I148M patatin-like phospholipase domain-containing 3 gene variant and severity of pediatric nonalcoholic fatty liver disease

Abstract

Nonalcoholic fatty liver disease (NAFLD) is one of the most common causes of chronic liver disease in children. Genetic variability, which is a main player in NAFLD, is especially characterized by polymorphisms in genes involved in the development and progression of the disease to nonalcoholic steatohepatitis (NASH). Recently, the rs738409 C>G adiponutrin/patatin-like phospholipase domain-containing 3 (PNPLA3) polymorphism, which encodes the I148M protein variant in the catalytic domain, has been associated with severe steatosis, NASH, and liver fibrosis in adults. In this study, we investigated the association between the rs738409 PNPLA3 gene polymorphism and NAFLD in 149 consecutive children and adolescents (age = 6-13 years) with biopsy-proven NAFLD. We analyzed the rs738409 polymorphism by a 5′-nuclease TaqMan assay and assessed its association with NASH: 41% of the subjects with NAFLD showed heterozygosity and 15% showed homozygosity for the at-risk G allele. The rs738409 genotype did not influence the body mass, adiposity, lipid levels, or insulin resistance and was not associated with alanine aminotransferase levels. Interestingly, the rs738409 G allele was strongly associated with the severity of steatosis (P < 0.0001), the presence of NASH (P < 0.0001), hepatocellular ballooning (P < 0.0001), lobular inflammation (P < 0.0001), and the presence of fibrosis (P = 0.01) independently of confounders. Individuals carrying two minor G alleles almost always had severe steatosis and NASH, heterozygotes were at intermediate risk, and patients negative for G alleles had milder and often uncomplicated steatosis. Conclusion: The PNPLA3 rs738409 polymorphism is associated with steatosis severity, hepatocellular ballooning, lobular inflammation, and perivenular fibrosis in pediatric NAFLD. (HEPATOLOGY 2010)

Pediatric nonalcoholic fatty liver disease (NAFLD) has become the most frequent chronic liver disease in children and adolescents in industrialized countries in tandem with the growing prevalence of childhood obesity and overweight.1-3 NAFLD affects 2.6% to 9.8% of children and adolescents, and this figure increases up to approximately 80% among obese individuals.3-6 A large survey found elevated alanine aminotransferase (ALT) levels in 8% of US adolescents (age = 12-19 years).7 In the two largest samples of biopsy-proven NAFLD described in the literature, 84% (Rome) and 68% (San Diego) of NAFLD children were diagnosed with nonalcoholic steatohepatitis (NASH).8, 9

NASH, which is considered the progressive form of NAFLD and is characterized by necroinflammatory changes, ballooning degeneration, and/or fibrosis, can progress to liver failure and hepatocarcinoma.10

Generally, the condition predisposing children to pediatric NAFLD is hyperalimentation associated with inadequate physical activity, which leads to a progressive increase in the body mass index (BMI) and visceral adiposity. Calorie intake greater than that needed for growth may cause overweight and obesity in children. This is becoming more and more widespread with the daily consumption of fast foods and soft drinks, which are associated with inactive leisure activities such as watching television and playing video games. However, familial, epidemiological, and twin studies have suggested that inherited factors may also play a pivotal role in determining the susceptibility to developing NASH.11-14 Single nucleotide polymorphisms (SNPs) in genes involved in inflammation, insulin signaling, oxidative stress, and fibrogenesis have been associated with the severity of liver damage in NAFLD,15-18 but these factors explain only a small portion of fibrosis variability.

Recently, genome-wide association studies have identified as a strong determinant of liver fat an SNP in adiponutrin/patatin-like phospholipase domain-containing 3 (PNPLA3), rs738409 C>G, which encodes the I148M protein variant.19-21 The I148M SNP influences liver fat without affecting the body mass, dyslipidemia, or insulin resistance. Adiponutrin expression in the liver and adipose tissue is increased by carbohydrate feeding and a Western-type diet.22-24 Furthermore, it has lipase activity against triglycerides and thus is likely involved in energy mobilization and storage in lipid droplets.25 It has been reported that the 148M PNPLA3 allele is a loss-of-function variant that predisposes patients to steatosis by decreasing triglyceride hydrolysis in hepatocytes.26

We recently showed that the rs738409 PNPLA3 SNP was strongly associated with severe steatosis, NASH, and the progression of liver fibrosis in a large series of Italian and UK patients with NAFLD.27 However, even though genetic factors likely play a stronger role in NASH development in children, no data are available concerning the role of the PNPLA3 genotype in this setting.

The aim of this study was to evaluate whether the rs738409 PNPLA3 SNP, encoding the functional I148M protein variant, is associated with a predisposition to NASH and progressive liver fibrosis in a large series of Italian pediatric patients with a histological diagnosis of NAFLD and may represent a noninvasive early marker of advanced disease.

Abbreviations

ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; CI, confidence interval; GGT, gamma-glutamyl transferase; HOMA-IR, homeostasis model assessment of insulin resistance; IGT, impaired glucose tolerance; NAFLD, nonalcoholic fatty liver disease; NASH, nonalcoholic steatohepatitis; NS, not significant; OR, odds ratio; PNPLA3, patatin-like phospholipase domain-containing 3; SD, standard deviation; SNP, single nucleotide polymorphism.

Patients and Methods

Patients

This prospective study included 149 consecutive untreated children and adolescents (93 males and 56 females) with biopsy-proven NAFLD who were referred to Bambino Gesù Children's Hospital between May 2006 and November 2009. All patients were tested for secondary causes of steatosis such as alcohol abuse (≥140 g/week), total parenteral nutrition, and the use of drugs known to precipitate steatosis (e.g., valproate, amiodarone, and prednisone). Hepatitis A, B, C, D, E, and G, cytomegalovirus, and Epstein-Barr virus infections were ruled out by appropriate tests. In all cases, autoimmune liver disease, metabolic liver disease, Wilson's disease, and alpha-1-antitrypsin were ruled out with standard clinical and laboratory evaluations as well as liver biopsy. All included subjects were Caucasians of Italian descent. The study protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki, and the study was performed according to the recommendations of the ethics committee of our hospital. Informed consent was obtained from each patient or responsible guardian.

Anthropometrical and Biochemical Measures

The height in meters, weight in kilograms, and BMI were calculated and converted into standard deviation (SD) scores. We examined aspartate aminotransferase (AST), ALT, and gamma-glutamyl transferase (GGT) levels as previously described.28

Liver Histology

Biopsy was performed in all children with an automatic core biopsy device (Biopince, Amedic, Sweden) with an 18-G, 150-mm-long needle that had the ability to cut tissue up to 33 mm long with extreme precision.29

Liver biopsy samples were at least 18 mm long and were read by a single liver pathologist who was unaware of the clinical and laboratory data of the patients. Biopsy samples were routinely processed (formalin-fixed and paraffin-embedded) and stained with hematoxylin and eosin and Van Gieson stains for the assessment of fibrosis and architectural changes.

The diagnosis of NASH was based on the pathologist's overall impression according to Kleiner et al.30 The main histological features commonly described for NAFLD, including steatosis, inflammation (portal and lobular), hepatocyte ballooning, and fibrosis, were scored according to the scoring system for NAFLD recently developed by the National Institutes of Health–sponsored NASH Clinical Research Network.30 Briefly, steatosis was graded on a four-point scale: (0) steatosis involving fewer than 5% of hepatocytes, (1) steatosis involving up to 33% of hepatocytes, (2) steatosis involving 33% to 66% of hepatocytes, and (3) steatosis involving more than 66% of hepatocytes. Lobular inflammation was graded on a four-point scale: (0) no foci, (1) fewer than two foci per 200× field, (2) two to four foci per 200× field, and (3) more than four foci per 200× field. Hepatocyte ballooning was graded from 0 to 2: (0) no balloon cells, (1) few balloon cells, and (2) many/prominent balloon cells. The stage of fibrosis was quantified with a five-point scale: (0) no fibrosis, (1) perisinusoidal or periportal fibrosis [(1a) mild, zone 3, perisinusoidal; (1b) moderate, zone 3, perisinusoidal; and (1c) portal/periportal], (2) perisinusoidal and portal/periportal fibrosis, (3) bridging fibrosis, and (4) cirrhosis.

Clinical and histological features of the patients included in the study are shown in Table 1.

Table 1. Demographic, Clinical, Histological, and Genetic Features of 149 Italian Pediatric Patients with NAFLD

Feature	Mean ± SD or n (%)
Abbreviation: HOMA-IR, homeostasis model assessment of insulin resistance.
Female sex	56 (37.6)
Age at first visit (years)	10.2 ± 2.6
BMI (centile)	94.9 ± 5.9
Obesity	135 (90.6)
Waist circumference (cm)	84.2 ± 10
Total cholesterol (mg/dL)	161.9 ± 35
Triglycerides (mg/dL)	111.1 ± 67
Glucose (mg/dL)	83.2 ± 13
HOMA-IR	2.6 ± 1.9
IGT or diabetes	70 (47.0)
Hypertension	54 (36.2)
ALT (IU/mL)	86.7 ± 59
GGT (IU/mL)	31.5 ± 19
NASH	70 (47)
Fibrosis stage (F0/F1/F2/F3/F4)	53/81/7/8/0 (36/54/5/5/0)
Adiponutrin/PNPLA3 rs738409 C>G genotype (CC/CG/GG)	65/61/23 (44/41/15)

Genetic Analysis.

DNA was extracted from peripheral blood by the phenol-chloroform method. The rate of success in extracting DNA was 100% for each study group. The PNPLA3 rs738409 C>G SNP, encoding I148M, was genotyped with a 5′-nuclease TaqMan assay (Assay on Demand for rs738409, Applied Biosystems, Foster City, CA) by personnel unaware of the clinical status of the patients and controls. Post–polymerase chain reaction allelic discrimination was carried out through the measurement of allele-specific fluorescence on the Opticon 2 detection system (MJ Research, Waltham, MA). Random samples were confirmed by direct genotyping, which provided concordant results in all cases; controls were included in all analyzed batches, and quality controls were used to verify the reproducibility of the results. Valid genotypic data were obtained for more than 99% of the analyzed subjects.27

Statistical Analysis.

Results are expressed as means and SDs. Mean values were compared by analysis of variance or Wilcoxon testing as appropriate, and frequencies were compared by Fisher's exact test for trends. The association between the I148M PNPLA3 SNP, steatosis severity, NASH, and fibrosis was evaluated by multivariate logistic regression analysis. Analyses were carried out with JMP 6.0 statistical analysis software (SAS Institute, Inc., Cary, NC).

Results

We previously showed that overtransmission of the rs738409 G allele affected patients in a subset of 71 family trios of patients included in this study, and this indicated that the rs738409 G allele is a genetic factor predisposing people to NAFLD development.31

In the present study, the frequency distribution of the rs738409 SNP was in Hardy-Weinberg equilibrium in the 149 patients with NAFLD. Heterozygosity for the at-risk G allele was observed in 41% of patients, and homozygosity was observed in 15% (Table 1).

As reported in adults, the rs738409 genotype and the presence of the rs738409 G allele were not significantly associated with the body mass, adiposity, lipid levels, or insulin resistance (Table 2). Furthermore, the rs738409 genotype and the G allele were not associated with basal insulin levels or insulin and glucose levels 30, 60, 90, and 120 minutes after oral glucose tolerance testing or with the quantitative insulin sensitivity check index, insulin sensitivity index, AST levels, and liver function tests [including the prothrombin time and albumin, pseudocholinesterase, and platelet levels (not shown in detail)].

Table 2. Association of the rs738409 C>G Adiponutrin/PNPLA3 SNP Encoding for the I148M Protein Variant with Metabolic Features and Liver Damage in 149 Italian Children with NAFLD

	I148M PNPLA3 Genotype
Data are presented as means and SDs or as numbers and percentages. Abbreviation: HOMA-IR, homeostasis model assessment of insulin resistance.
	CC (148I/I)	CG (148I/M)	GG (148M/M)	P
Number	65	61	23
Age at first visit (years)	10.4 ± 2.5	9.9 ± 2.6	10.2 ± 2.6	0.51
Female sex	23 (35)	22 (36)	11 (48)	0.55
BMI (centile)	95.2 ± 5.5	95.0 ± 5.8	94.1 ± 6.9	0.75
Waist circumference (cm)	84.9 ± 9.8	82.7 ± 11.1	85.9 ± 9.3	0.34
Total cholesterol (mg/dL)	165.5 ± 36	159.6 ± 32	158.1 ± 36	0.54
Triglycerides (mg/dL)	112.3 ± 61	105.4 ± 64	122.9 ± 92	0.56
Glucose (mg/dL)	81.9 ± 12	83.9 ± 14	84.8 ± 15	0.58
HOMA-IR index	2.5 ± 1.5	2.7 ± 2.4	2.4 ± 1.3	0.76
IGT or diabetes	31 (48)	29 (48)	10 (43)	0.93
Hypertension	43 (41.7)	49 (43.0)	17 (47.2)	0.81
ALT (IU/mL)	86.0 ± 58	87.3 ± 58	86.8 ± 70	0.99
GGT (IU/mL)	32.4 ± 19	30.6 ± 19	31.5 ± 21	0.86

In contrast to what was observed in adults, the rs738409 genotype was not associated with ALT levels in this series of pediatric patients.

The relationship between the rs738409 genotype and the severity of liver steatosis (grades 1-3) is shown in Fig. 1. The rs738409 G allele was strongly associated with the severity of steatosis (P < 0.0001) in a dose-dependent manner. In particular, the prevalence of grade 2 steatosis was higher in patients with the GG genotype versus those with the CG genotype, and the prevalence of grade 3 steatosis was higher in patients with the GG genotype versus those with the CG and CC genotypes (P < 0.05).

Figure 1. Relationship between the rs738409 genotype and the severity of liver steatosis (grades 1-3) in 149 pediatric patients with NAFLD.

Independent predictors of grade 2 to 3 steatosis are shown in Table 3. Moderate/severe steatosis was associated with the rs738409 genotype independently of the age at presentation, body mass, and presence of metabolic syndrome [odds ratio (OR) = 18.86, 95% confidence interval (CI) = 7.1-47].

Table 3. Independent Predictors of Grade 2 to 3 Steatosis in 149 Italian Pediatric Patients with NAFLD

	OR	95% CI	P
Abbreviation: NS, not significant. * *Number of G alleles: (0) CC genotype, (1) CG genotype, and (2) GG genotype.
Age at first visit (years)	0.71	0.57-0.87	0.002
Metabolic syndrome	0.60	0.35-0.99	0.05
BMI (centile)	1.03	0.93-1.13	NS
rs738409 PNPLA3 G alleles (n)*	18.86	7.1-47	<0.0001

The prevalence of NASH was 3% in children with the CC genotype (2/65), 74% in those with the CG genotype (45/61), and 100% in those with the GG genotype (23/23; P < 0.0001; Fig. 2). Because of the almost complete association of the rs738409 GG genotype (i.e., two at-risk alleles) with NASH and the occurrence of all cases of simple steatosis (i.e., the absence of NASH) in patients with the rs738409 CC genotype (no at-risk alleles), it was not even possible to estimate reliable ORs of NASH for the rs738409 genotype.

Figure 2. Relationship between the rs738409 genotype and the presence of NASH in 149 pediatric patients with NAFLD (P < 0.0001).

The PNPLA3 genotype was associated with the severity of both lobular necroinflammation [a grade > 1 was observed in 2 of 65 children with the CC genotype (3%), in 18 of 61 with the CG genotype (30%), and in 16 of 23 with the GG genotype (70%); P < 0.0001] and hepatocellular ballooning [observed in 12 of 65 children with the CC genotype (18%), in 34 of 61 with the CG genotype (56%), and in 20 of 23 with the GG genotype (87%); P < 0.0001].

There was a significant association between the PNPLA3 genotype and the presence of fibrosis (P = 0.03; Fig. 3). In particular, the rs738409 genotype was strongly associated with the presence of perivenular or higher grade fibrosis [in 20 of 65 patients with the CC genotype (31%), in 29 of 61 with the CG genotype (48%), and in 17 of 23 with the GG genotype (74%); P = 0.0005]. In contrast, the rs738409 G allele did not predispose children to periportal fibrosis (grade 1c). The prevalence of periportal fibrosis was 26% in patients with the CC genotype (17/65), 18% in patients with the CG genotype (11/61), and 9% in patients with the GG genotype (2/23).

Figure 3. Relationship between the rs738409 genotype and the presence of fibrosis in 149 pediatric patients with NAFLD (P = 0.03).

Independent predictors of the presence of fibrosis are shown in Table 4. The presence of fibrosis was associated with the rs738409 genotype independently of the age at presentation, waist circumference, impaired glucose tolerance (IGT) or diabetes status, and ALT levels (OR = 1.94, 95% CI = 1.14-3.45 per number of G alleles).

Table 4. Independent Predictors of the Presence of Fibrosis in 149 Italian Children with NAFLD

	OR	95% CI	P
* Number of G alleles: (0) CC genotype, (1) CG genotype, and (2) GG genotype.
Age at first visit (years)	0.70	0.52-0.92	0.0096
Waist circumference (cm)	1.11	1.04-1.20	0.0017
IGT or diabetes	1.43	0.98-2.12	0.066
ALT (IU/mL)	1.01	1.01-1.03	0.0078
rs738409 PNPLA3 G alleles (n)*	1.94	1.14-3.95	0.0142

Discussion

Paralleling the epidemic of childhood obesity, pediatric NAFLD has become the most frequent chronic, potentially progressive liver disease10 in children and adolescents in industrialized countries.1-3 Because NASH has a strong genetic component,11-14 hypothesizing that inherited factors are particularly important in early-onset cases, we evaluated whether the rs738409 SNP of PNPLA3, recently identified as a determinant of liver fat content and NASH susceptibility in adults,19, 27, 32, 33 influences the severity of liver diseases in pediatric patients with NAFLD and may represent a noninvasive early marker able to identify patients at high risk of advanced disease.

PNPLA3 is a phospholipase induced by steatosis in hepatocytes and is likely involved in triglyceride metabolism,23, 26 and the rs738489 C>G SNP encoding the I148M protein variants that affect the catalytic subunit results in decreased enzymatic activity and hepatocellular fat accumulation.26

Our results, obtained from a large series of obese pediatric patients with histologically proven NAFLD, indicate that the rs738409 G allele represents the strongest determinant of steatosis severity, with severe steatosis occurring almost exclusively and importantly almost always in the 15% of patients carrying two at-risk G alleles (the GG genotype).

The strength of this association, which by far surpasses the link between the PNPLA3 genotype and steatosis observed in adult NAFLD patients,27 suggests that the rs738409 genotype may represent a critical factor that determines whether the increased hepatic free fatty acid flux related to obesity translates into mild, uncomplicated steatosis or severe, progressive steatohepatitis in obese children. It can be hypothesized that we observed a stronger link between PNPLA3 and steatosis in children and adolescents versus adults because of the lower number of confounding factors in pediatric patients (e.g., the duration of disease, presence of obesity, lifestyle habits, comorbidities, and drugs) and the likely more important role played by genetic factors in early-onset disease.

In NAFLD, the steatosis grade parallels the severity of necroinflammatory changes, and this suggests that liver damage is strictly entangled with lipid metabolism alterations.34 Indeed, in patients with the rs738409 GG genotype, severe steatosis was associated with increased lobular inflammation and hepatocellular ballooning, and NASH was present in all cases. In contrast, simple, uncomplicated steatosis was largely the predominant histological picture observed in patients who did not carry any G allele (the CC genotype). Patients carrying only one G allele (the CG genotype) were at intermediate risk.

The rs738409 G allele not only predisposes patients to severe steatosis, lobular necroinflammation, ballooning, and NASH but also is associated with the presence of fibrosis and particularly perisinusoidal fibrosis, the typical manifestation of chronic liver damage in adult and pediatric patients with type 1 NASH.8, 30, 35 This suggests that these subjects are at increased risk of advanced liver disease later in life.

Whether the association between the PNPLA3 genotype and increased hepatocellular damage is mediated by increased steatosis or the PNPLA3 genotype also directly influences proinflammatory pathways in the liver remains to be determined.27, 36

Moreover, the PNPLA3 genotype did not predispose patients to periportal fibrosis (stage 1c according to the NASH Clinical Research Network scoring system),30 which has been reported to represent a marker of disease severity and progression in adult and pediatric patients with NASH37, 38 and particularly in Hispanic and Asian children.8 Our results support the hypothesis that the PNPLA3 148M allele favors severe steatosis, lobular inflammation with ballooning, and perisinusoidal/centrilobular fibrosis, whereas periportal fibrosis may be caused by altered regeneration and a ductular reaction possibly related to insulin resistance.37, 38 However, because we did not evaluate children of other ethnic groups with NAFLD, we cannot exclude the idea that the PNPLA3 I148M SNP has a different histological expression according to the genetic background.

Importantly, in line with previous findings in adults, the association between the PNPLA3 148M allele and NASH-related fibrosis was independent of confounding factors such as the age at presentation, adipose tissue mass and distribution, presence of diabetes or IGT,19, 27, 32, 39 and ALT levels, and this indicates that the evaluation of the rs738409 genotype may provide additional useful information for clinical identification of patients at risk of progressive disease. However, because of the limited number of subjects considered, the lack of an association between the PNPLA3 SNP and liver enzymes, despite increased liver damage in children, should be further confirmed in larger series.

The effect of the rs738409 genotype on fibrosis progression should be evaluated in prospective studies enrolling young patients with NAFLD.40 Nevertheless, the unexpectedly striking association between the PNPLA3 148M variant and liver damage reinforces the suggestion that liver biopsy29 and aggressive treatment should be indicated for pediatric patients carrying the rs738409 GG genotype.

However, although it was previously observed in a multiethnic US population that the same rs738409 SNP explained a good part of the increased susceptibility to steatosis of Hispanics versus subjects of European descent,19 we should note that the present results apply only to Caucasian children with NAFLD.

In conclusion, the PNPLA3 rs738409 SNP is associated with steatosis severity and the presence of NASH and fibrosis in pediatric Italian patients with NAFLD. Individuals carrying the GG allele have a very high risk of progressive liver disease. Screening analysis of the PNPLA3 rs738409 SNP could be a useful tool for discriminating among the at-risk pediatric subjects who require continuous monitoring over time because they are extremely susceptible to NASH and fibrosis.