SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

The genetic factors associated with susceptibility to nonalcoholic fatty liver disease (NAFLD) in pediatric obesity remain largely unknown. Recently, a nonsynonymous single-nucleotide polymorphism (rs738409), in the patatin-like phospholipase 3 gene (PNPLA3) has been associated with hepatic steatosis in adults. In a multiethnic group of 85 obese youths, we genotyped the PNLPA3 single-nucleotide polymorphism, measured hepatic fat content by magnetic resonance imaging and insulin sensitivity by the insulin clamp. Because PNPLA3 might affect adipogenesis/lipogenesis, we explored the putative association with the distribution of adipose cell size and the expression of some adipogenic/lipogenic genes in a subset of subjects who underwent a subcutaneous fat biopsy. Steatosis was present in 41% of Caucasians, 23% of African Americans, and 66% of Hispanics. The frequency of PNPLA3(rs738409) G allele was 0.324 in Caucasians, 0.183 in African Americans, and 0.483 in Hispanics. The prevalence of the G allele was higher in subjects showing hepatic steatosis. Surprisingly, subjects carrying the G allele showed comparable hepatic glucose production rates, peripheral glucose disposal rate, and glycerol turnover as the CC homozygotes. Carriers of the G allele showed smaller adipocytes than those with CC genotype (P = 0.005). Although the expression of PNPLA3, PNPLA2, PPARγ2(peroxisome proliferator-activated receptor gamma 2), SREBP1c(sterol regulatory element binding protein 1c), and ACACA(acetyl coenzyme A carboxylase) was not different between genotypes, carriers of the G allele showed lower leptin (LEP)(P = 0.03) and sirtuin 1 (SIRT1) expression (P = 0.04). Conclusion: A common variant of the PNPLA3 gene confers susceptibility to hepatic steatosis in obese youths without increasing the level of hepatic and peripheral insulin resistance. The rs738409 PNPLA3 G allele is associated with morphological changes in adipocyte cell size. (HEPATOLOGY 2010.)

Nonalcoholic fatty liver disease (NAFLD) has emerged as the most common cause of chronic liver disease in pediatrics, affecting an alarming 38% of obese children.1-3 NAFLD varies from steatosis to steatohepatitis to advanced fibrosis with cirrhosis.4, 5 Children with NAFLD may develop end-stage liver disease with a consequent need for liver transplantation.4

Recently, a nonsynonymous single-nucleotide polymorphism (SNP)(rs738409), characterized by a C-to-G substitution encoding an isoleucine-to-methionine substitution at the amino acid position 148 (I148M), in the patatin-like phospholipase 3 gene (PNPLA3) was found to be associated with hepatic steatosis measured by proton magnetic resonance spectroscopy in a multiethnic cohort of adults.6 Subsequently, other studies in adults confirmed this association,7-11 with some studies indicating an association also with the severity of NAFLD10, 11 and a study suggesting that this polymorphism predisposes obese children and adolescents to exhibit early hepatic damage.12

The PNPLA3 gene product, known as adiponutrin, was originally identified as a member of the calcium-independent phospholipase A2 family.13 However, it has both triacylglycerol hydrolase and acylglycerol transacetylase activity.13 In animals and humans, adiponutrin is primarily expressed in white adipose tissue and liver,14 its expression is nutritionally regulated,15 and it increases with obesity.8 Moreover, it has been recently shown that the PNPLA3 gene product may also have a role in adipogenesis, being up-regulated during the differentiation of white adipocytes.16

Although adiponutrin expression is influenced by insulin, it is still unclear whether its expression is decreased in subjects with insulin resistance. In contrast to the existing information in adults, little is known about the potential role of variants in the PNLPA3 gene early in the development of fatty liver in pediatric obesity. Therefore, in this study, we determined: (1) the association between the PNPLA3 rs738409 SNP and fatty liver in a multiethnic group of obese youths; (2) whether the SNP influences hepatic and peripheral insulin sensitivity as measured by the hyperinsulinemic-euglycemic clamp17; and (3) in a subgroup of subjects undergoing a subcutaneous fat biopsy, we explored whether the polymorphism might be associated with changes in expression of PNPLA3 gene, which in turn, in light of adiponutrin lipogenic activity, might be associated with changes in adipocyte size as well as with changes in the expression of genes regulating adipogenesis and lipid metabolism.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

Subjects

We studied 85 obese (42 girls, with 34 Caucasian, 22 African American, and 29 Hispanic, age range = 8.1-18.7) children and adolescents recruited from the Yale Pediatric Obesity Clinic. The three ethnic groups did not differ for mean age (Caucasians = 13.4, 95% confidence interval [CI] = 12.6-14.3; African Americans =13.7, 95% CI = 12.6-14.7; Hispanics = 12.6, 95% CI = 11.7-13.5; P = 0.3) or for body mass index (BMI) z-score (Caucasians = 2.27, 95% CI = 2.12-2.42; African Americans = 2.48, 95% CI = 2.29-2.66; Hispanics = 2.26, 95% CI = 2.10-2.42). The prevalence of subjects showing impaired glucose tolerance (IGT) or type 2 diabetes did not differ among the groups. Thirteen Caucasians (eight females), five African Americans (all females), and 10 Hispanics (five females) showed IGT, whereas one Caucasian (female) and one African American (male) showed type 2 diabetes (P = 0.3).

To be eligible for this study, subjects could not be on medications known to affect liver function or alter glucose or lipid metabolism. Information relating to alcohol consumption was obtained in all subjects using a questionnaire. Autoimmune hepatitis, Wilson disease, alpha-1-antitrypsin deficiency, hepatitis B and C, and iron overload were excluded with appropriate tests in subjects with persistent elevation in alanine aminotransferase (>6 months).

The study was approved by the Yale University Human Investigation Committee. Parental informed consent and child assent were obtained from all participants.

Genotyping

Genomic DNA was extracted from peripheral blood leukocytes. To genotype the rs738409 SNP, the following pair of primers was used: forward = 5′-GCC CTG CTC ACT TGG AGA AA-3′ and reverse = 5′-TGA AAG GCA GTG AGG CAT GG-3′. Polymerase chain reaction (PCR) was carried out using the following conditions: denaturation at 95°C for 5 minutes followed by 35 cycles of 30 seconds at 94°C, 30 seconds at 55°C, and 30 seconds at 72°C. PCR products were analyzed by automated sequencing through the Yale W.M. Keck facility.

Metabolic Studies

All metabolic studies were done at the Yale Clinical Center Investigation at 8:00 AM following a 10-hour to 12-hour overnight fast.

Oral Glucose Tolerance Test

A standard oral glucose tolerance test (OGTT; 1.75 g/kg body weight, up to 75 g) was performed in all subjects. Whole Body Insulin Sensitivity Index (WBISI) was used as index of insulin sensitivity, recently validated for the use in obese children and adolescents.18, 19

Direct Measurement of Hepatic and Peripheral Insulin Sensitivity

The hyperinsulinemic-euglycemic clamp was performed in a subgroup of 41 subjects (16 male/25 female; 17 Caucasian/13 African American/11 Hispanic, mean age = 13.2, 95% CI = 11.9-14.5; mean BMI z-score = 2.39, 95% CI = 2.17-2.58). Twenty-six were normal glucose tolerant, 13 were IGT, and two showed type 2 diabetes. This subgroup did not differ from the main cohort for age, sex, race, BMI z-score, glucose tolerance, hepatic fat fraction (HFF), and body fat.

Two intravenous catheters (one for blood sampling and one for infusion of glucose, insulin, and stable isotopes) were inserted in the antecubital vein of each arm after local lidocaine infiltration.17 The sampling arm was kept in a heated box for arterialization of blood. Hepatic and peripheral insulin sensitivity was measured by a two-step hyperinsulinemic-euglycemic clamp by infusing insulin as a primed continuous infusion at 4 mU·m−2·minute−1·and 80 mU·m−2·minute−1.

The glucose infusion rates were calculated during the last 30 minutes of each step of the clamp and expressed as milligrams of glucose per minute per meter squared. Endogenous hepatic glucose production and glycerol turnover at baseline and during the two steps of the insulin clamp, along with the clamped glucose disposal rates, were calculated as previously reported.17

Imaging Studies

DEXA.

Total body composition was measured by dual-energy X-ray absorptiometry (DEXA) with a Hologic scanner.

Abdominal Magnetic Resonance Imaging.

Magnetic resonance imaging (MRI) studies were performed on a GE or Siemens Sonata 1.5 Tesla system.21

Fast-MRI.

Measurement of liver fat content was performed by MRI using the two-point Dixon (2PD) method as modified by Fishbein et al.22 Using the MRIcro software program, five regions of interest were drawn on each image and the mean pixel signal intensity level was recorded. The HFF was calculated in duplicate from the mean pixel signal intensity data using the formula: [(Sin− Sout)/(2 × Sin)] × 100.23

Liver Biopsy

Liver biopsy was performed in six subjects. All the information concerning the liver biopsy has been included as Supporting Information Material.

Adipose Tissue Biopsy, Cell Measurements, and Gene Expression

Of the 85 subjects, only a subgroup of 18 subjects (three male/three female Caucasians, three male/four female African Americans, and three male/two female Hispanics) consented to undergo a subcutaneous fat biopsy. This subgroup had a higher mean age (age = 15.1, 95% CI = 10-19) than the main group (P = 0.004), but similar BMI z-score, percent HFF, sex distribution, ethnicity, and glucose tolerance.

After administration of 0.25% lidocaine, a 1-cm scalpel incision was made inferior to the umbilicus, from which 2 g of subcutaneous adipose tissue was removed. Two samples of 20-30 mg of tissue were immediately fixed in osmium tetroxide and incubated in a water bath at 37°C for 48 hours.24

Cell size, determined by the Beckman Coulter (Miami, FL) Multisizer III, was described via a mathematical model.24 Individual analysis of adipose cell size distribution from Multisizer graphs entailed identification of the nadir, defined as the low point (in frequency) between the two cell populations, i.e. where the curve between the two populations was flat.25, 26 The number of adipocytes above and below this point was calculated by the Multisizer software, and expressed as the “percent above”(percent large cells) and “percent below”(percent small cells) the nadir. Finally, the Multisizer software calculated the mean, median, and mode of the overall cell size for each subject.

Quantitative Real-Time PCR.

In subcutaneous adipose tissue from those 18 subjects, we evaluated the expression of adiponutrin (PNPLA3), leptin (LEP), genes involved in adipogenesis/lipogenesis(peroxisome proliferator-activated receptor gamma 2 [PPARγ2], sterol regulatory element binding protein-1c [SREBP1c], acetyl coenzyme A carboxylase [ACACA]) and lipolysis(adipose triglyceride lipase [PNPLA2], and sirtuin 1 [SIRT1]). Gene expression data from these 18 subjects are included in two other submitted articles that are focused on two different topics: (1) association between cellularity of the adipose tissue and gene profiling and (2) the relationship between SIRT1 and inflammation. All the procedures concerning the gene expression analysis have been explained in detail in the Supporting Information Material.

Biochemical Analyses

Plasma glucose was determined using a glucose analyzer by the glucose oxidase method (Beckman Instruments, Brea, CA). Plasma insulin was measured by the Linco RIA, lipid levels were determined with an Auto-Analyzer (model 747-200), and liver enzymes were measured, using standard automated kinetic enzymatic assays. Analysis of enrichments of 6,6-[2H]-glucose and [2H]5-glycerol in plasma and infusates were done by gas chromatography/mass spectrometry.17

Statistical Analyses

To test the effect of the at risk allele on the development of hepatic steatosis, first we used the Cochran-Mantel-Haenszel test to assess if the odds ratio differed between the three different ethnic groups. The P value was 2.35 × 10−5, indicating that the three different groups needed to be analyzed separately. Then, within each group, four statistical tests were used to test the association between genotype and phenotype under different diseases models including: Cochran-Armitage trend test, allele association, dominant and recessive model. Except for the trend test, P value was calculated via Fisher's exact test. We tested four models due to the lack of knowledge on the underlying genetic model. For each population, the model with the minimal P value was considered the best model for describing the genotype-phenotype relationship. Because four models were used, we employed a permutation procedure to account for multiple comparisons. More specifically, we took the minimum P value among the observed test statistics and compared it to the minimum P value among the tests statistics from permuted data sets. For each permutation, we randomly shuffled the disease status among the cases and controls and redid the analysis based the permuted data sets and recorded the minimum P value. Then the empirical P value was calculated as the proportion of the permuted samples with equal or smaller minimum P value than the observed one. Ten thousand permutations were used to derive the empirical significance levels.

For the analysis evaluating the association between the genotypes and quantitative traits, the heterozygous and the minor allele homozygous were grouped and the analyses of variance or of covariance were run. When subjects carrying the rs738409 PNPLA3 minor allele (G) were compared to common allele homozygotes (CC) to assess differences in hepatic fat content (HFF), we used age, sex, BMI z-score, glucose tolerance status, and visceral fat as covariates. Non-normally distributed variables were log-transformed, and except for HFF, the square root transformation was used, ensuring a better normalization. Unless differently specified, for all the data, raw means and 95% confidence intervals are shown. Given the small sample size of the group of subjects who underwent the fat biopsy, to compare the traits between genotypes, the three ethnicities were combined and the Mann-Whitney test was used. In the same subgroup, the Spearman correlation was used to correlate the percent of HFF with percent of small cells. All P < 0.05 were considered statistically significant.

Results

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

Association of PNPLA3 SNP rs738409 with Hepatic Steatosis

We examined this association by dividing each ethnic group into case and control subgroups based on the presence or absence of hepatic steatosis (HFF < 5.5%). Fourteen Caucasians (41%), five African Americans (23%), and 19 Hispanics (66%) showed fatty liver.

The PNPLA3 rs738409 minor allele (G) frequency was 0.324 in Caucasians, 0.183 in African Americans, and 0.483 in the Hispanics. The genotype distribution was in Hardy Weinberg equilibrium in all ethnic groups (Table 1).

Table 1. Summary of the Allele and Genotype Frequency
Ethnic GroupsAffection StatusAllele FrequencyGenotype CountHardy-Weinberg EquilibriumInheritance Model
CGCCCGGGCochran-ArmitageGenotypicDominantRecessivePermutation
  1. The P values referred to in the Cochran-Armitage, Genotypic, Dominant, and Recessive tests are nominal P value. The P values referred to in the Permutation are empiric P values.

CaucasiansAll0.680.32171250.264.9 × 10–43.9 × 10–43.6 × 10–40.1460.0006
Affected0.570.432940.61  
Unaffected0.870.1315310.26  
African AmericansAll0.820.1815610.540.0040.0210.0210.2270.003
Affected0.500.501311  
Unaffected0.910.0914301  
HispanicsAll0.520.48814710.3680.5980.3901.0000.452
Affected0.530.4741051  
Unaffected0.600.404420.57  

We then used the Cochran-Mantel-Haenszel test to evaluate the evidence of heterogeneity of the allele frequency among the three ethnic groups. The P value was 2.350 × 10−5, indicating significant difference among the groups. Therefore, the association was evaluated separately among the ethnic groups. Table 1 shows the summary of the allele and genotype frequency. Three tests of association—including the Cochran-Armitage trend test, the genotype test, and the test based on dominant inheritance model—showed significant association among Caucasian and African American individuals only; the test based on the recessive inheritance model did not show any significant association among all three ethnic groups. Therefore, these tests indicate a dominant inheritance model where people with both CG and GG genotype tend to have increased risk of hepatic steatosis. The prevalence of G allele was significantly higher in subjects showing hepatic steatosis who were Caucasian (0.43 versus 0.13; P = 3.6×10−4) and African American (0.50 versus 0.09; P = 0.012), but not in those who were Hispanic (0.47 versus 0.40; P = 0.52). Consistently, subjects carrying the G allele clearly showed a higher hepatic fat content than common allele homozygotes (Fig. 1). This difference was statistically significant in Caucasians and African Americans, but not in Hispanics, although a similar trend was observed in this group. This association was independent of the BMI z-score, visceral fat, and glucose tolerance in Caucasians (P = 0.0001) and in African Americans (P = 0.01).

thumbnail image

Figure 1. Association between PNPLA3 rs738409 SNP and hepatic fat content (hepatic fat fraction [%HFF]) in Caucasians, African Americans, and Hispanics. The %HFF is expressed using box plots; dark blue indicates CC carriers and light blue the CG/GG carriers. P values are adjusted for age, sex, and BMI z-score. Data are shown as median and 95% CI. *P < 0.001; **P = 0.009.

Download figure to PowerPoint

Metabolic Phenotypes by Genotype

Carriers of the G allele, regardless of ethnicity, did not show any clear significant difference in overall degree of adiposity, abdominal fat distribution, and in the fasting lipid profile. Alanine and aspartate aminotransferases, which are markers of hepatic injury, tended to be higher in the G carriers in each ethnic group; however, the difference was not significant because of the small sample size.

The index of insulin sensitivity (WBISI) was also not different by genotype (Table 2). The lack of differences in insulin sensitivity was further confirmed by using the hyperinsulinemic clamp. Indeed, carriers of the G allele, despite having greater fat accumulation in the liver, did not manifest a greater peripheral insulin resistance (Fig. 2). There were no differences in hepatic insulin sensitivity and lipolysis between the CC and CG/GG carriers. Indeed, we found similar hepatic glucose production rates as well as glycerol turnover rates at baseline and during both steps of the clamps, in both the CC and CG/GG groups (Fig. 2).

thumbnail image

Figure 2. Hepatic fat content (%HFF)(top left) and insulin sensitivity (glucose disposal rate)(top right) by genotype. Basal, low dose, and high dose insulin dose clamp hepatic glucose production rates (bottom left) and glycerol turnover rates (bottom right) by genotype are expressed using box plots; dark blue indicates CC carriers and light blue the CG/GG carriers. Data are shown as median and 95% CI. *P = 0.04. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Download figure to PowerPoint

Table 2. Clinical Characteristics of the Subjects Stratified by Ethnicity and PNPLA3 Genotype
CharacteristicCaucasiansAfrican AmericansHispanics
CC (17)CG/GG (12/5)PCC (15)CG/GG (6/1 )PCC (8)CG/GG (14/7)P
  • *

    Log-transformed and adjusted for age, sex, and BMI z-score. Raw means and 95% CI are shown.

Age (years)13.7; 12.7-14.913.3; 12.1-14.30.513.7; 11.9-15.613.9; 11.7-16.00.913.0; 11.4-14.712.6; 11.5-13.70.7
Sex (male/female)9/86/110.37/82/50.54/415/60.5
NGT/IGT/T2D10/7/010/6/10.913/2/03/3/10.15/3/014/7/00.6
Anthropometrics
 BMI (kg/m 2)34.0; 30.7-37.234.2; 31.0-37.40.935.6; 32.1-39.241.3; 36.1-46.60.0832.4; 27.8-37.032.2; 29.1-35.20.9
 BMI z-score2.24; 2.04-2.432.30; 2.11-2.490.72.37; 2.15-2.582.64; 2.33-2.960.12.03; 1.54-2.522.22; 1.89-2.540.5
 Body fat (%)41.6; 37.0-46.245.0; 40.6-49.50.344.9; 41.3-48.548.5; 43.2-53.70.341.3; 33.6-49.045.0; 40.0-50.00.4
 Visceral *(cm2)71.8; 59.8-83.979.9; 67.8-98.90.244.8; 33.1-56.565.6; 48.4-82.70.456.0; 36.1-75.969.9; 56.5-83.20.5
 Subcutaneous *(cm2)546.9; 454.4-639.4547.1; 454.7-639.70.5539.4; 429.0-649.1670.1; 509.3-831.40.9436.5; 297.1-575.0517.4; 423.9-610.80.1
Lipid profile
 Cholesterol *(mg/dL)159.7; 140.2-179.3161.0; 141.4-180.50.7164.0; 151.7-177.0150.9; 133.2-170.10.6181.3; 152.8-209.8155.5; 135.3-175.70.2
 HDL*(mg/dL)48.8; 39.5-58.141.0; 31.7-50.30.642.6; 37.9-47.338.4; 31.8-45.00.542.2; 36.5-47.941.8; 37.8-45.80.8
 LDL*(mg/dL)96.1; 83.2-109.094.3; 81.8-106.70.9101.7; 89.2-114.296.3; 78.6-113.90.9110.7; 84.6-136.989.0; 70.8-108.90.5
 Triglycerides*(mg/dL)141.6; 66.5-215.4152.5; 80.4-224.50.996.8; 65.4-128.988.6; 42.0-134.40.9141.5; 93.6-189.4131.9; 95.9-167.80.6
Glucose and insulin levels
 Fasting glucose*(mg/dL)96.9; 91.8-101.998.1; 93.0-103.20.696.5; 89.5-103.4104.6; 94.5-114.70.397.0; 90.6-103.498.0; 93.8-102.50.8
 Glucose 120*(mg/dL)132.4; 111.8-153.0142.5; 127.5-157.40.6121.6; 101.8-141.3139.5; 110.5-168.50.7140.3; 119.2-161.5121.7; 105.8-137.70.3
 Fasting insulin*(μU/mL)31.4; 17.4-45.442.3; 27.4-57.30.231.4; 22.5-40.342.4; 28.9-55.90.537.2; 23.5-50.937.2; 28.0-46.31
 WBISI*1.64; 1.27-2.011.45; 1.05-1.850.21.56; 1.11-2.131.40; 0.71-2.130.41.42; 0.76-2.081.50; 1.06-1.950.7
Liver enzymes
 ALT*(UI/L)27.8; 7.7-47.947.4; 26.6-68.20.213.3; 8.74-17.719.6; 13.7-25.40.427.1; 9.3-63.557.9; 31.3-84.50.4
 AST*(UI/L)22.2; 12.1-32.234.3; 23.8-44.60.219.8; 17.4-22.020.8; 17.8-23.90.427.2; 8.8-45.639.2; 25.8-52.90.6

Fat Biopsy and Gene Expression Analysis

Of the 18 subjects undergoing the subcutaneous fat biopsy, 11 carried the minor allele (G). In particular one Caucasian, three African Americans, and three Hispanics showed the CC genotype; four Caucasians, four African Americans, and two Hispanics were heterozygotes; one Caucasian showed the GG genotype. As in the main group, the G carriers tended to show higher %HFF than the CC group (P = 0.05). Given the small sample size, the subjects were merged according to the genotype independently of the ethnicity in order to explore the association between PNPLA3 gene variant with the adipocyte size and gene expression. CG/GG carriers showed significantly higher percent of small cells (P = 0.005) as well as a trend for lower median adipocyte cell size than the CC group (P = 0.05; Fig. 3). Potential ethnic differences have been addressed by a within-ethnic-group permutation test; this is described in detail in the Supporting Information Material.

thumbnail image

Figure 3. Shown are (A) adipose cell size, as amount of small cells in percentage and (B) adipocyte median in micrometers, by genotype (dark blue indicates CC homozygous and light blue the G carriers). Box plots in (C) illustrate the difference in gene expression of PNPLA3, PNPLA2, SREBP1c, ACACA, PPARγ2, LEP, and SIRT1 by genotype (dark blue indicates CC homozygous and light blue the G carriers). Data are shown as median and 95% CI (*P = 0.005; **P = 0.05; ♦P = 0.03; ♦♦P = 0.04). [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Download figure to PowerPoint

The %HFF correlated positively with the percentage small cells (r = 0.50; P = 0.03). The expression of PNPLA3 and PNLPA2 in adipose tissue did not vary between genotypes (P = 0.7 and P = 0.1). LEP and SIRT1 gene expression was significantly lower in the CG/GG group (P = 0.037 and P = 0.046, respectively; Fig. 3). The expression of PPARγ2, SREBP1C, and ACACA was lower in subjects carrying the G allele; however, the differences did not reach significance.

Discussion

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

In this study, we observed that obese children and adolescents carrying the G allele have higher hepatic fat content (HFF) than C allele homozygotes. This association was significant in Caucasians and African Americans, but not in Hispanics, although this latter group showed the same trend. The lack of association in Hispanics may be due to the high prevalence of hepatic steatosis (65%) and the small sample size. The association between this SNP and hepatic steatosis in Caucasians and African Americans was independent of BMI, visceral fat, and glucose tolerance status. These findings support the hypothesis of a pivotal role of the PNPLA3 rs738409 SNP in the development of early onset NAFLD in obese youths.

Dissociation Between Genotype and Metabolic Phenotype

An interesting observation that surfaced was that G carriers, despite having hepatic steatosis were not more insulin resistant than the C homozygote. Although our results would suggest that this polymorphism may not influence insulin sensitivity, caution in the interpretation of the data is still needed because all the subjects were obese with variable degree of hepatic and peripheral insulin resistance. Although some transgenic mouse studies have disassociated hepatic steatosis from hepatic insulin resistance27 other studies28-32 in rodent models of NAFLD have demonstrated that diacylglycerol activation of PKCε is the key trigger in the pathogenesis of NAFLD associated hepatic insulin resistance. Taken together, it is possible that alterations in adiponutrin expression/activity lead to increased hepatic triglyceride content independent of changes in hepatocellular diacylglycerol content and PKCε activation. It is also conceivable that other factors associated with steatosis, such as inflammation, circulating adipokines, endoplasmic reticulum (ER) stress affect insulin sensitivity without necessarily being directly related with hepatic lipid accumulation.33

Adipocyte Morphology and Gene Expression in G Carriers

A further aim was to verify whether this polymorphism might influence the expression of PNPLA3 and thus be associated with changes in the size of adipocytes and the expression of adipogenic genes. We found that subjects carrying the rs738409 minor allele showed an increased number of small adipocytes. Moreover, genes known to be involved in adipogenesis and lipogenesis, like PPARγ2, SREBP1c, and ACACA, tended to be down-regulated without reaching significance. These data suggest that both adipogenesis and lipogenesis could be the pathways compromised in subjects carrying the rs738409 G allele. Although this observation has been noted in a small number of subjects and cannot be conclusive, these data suggest that PNPLA3 rs738409 (G) allele may contribute to the development of hepatic steatosis by modulating adipocyte size. Adipocyte size, in fact, reflects the amount of lipid storage in the subcutaneous fat depot. Limited expandability of the subcutaneous adipose tissue depot has been considered as a possible factor leading to fat accumulation in ectopic tissues and organs such as the liver.34 It is conceivable that adipose tissue capability to store triglycerides is much more reduced in subjects carrying the G allele compared to those who are C homozygous. Moreover, it has been demonstrated that during postabsorptive conditions, the major source of free fatty acids delivered to the liver is derived from free fatty acids released from subcutaneous adipose tissue, which enter the systemic circulation and are then transported to the liver by the hepatic artery and portal vein, after passage through splanchnic tissues.33 Thus, in the presence of smaller adipocytes in the subcutaneous adipose tissue, we may have an overflow of free fatty acids to the liver in which they accumulate as triglycerides. Furthermore, subjects carrying the minor allele showed a significant reduced expression of SIRT1, a gene involved in lipolysis that could be both the result of the higher prevalence of small cells as well as a mechanism to compensate for the storage defect.35 Our data are in line with a recent report that investigated SIRT1-overexpressing mice, which had decreased nuclear factor κB activity, protecting them from lipid-induced hepatic inflammation, glucose intolerance, and NAFLD.36 Not surprising was the significant reduced expression of LEP in subjects carrying the minor allele, given that their adipocyte size was decreased. It is well known that adipocyte size positively correlates with secretion and messenger RNA expression of leptin.37

Although the expression of PNPLA3 messenger RNA during the differentiation of white adipocytes and its response to classical regulatory hormones of lipid synthesis would suggest an important role of the protein in adipogenesis,17 it is known that the PNPLA3 gene product, adiponutrin, has a transacetylase activity, which catalyzes triglyceride synthesis in adipocytes,14 being up-regulated by insulin38 and refeeding.39 Our understanding of how this polymorphism might be linked to impaired subcutaneous adipocyte size is that this mutation may impair the lipogenic activity of the PNPLA3 gene product, adiponutrin, and/or impair the up-regulation of adiponutrin by insulin and food intake, which in obese subjects may lead the subcutaneous adipocytes to contain less triglycerides, being consequently smaller. On the other hand, it has to be taken into account that adipose cell size regulation is a complex trait depending on several molecules such as PNPLA2, a major lipolytic enzyme, which has been recently demonstrated to be a regulatory factor of lipid droplet size and, as a consequence, of adipose cell size.40

Other mechanisms by which variation in PNPLA3 affects liver triglyceride content have been hypothesized. Recent studies by Hobbs's group41 would suggest that PNPLA3 is a lipid droplet protein that can catalyze hydrolysis of triglyceride in vitro. Although the overexpression of wild-type PNPLA3 did not affect triglyceride concentration in the hepatocytes, the I148M mutation promoted triglyceride accumulation when fatty acid acylation was inhibited. These findings are consistent with the notion of I148M substitution interfering with hepatic triglyceride hydrolysis as a way of promoting hepatic steatosis.40

In summary, our results suggest that the G allele of the PNPLA3 rs738409 SNP increases susceptibility to liver steatosis in obese youths. However, we did not find an association with both hepatic and peripheral insulin resistance as well as with insulin's ability to suppress lipolysis. Moreover, subjects carrying the rs738409 PNPLA3 G allele showed smaller adipocytes; this latter observation warrants further studies to unravel the mechanisms explaining the relationship among adipose cell size and adipogenesis, hepatic steatosis, and PNPLA3 genotype.

References

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
HEP_23832_sm_SuppFig-1.tif12165KSupporting Figure 1
HEP_23832_sm_SuppMat.doc24KSupporting Material

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.