Low clusterin levels in high-density lipoprotein associate with insulin resistance, obesity, and dyslipoproteinemia

Abstract

Objective: To determine whether obesity and insulin resistance associate with changes in the protein content of high-density lipoprotein (HDL) in 2 different groups of men by using targeted proteomics.

Methods and results: Insulin resistance and obesity are hallmarks of type 2 diabetes mellitus and the metabolic syndrome, which confer an increased risk of cardiovascular disease. Recent studies suggest that the protein cargo of HDL makes important contributions to the lipoprotein's cardioprotective effects. In a discovery study, we used isotope dilution mass spectrometry to quantify the relative concentrations of 5 proteins previously implicated in HDL's cardioprotective effects in 3 groups of healthy subjects: lean insulin-sensitive, lean insulin-resistant, and obese insulin-resistant individuals. We validated our findings in a different group of subjects. The clusterin concentration in HDL strongly and negatively associated with insulin resistance and body mass index in both populations. HDL clusterin levels were lower in subjects with low HDL and high triglycerides, key components of the metabolic syndrome. There was an inverse correlation between clusterin levels in HDL and very-low-density lipoprotein/low-density lipoprotein.

Conclusions: Clusterin levels in HDL are lower in men with reduced insulin sensitivity, higher body mass index, and an unfavorable lipid profile. Our observations raise the possibility that clusterin depletion contributes to the loss of HDL's cardioprotective properties.

Figures

Figure 1. Relationships between relative HDL clusterin concentration and clinical characteristics

Relative HDL clusterin concentration (per g total HDL protein) is plotted against loge-transformed BMI (A), loge-transformed HOMA-IR (B), and mean arterial blood pressure (C). Results represent the discovery population (open circles) and validation population (closed circles). Lines represent linear regression analysis of the relationship between HDL clusterin and clinical parameters for the discovery (dashed lines), validation (dotted lines), and combined populations (solid lines). Pearson correlation coefficients (r) and associated p values are presented for the combined population.

Figure 2. Relative concentration of HDL clusterin in control subjects, and subjects with diabetes, the metabolic syndrome, or dyslipoproteinemia

Diabetic subjects were compared with age-matched normal subjects in the validation population (A). Subjects with metabolic syndrome were compared with normal subjects in the discovery population (B) as described in Supplemental Methods. Also shown are subjects categorized by plasma HDL-C (C; normal ≥40 mg/dL), plasma triglycerides (D; normal

Figure 3. Correlations of lipid parameters with relative HDL clusterin concentration

Linear regression analysis is shown for HDL clusterin and plasma HDL-C (A), plasma triglycerides (B), and clusterin concentration in VLDL/LDL particles (C). Data are presented as in Fig. 1, except that the regression line presented in C is for the validation population only.