A biproduct of the worldwide obesity epidemic is that nonalcoholic fatty liver disease (NAFLD) is now the most common chronic liver condition, ranging from isolated steatosis to more-severe nonalcoholic steatohepatitis (NASH) and NAFLD-related cirrhosis. In addition to liver-related morbidity, NAFLD is associated with other metabolic derangements and increased cardiovascular disease (CVD).[2,3] Because the liver plays a key role in lipoprotein synthesis, the lipoprotein proteome may have utility in assessing the severity of liver disease. We previously developed a method for enrichment of high-density lipoprotein (HDL) particles in human serum that enabled deployment of a multiplexed, quantitative, targeted proteomic method to identify key proteins for estimation of cholesterol efflux capacity and CVD risk.[4,5] In the current study, we applied a similar proteomic approach to assess whether targeted proteins involved in renal and hepatic function could be used to help assess NAFLD severity.
Patients (N=185) aged 21 to 70 with no evidence of secondary causes for liver steatosis or use of medications that affect intrahepatic triglyceride content were recruited for this study. 45(24%) did not have NAFLD and were used as controls. The remaining 140 patients underwent percutaneous liver biopsy and were divided based on presence or absence of advanced fibrosis (fibrosis stage 3 or 4).
HDL particles from human serum were enriched by affinity to lipid-free, stable isotope labeled, His-tagged Apolipoprotein A-I as previously described . Enriched fractions were subjected to Lys-C digestion for subsequent LC-MS/MS analysis with a 3-stream Agilent 1260 HPLC coupled to an Agilent 6495 QqQ mass spectrometer using StreamSelect. A multiple-reaction monitoring method was designed to target 28 proteins associated with lipid transport and metabolism in addition to hepatic and renal function.
Comparisons between two groups were performed by t-test or Kruskal-Wallis for continuous measures (depending on distribution) and by chi-squared or Fisher’s exact test for categorical measures. Comparisons among >2 groups were performed by one-way ANOVA. Differences in HDL-bound proteins were adjusted for multiple comparisons, controlling for a false discovery rate of 0.10 by the Benjamini-Hochberg method (p<0.017 was considered statistically significant).
Of 28 proteins quantified, 5 were significantly different between groups with and without advanced fibrosis after adjustment for multiple testing. ApoC-I (p=0.001), ApoC-IV (p=010), ApoM (p=0.015), LCAT (p=0.014), and SAA4 (p=0.015) were significantly reduced in advanced fibrosis patients, even after adjustment for presence of diabetes.
Differences were observed in levels of these 5 proteins with respect to disease progression. While ApoC-I and ApoC-IV demonstrated a stepwise reduction with progression of fibrosis from moderate (stage 2) to advanced (stages 3 and 4), ApoM, LCAT, and SAA4 demonstrated abrupt reduction in advanced fibrosis, but with no apparent reduction in moderate fibrosis.
When comparing patients with and without NAFLD, two proteins (ApoA-I [p=0.001] and PON3 [p<0.001]) were significantly reduced in patients with NAFLD after adjusting for multiple comparisons and for the presence of diabetes. Notably, the same 2 proteins were also independently associated for the presence of diabetes (p<0.001 for both). These results are in-line with previous studies in patients with chronic liver disease resulting from NAFLD and other etiologies, suggesting a mechanistic link between fibrosis and lipoprotein synthesis impairment, and a direct relationship between liver disease and CVD risk.
1. Godinez-Leiva, E. et al Curr Hypertens Rev (2021) 17, 94-111.
2. Bril, F. et al J Clin Endocrinol Metab (2016) 101, 644-652.
3. Targher G. et al J Hepatol (2016) 65, 589-600.
4. Collier T. et al J Proteome Res (2018) 17, 1183-93.
5. Jin, Z. et al Clin Chem (2019) 65, 282-290.