ABSTRACT
AIMS/HYPOTHESIS: The rs738409 C>G single-nucleotide polymorphism in PNPLA3 leads to a missense mutation (I148M) which increases liver fat but does not cause insulin resistance. We hypothesised that patients with non-alcoholic fatty liver disease (NAFLD) due to the PNPLA3 variant ('PNPLA3 NAFLD' = PNPLA3-148MM) do not have adipose tissue (AT) inflammation in contrast with those with NAFLD due to obesity ('obese NAFLD'). METHODS: Biopsy specimens of AT were taken, and PNPLA3 genotype and liver fat ((1)H-magnetic resonance spectroscopy) were determined in 82 volunteers, who were divided into groups based on either median BMI (obese 36.2 ± 0.7 kg/m(2); non-obese 26.0 ± 0.4 kg/m(2)) or PNPLA3 genotype. All groups were similar with respect to age and sex. The PNPLA3 subgroups were equally obese (PNPLA3-148MM, 31.1 ± 1.3 kg/m(2); PNPLA3-148II, 31.2 ± 0.8 kg/m(2)), while the obese and non-obese subgroups had similar PNPLA3 genotype distribution. Gene expression of proinflammatory (MCP-1, CD68) and anti-inflammatory (Twist1, ADIPOQ) markers was measured using quantitative real-time RT-PCR. RESULTS: Liver fat was similarly increased in obese NAFLD (9.5 ± 1.3% vs 5.1 ± 0.9%, obese vs non-obese, p = 0.007) and PNPLA3 NAFLD (11.4 ± 1.7% vs 5.3 ± 0.8%, PNPLA3-148MM vs PNPLA3-148II, p < 0.001). Fasting serum insulin was higher in the obese than the non-obese group (76 ± 6 vs 47 ± 6 pmol/l, p < 0.001), but similar in PNPLA3-148MM and PNPLA3-148II (60 ± 8 vs 62 ± 5 pmol/l, NS). In obese vs non-obese, MCP-1 and CD68 mRNAs were upregulated, whereas those of Twist1 and ADIPOQ were significantly downregulated. AT gene expression of MCP-1, CD68, Twist1 and ADIPOQ was similar in PNPLA3-148MM and PNPLA3-148II groups. CONCLUSIONS/INTERPRETATION: PNPLA3 NAFLD is characterised by an increase in liver fat but no insulin resistance or AT inflammation, while obese NAFLD has all three of these features.
Subject(s)
Adipose Tissue/pathology , Fatty Liver/complications , Fatty Liver/pathology , Lipase/genetics , Membrane Proteins/genetics , Obesity/complications , Adiponectin/metabolism , Adult , Body Composition , Fatty Liver/genetics , Female , Gene Expression Regulation , Genetic Variation , Genotype , Humans , Liver/pathology , Male , Middle Aged , Mutation, Missense , Non-alcoholic Fatty Liver Disease , Obesity/geneticsABSTRACT
OBJECTIVE: Studies in mice have suggested that endocannabinoid blockade using the cannabinoid receptor type 1 (CB1) blocker rimonabant prevents obesity-induced hepatic steatosis. DESIGN AND SUBJECTS: To determine effects of rimonabant on liver fat in humans, we measured liver fat content by proton magnetic resonance spectroscopy in 37 subjects who used either a CB1 blocker rimonabant or placebo in a double-blind, randomized manner. This was retrospectively compared with a historical hypocaloric diet weight loss group (n=23). RESULTS: Weight loss averaged 8.5±1.4 kg in the rimonabant, 1.7±1.0 kg in the placebo and 7.5±0.2 kg in the hypocaloric diet group (P<0.001, rimonabant vs placebo; NS, rimonabant vs hypocaloric diet). Liver fat decreased more in the rimonabant (5.9% (2.5-14.6%) vs 1.8% (0.9-3.5%), before vs after) than in the placebo group (6.8% (2.2-15.7%) vs 4.9% (1.6-7.8%), before vs after, P<0.05). The percentage change in body weight correlated closely with the percentage loss of liver fat (r=0.70, P>0.0001). The decreases in liver fat were comparable between the rimonabant and the young historical hypocaloric diet groups. CONCLUSIONS: We conclude that, unlike in mice, in humans rimonabant decreases liver fat in proportion to weight loss.
Subject(s)
Cannabinoid Receptor Antagonists/therapeutic use , Fatty Liver/drug therapy , Liver/drug effects , Metabolic Syndrome/drug therapy , Obesity/drug therapy , Piperidines/therapeutic use , Pyrazoles/therapeutic use , Weight Loss/drug effects , Adult , Aged , Double-Blind Method , Fatty Liver/epidemiology , Fatty Liver/pathology , Female , Finland/epidemiology , Humans , Liver/pathology , Male , Metabolic Syndrome/epidemiology , Metabolic Syndrome/pathology , Middle Aged , Obesity/epidemiology , Obesity/pathology , Retrospective Studies , Rimonabant , Treatment OutcomeABSTRACT
BACKGROUND: Long-term use of both zidovudine (AZT) and stavudine (d4T) is associated with lipoatrophy, but it occurs possibly through different mechanisms. METHODS: Surgical biopsy specimens of subcutaneous adipose tissue were obtained from 18 human immunodeficiency virus type 1 (HIV-1)-infected lipoatrophic patients (the LA+ group) who were treated with either zidovudine (the AZT+LA+ group; n = 10) or stavudine (the d4T+LA+ group; n = 8) and from 10 nonlipoatrophic HIV-1-infected patients (the LA- group) who received antiretroviral therapy. Mitochondrial DNA (mtDNA) copy numbers, gene expression, and immunohistochemistry data were analyzed. RESULTS: mtDNA copy numbers were significantly reduced in the LA+ group, compared with the LA- group, and in the d4T+LA+ group, compared with the AZT+LA+ group. The ratio of mtDNA-encoded cytochrome COX3 to nuclear DNA-encoded COX4 expression was significantly lower in the LA+ group than in the LA- group. Compared with the LA- group, the LA+ group had significantly lower expression of genes involved in adipogenesis (SREBP1c and CEBPB), lipid (fatty acid synthase), and glucose (GLUT4) metabolism. Expression of genes involved in mitochondrial biogenesis (PGC1B), apoptosis (FAS), inflammation (IL1B), oxidative stress (PCNA and SOD1), and lamin B was significantly higher in the LA+ group than in the LA- group. The d4T+LA+ group had significantly lower expression of genes involved in mitochondrial biogenesis (POLG1), energy metabolism (the COX3/COX4 ratio), adipogenesis (SREBP1c and CEBPA), perilipin, and hexokinase than did the AZT+LA+ group. There were 7-fold more macrophages in adipose tissue specimens obtained from patients in the LA+ group, compared with the LA- group. CONCLUSIONS: Lipoatrophy is characterized by mtDNA depletion, inflammation, and signs of apoptosis. Changes were more profound in the d4T+LA+ group than in the AZT+LA+ group.