNPLA3 rs738409 between the advanced fibrosis group and less advanced fibrosis group according to the nonalcoholic fatty liver disease fibrosis score (NFS) or BARD score in nonalcoholic fatty liver disease patients. a In 144 nonalcoholic fatty liver patients, the intermediate or high NFS group (NFS C -1.455) displayed a higher frequency of CG?GG genotypes than the low NFS group (NFS \ -1.455, p = 0.012). b In 155 nonalcoholic fatty liver patients, the high BARD score group (BARD score C 2) displayed a higher frequency of CG?GG genotypes than the low BARD score group (BARD score \ 2, p = 0.046). The p value is calculated using a chi-square test. NFS nonalcoholic fatty liver disease fibrosis score
genotype is associated with advanced fibrosis. Moreover, the high BARD score group (BARD score C2, n = 48) displayed a significantly higher frequency of CG?GG genotypes than the low BARD score group (BARD score \2, n = 107, p = 0.046) (Fig. 1b). To confirm this finding, univariate and multivariate logistic regression analyses were performed. The multivariable analysis indicated that the presence of the CG?GG genotype of rs738409 was an independent factor associated with advanced liver fibrosis in NAFLD, regardless of using the criteria of NFS or BARD score (Tables 4, 5). However, the significance was observed only in the dominant model and not in the additive model. On the other hand, no significant associations between APRI and genotype of rs738409 were observed (data not shown).
In this study, the frequency of the PNPLA3 rs738409 G allele was 0.45 in the controls and 0.56 in NAFLD patients in the Korean study population, which suggests that there was a significant association between this SNP and the development of NAFLD (p = 0.006). In comparison with the CC genotype, the CG?GG genotype under a dominant model raised the risk of NAFLD 2.6-fold after adjustment for age, gender, BMI, and TG level (95 % CI 1.109–5.945, p = 0.028), whereas, under an additive model, the OR for NAFLD was 1.9 (95 % CI 1.200–3.283, p = 0.008). Moreover, using the indirect serum markers of fibrosis, CG?GG genotype was an independent factor for advanced liver fibrosis in NAFLD patients. Therefore, the PNPLA3 SNP is significantly related to the development of NAFLD as well as to the development of liver fibrosis in Korean NAFLD patients. A recent meta-analysis revealed that the rs738409 I148M variant of PNPLA3 is a strong modifier of the natural history of NAFLD in different ethnic populations around the world [18]. However, only 2 studies from the Asian region, one from Taiwan [19] (pediatric population) and the other from Japan [20], were included in the metaanalysis. Moreover, there was no report on the genotype frequency or effect of the SNP in the development of NAFLD-associated fibrosis in Korean population. In the population without NAFLD, the reported G-allele frequency of rs738409 PNPLA3 is 0.17 in African Americans, 0.23 in Europeans, 0.40 in the Taiwanese, 0.44 in Japanese, and 0.49 in the Hispanics [8, 20, 21]. The G-allele frequency of 0.45 in the Korean healthy controls of this study is comparable to that of examined Japanese populations, demonstrating the high-frequency range among various ethnic groups. The prevalence of NAFLD is concordant with the frequencies of the PNPLA3 G allele in American, Argentinean, and German populations [22–24]. Moreover, this SNP was identified as a risk factor of NAFLD in normoglycemic Taiwanese and obese pediatric populations [19, 21], and in a Japanese population [20]. In the present study, the G-allele frequency in NAFLD patients was significantly higher than in healthy controls, and the PNPLA3 genotype based on both, the dominant and additive model, was an independent factor for NAFLD according to the multivariate analysis. Therefore, the carriers of the PNPLA3 rs738409 G allele were susceptible to NAFLD in the Korean population, which is concordant with the results of studies of various ethnic groups. A previous meta-analysis revealed the association of rs738409 with increased ALT levels in NAFLD, but there was no significant association with insulin resistance phenotype, such as HOMR-IR, fasting insulin level, or fasting
123
2972
Dig Dis Sci (2014) 59:2967–2974
Table 4 Univariate and multivariate analyses of predictors of advanced liver fibrosis (nonalcoholic fatty liver disease fibrosis score C -1.455) in patients with nonalcoholic fatty liver disease according to nonalcoholic fatty liver disease fibrosis score Variables
Univariate OR
Multivariate Dominant modela
p
OR (95 % CI)
Additive modelb p
OR (95 % CI)
p
PNPLA3 rs738409 genotype CG or GG genotype
5.700
0.014
CC/CG/GG genotype
1.656
0.070
Age, [46 years
9.000
\0.001
0.405
0.020
Gender, male 2
18.573 (2.035–169.526)
0.010
31.569 (5.974–166.831)
\0.001
1.765 (0.823–3.785) 1.596 (0.422–6.041)
0.491
26.526 (5.570–126.334)
0.144 \0.001
1.858 (0.496–6.966)
0.358
BMI, [25 (kg/m )
2.191
0.131
3.810 (0.991–14.650)
0.052
3.092 (0.857–11.148)
0.085
Hypertension
5.133
\0.001
1.900 (0.643–83.271)
0.246
1.779 (0.615–5.145)
0.288
Diabetes
9.889
\0.001
21.301 (5.449–83.271)
\0.001
15.788 (4.644–53.669)
\0.001
Available data in 144 patients NAFLD nonalcoholic fatty liver disease, CI confidence interval, OR odds ratio, BMI body mass index, TG triglyceride a CG or GG genotype for rs738409 b
Odds ratio per number of variant 148 M alleles
Table 5 Univariate and multivariate analyses of predictors of advanced liver fibrosis (BARD score C 2) in patients with nonalcoholic fatty liver disease according to BARD score (n = 155) Variables
Univariate OR
Multivariate Dominant modela
p
OR (95 % CI)
Additive modelb p
OR (95 % CI)
p
PNPLA3 rs738409 genotype CG or GG genotype
2.760
0.052
CC/CG/GG genotype
1.483
0.118
Age, [ 46 years
5.214
Gender, male
0.263 2
4.040 (1.084–15.048)
0.037 1.426 (0.774–2.628)
0.256
\0.001
2.536 (0.825–7.792)
0.104
2.726 (0.913–8.142)
0.072
\0.001
0.245 (0.075–0.799)
0.020
0.284 (0.091–0.890)
0.031
BMI, [ 25 (kg/m )
1.859
0.163
3.151 (1.001–9.921)
0.050
2.790 (0.904–8.610)
0.074
Hypertension
3.864
\0.001
1.715 (0.664–4.428)
0.265
1.633 (0.639–4.171)
0.305
Diabetes
7.886
\0.001
8.779 (3.364–22.915)
\0.001
8.502 (3.259–22.175)
\0.001
NAFLD nonalcoholic fatty liver disease, CI confidence interval, OR odds ratio, BMI body mass index, TG triglyceride a b
CG or GG genotype for rs738409 odds ratio per number of variant 148 M alleles
glucose level [18]. However, there were no differences in age, gender, BMI, diabetes, hypertension, and ALT level among the 3 genotypes in both, the controls and NAFLD patients, in our study; this was most likely because of the small size or different subject profiles. Recently, several noninvasive scoring systems have been proposed to predict advanced liver fibrosis without invasive liver biopsy, such as the NFS, BARD score, and APRI. The NFS is composed of 6 easily available variables and can accurately predict the presence or absence of advanced fibrosis in NAFLD patients. According to the previous results [14], the low cutoff score (-1.455) had an 88–93 % negative predictive value to exclude advanced
123
fibrosis, whereas the high cutoff (0.676) had an 82–90 % positive predictive value in identifying the presence of advanced fibrosis. A recent meta-analysis showed that NFS has been independently validated in different ethnicities including American Caucasians, blacks, and Hispanics, Asians, and Europeans [25]. In addition, Chang et al. [26] showed that NAFLD and its severity using NFS was associated with increased risk of diabetes in a healthy Korean population. The BARD score is another simple noninvasive scoring system for assessing the severity of fibrosis in NAFLD. A BARD score of 0 or 1 point had a 96 % negative predictive value (96 %) to identify patients without advanced
Dig Dis Sci (2014) 59:2967–2974
liver fibrosis, whereas the positive predictive value of a high BARD score (2–4 point) was not very good (43 %) [15]. In a study by Ruffillo et al. [27], both scoring systems had higher specificity than sensitivity. By using the NFS and BARD score in the present study, we demonstrated that there was a significant association between the CG?GG genotype of rs738409 and fibrosis of NAFLD, and the CG?GG genotype was an independent predictor of advanced liver fibrosis in NAFLD in the multivariate analysis. This finding suggests that the influence of the I148M variant on the susceptibility to advanced liver fibrosis may follow a dominant model. Consistent with our study, a previous meta-analysis revealed that the influence of PNPLA3 genotypes on NAFLD followed an additive genetic model, except for liver fibrosis, which followed a dominant genetic model [18]. APRI has been proposed as a noninvasive tool for the assessment of liver fibrosis in chronic liver disease [16]; however, we could not find a significant association between the PNPLA3 genotype and APRI score. In general, approximately 15–25 % of patients with NASH progress to cirrhosis [7]. The G allele of rs738409 in PNPLA3 is associated not only with histologic steatosis but also with fibrosis [28–30]. The association of this SNP with advanced liver fibrosis and steatosis was observed in previous Caucasian, Argentinean, Caucasian, and Asian populations [23, 28, 29, 31]. Moreover, several new studies in the Asian showed that the PNPLA3 rs738409 variant is strongly associated with the severity of NAFLD [32–34]. A recent meta-analysis revealed that PNPLA3 is associated with an increased risk of both advanced fibrosis and HCC among patients with liver disease [35]. On the other hand, the PNPLA3 rs738409 variant was strongly associated with alcoholic cirrhosis, which shares many similar features with NAFLD-related cirrhosis [36, 37], and with HCVrelated cirrhosis [38]. The I148M PNPLA3 variant influences the liver damage of steatohepatitis in chronic liver disease with metabolic, alcoholic, and viral risk factors [39]. This is the first study to elucidate a significant role of PNPLA3 rs738409 in the development of NAFLD and fibrosis in NAFLD in the Korean population. However, it has several limitations. First, this was a single-center, cross-sectional study with a relatively small sample size. Second, only 10 patients with NAFLD underwent liver biopsy, and we did not measure insulin resistance or quantitative intrahepatic fat content in the subject, so that the relationship between the SNP and those indices could not be investigated. Moreover, only 1 SNP of PNPLA3 (rs738409) was evaluated in this study, although 4 new SNPs were reported in recent papers [40]. Lastly, the study population were enrolled about 10 years age, so that some
2973
clinical features such BMI may not exactly reflect current status. Conclusion In conclusion, the G-allele frequency of the PNPLA3 SNP rs738409 in the healthy Korean population was 0.45, and the carriers of the PNPLA3 rs738409 G allele were susceptible to NAFLD. The CG?GG genotype may confer a significant risk of fibrosis in Korean patients with NAFLD. Therefore, NAFLD patients carrying a risk genotype may be a priority target group for intensive lifestyle modifications and therapeutic trials. Acknowledgments This study was supported by a research grant from an intramural fund of Seoul National University Bundang Hospital (No. 02-2012-046). Conflict of interest
Nothing to declare for all authors.
References 1. Angulo P. Nonalcoholic fatty liver disease. N Engl J Med. 2002;346:1221–1231. 2. Chae HB, Kim JH, Kim JK, Yim HJ. Current status of liver diseases in Korea: hepatitis B. Korean J Hepatol. 2009;15:S13– S24. 3. Park SH, Jeon WK, Kim SH, et al. Prevalence and risk factors of non-alcoholic fatty liver disease among Korean adults. J Gastroenterol Hepatol. 2006;21:138–143. 4. Park SH. Current status of liver disease in Korea: nonalcoholic fatty liver disease. Korean J Hepatol. 2009;15:S34–S39. 5. Chitturi S, Wong VW, Farrell G. Nonalcoholic fatty liver in Asia: firmly entrenched and rapidly gaining ground. J Gastroenterol Hepatol. 2011;26:163–172. 6. Vernon G, Baranova A, Younossi ZM. Systematic review: the epidemiology and natural history of non-alcoholic fatty liver disease and non-alcoholic steatohepatitis in adults. Aliment Pharmacol Ther. 2011;34:274–285. 7. Sanyal AJ. American gastroenterological A. AGA technical review on nonalcoholic fatty liver disease. Gastroenterology. 2002;123:1705–1725. 8. Romeo S, Kozlitina J, Xing C, et al. Genetic variation in PNPLA3 confers susceptibility to nonalcoholic fatty liver disease. Nat Genet. 2008;40:1461–1465. 9. Baulande S, Lasnier F, Lucas M, Pairault J. Adiponutrin, a transmembrane protein corresponding to a novel dietary- and obesity-linked mRNA specifically expressed in the adipose lineage. J Biol Chem. 2001;276:33336–33344. 10. Lake AC, Sun Y, Li JL, et al. Expression, regulation, and triglyceride hydrolase activity of Adiponutrin family members. J Lipid Res. 2005;46:2477–2487. 11. He S, McPhaul C, Li JZ, et al. A sequence variation (I148M) in PNPLA3 associated with nonalcoholic fatty liver disease disrupts triglyceride hydrolysis. J Biol Chem. 2010;285:6706–6715. 12. Pirazzi C, Adiels M, Burza MA, et al. Patatin-like phospholipase domain-containing 3 (PNPLA3) I148M (rs738409) affects hepatic VLDL secretion in humans and in vitro. J Hepatol. 2012;57:1276–1282.
123
2974 13. Kumari M, Schoiswohl G, Chitraju C, et al. Adiponutrin functions as a nutritionally regulated lysophosphatidic acid acyltransferase. Cell Metabol. 2012;15:691–702. 14. Angulo P, Hui JM, Marchesini G, et al. The NAFLD fibrosis score: a noninvasive system that identifies liver fibrosis in patients with NAFLD. Hepatol. 2007;45:846–854. 15. Harrison SA, Oliver D, Arnold HL, Gogia S, Neuschwander-Tetri BA. Development and validation of a simple NAFLD clinical scoring system for identifying patients without advanced disease. Gut. 2008;57:1441–1447. 16. Wai CT, Greenson JK, Fontana RJ, et al. A simple noninvasive index can predict both significant fibrosis and cirrhosis in patients with chronic hepatitis C. Hepatology. 2003;38:518–526. 17. Angulo P, Bugianesi E, Bjornsson ES, et al. Simple noninvasive systems predict long-term outcomes of patients with nonalcoholic fatty liver disease. Gastroenterology. 2013;145:e784. 18. Sookoian S, Pirola CJ. Meta-analysis of the influence of I148M variant of patatin-like phospholipase domain containing 3 gene (PNPLA3) on the susceptibility and histological severity of nonalcoholic fatty liver disease. Hepatology. 2011;53:1883–1894. 19. Lin YC, Chang PF, Hu FC, Yang WS, Chang MH, Ni YH. A common variant in the PNPLA3 gene is a risk factor for nonalcoholic fatty liver disease in obese Taiwanese children. J Pediatrics. 2011;158:740–744. 20. Hotta K, Yoneda M, Hyogo H, et al. Association of the rs738409 polymorphism in PNPLA3 with liver damage and the development of nonalcoholic fatty liver disease. BMC Med Genet. 2010;11:172. 21. Wang CW, Lin HY, Shin SJ, et al. The PNPLA3 I148M polymorphism is associated with insulin resistance and nonalcoholic fatty liver disease in a normoglycaemic population. Liver Int. 2011;31:1326–1331. 22. Kotronen A, Johansson LE, Johansson LM, et al. A common variant in PNPLA3, which encodes adiponutrin, is associated with liver fat content in humans. Diabetologia. 2009;52:1056–1060. 23. Sookoian S, Castano GO, Burgueno AL, Gianotti TF, Rosselli MS, Pirola CJ. A nonsynonymous gene variant in the adiponutrin gene is associated with nonalcoholic fatty liver disease severity. J Lipid Res. 2009;50:2111–2116. 24. Kantartzis K, Peter A, Machicao F, et al. Dissociation between fatty liver and insulin resistance in humans carrying a variant of the patatin-like phospholipase 3 gene. Diabetes. 2009;58:2616–2623. 25. Musso G, Gambino R, Cassader M, Pagano G. Meta-analysis: natural history of non-alcoholic fatty liver disease (NAFLD) and diagnostic accuracy of non-invasive tests for liver disease severity. Ann Med. 2011;43:617–649. 26. Chang Y, Jung HS, Yun KE, Cho J, Cho YK, Ryu S. Cohort study of non-alcoholic fatty liver disease, NAFLD fibrosis score, and the risk of incident diabetes in a Korean population. Am J Gastroenterol. 2013;108:1861–1868.
123
Dig Dis Sci (2014) 59:2967–2974 27. Ruffillo G, Fassio E, Alvarez E, et al. Comparison of NAFLD fibrosis score and BARD score in predicting fibrosis in nonalcoholic fatty liver disease. J Hepatol. 2011;54:160–163. 28. Valenti L, Al-Serri A, Daly AK, et al. Homozygosity for the patatin-like phospholipase-3/adiponutrin I148M polymorphism influences liver fibrosis in patients with nonalcoholic fatty liver disease. Hepatology. 2010;51:1209–1217. 29. Rotman Y, Koh C, Zmuda JM, Kleiner DE, Liang TJ, Nash CRN. The association of genetic variability in patatin-like phospholipase domain-containing protein 3 (PNPLA3) with histological severity of nonalcoholic fatty liver disease. Hepatology. 2010;52:894–903. 30. Speliotes EK, Butler JL, Palmer CD, et al. PNPLA3 variants specifically confer increased risk for histologic nonalcoholic fatty liver disease but not metabolic disease. Hepatology. 2010;52:904–912. 31. Li Y, Xing C, Cohen JC, Hobbs HH. Genetic variant in PNPLA3 is associated with nonalcoholic fatty liver disease in China. Hepatology. 2012;55:327–328. 32. Kitamoto T, Kitamoto A, Yoneda M, 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. 2013;132:783–792. 33. Zain SM, Mohamed R, Mahadeva S, et al. A multi-ethnic study of a PNPLA3 gene variant and its association with disease severity in non-alcoholic fatty liver disease. Hum Genet. 2012;131:1145–1152. 34. Kawaguchi T, Sumida Y, Umemura A, et al. Genetic polymorphisms of the human PNPLA3 gene are strongly associated with severity of non-alcoholic fatty liver disease in Japanese. PloS One. 2012;7:e38322. 35. Singal AG, Manjunath H, Yopp AC, et al. The effect of PNPLA3 on fibrosis progression and development of hepatocellular carcinoma: a meta-analysis. Am J Gastroenterol. 2014;109:325–334. 36. Tian C, Stokowski RP, Kershenobich D, Ballinger DG, Hinds DA. Variant in PNPLA3 is associated with alcoholic liver disease. Nat Genet. 2010;42:21–23. 37. Stickel F, Buch S, Lau K, et al. Genetic variation in the PNPLA3 gene is associated with alcoholic liver injury in caucasians. Hepatology. 2011;53:86–95. 38. Valenti L, Fargion S. Patatin-like phospholipase domain containing-3 Ile148 Met and fibrosis progression after liver transplantation. Hepatology. 2011;54:1484. 39. Valenti L, Alisi A, Nobili V. I148M PNPLA3 variant and progressive liver disease: A new paradigm in hepatology. Hepatology 2012. 40. Speliotes EK, Yerges-Armstrong LM, Wu J, et al. Genome-wide association analysis identifies variants associated with nonalcoholic fatty liver disease that have distinct effects on metabolic traits. PLoS Genet. 2011;7:e1001324.
