Proteomic analysis of aortae from human lipoprotein(a) transgenic mice shows an early metabolic response independent of atherosclerosis

Abstract

Background: Elevated low density lipoprotein (LDL) and lipoprotein(a) are independent risk factors for the development of atherosclerosis. Using a proteomic approach we aimed to determine early changes in arterial protein expression in transgenic mice containing both human LDL and lipoprotein(a) in circulation.

Methods and results: Plasma lipid analyses showed the lipoprotein(a) transgenic mice had significantly higher lipid levels than wildtype, including a much increased LDL and high density lipoprotein (HDL) cholesterol. Analysis of aortae from lipoprotein(a) mice showed lipoprotein(a) accumulation but no lipid accumulation or foam cells, leaving the arteries essentially atherosclerosis free. Using two-dimensional gel electrophoresis and mass spectrometry, we identified 34 arterial proteins with significantly altered abundance (P<0.05) in lipoprotein(a) transgenic mice compared to wildtype including 17 that showed a ≥2 fold difference. Some proteins of interest showed a similarly altered abundance at the transcript level. These changes collectively indicated an initial metabolic response that included a down regulation in energy, redox and lipid metabolism proteins and changes in structural proteins at a stage when atherosclerosis had not yet developed.

Conclusions: Our study shows that human LDL and lipoprotein(a) promote changes in the expression of a unique set of arterial proteins which may be early indicators of the metabolic disturbances preceding atherosclerosis.

Figure 1. Plasma lipids in wildtype and Lp(a) mice.

A, Lp(a) mice (n = 12) had significantly higher concentrations of total cholesterol (TC), triglyceride (TG) and phospholipids (PL) in plasma than wildtype mice (n = 12). Data is represented as mean concentration ± SEM. ^* P<0.05, ^*** P<0.001 versus wildtype. B, Plasma lipoproteins were separated by gel permeation chromatography and the cholesterol content of each fraction measured. Lp(a) mice had elevated LDL and HDL with the cholesterol relatively evenly distributed between the two. In the wildtype mice the cholesterol was largely all in the HDL.

Figure 2. Representative immunoblot of Lp(a) mouse plasma.

Plasma was separated by SDS PAGE under nonreducing conditions and subject to immunoblot analysis. A, An anti-human apoB-specific antibody was used to detect both LDL and Lp(a), indicating an excess of unbound apoB (LDL) to bound apoB [Lp(a)]. B, An anti-human apo(a)-specific antibody was used to detect Lp(a) and free apo(a), which showed all apo(a) was bound to human apoB in the form of Lp(a). Plasma from human apoB and human apo(a) only mice were included as controls.

Figure 3. Comparison of aorta lipids in wildtype and Lp(a) mice.

A, Total cholesterol (TC), triglyceride (TG) and phospholipids (PL) in homogenized aorta lipid extracts (n = 6). TC and TG concentrations in the aorta of Lp(a) mice were significantly reduced compared to wildtype mice. B, Lp(a) mice had a significantly elevated concentration of thiobarbituric acid-reactive substances (TBARS) in the aorta compared to wildtype suggesting an accumulation of aldehydes from lipid oxidation. Data represented as mean concentration ± SEM. ^** P<0.01 versus wildtype.

Figure 4. Representative histological analysis of aortae from wildtype and Lp(a) mice.

Aortic arches and sinuses of wildtype and Lp(a) mice (n = 8) were stained with haematoxylin and eosin (A and B) or Verhoeff's elastic stain and Curtis' modified van Gieson stain (C and D), which stained the elastic laminar (black), smooth muscle (brown), and collagen-rich fibrous tissue (red/pink). There was no evidence of atherosclerosis in the arteries of Lp(a) mice or wildtype mice, including no evidence of foam cells in the aortic arch or aortic sinus. Aortic arches were also immunostained with an anti-human Lp(a) antibody (brown) and counterstained with haematoxylin (blue). E, The wildtype mice were negative for Lp(a). F, The Lp(a) mice showed staining with the Lp(a)-specific antibody, indicating retention of Lp(a) in the arterial wall. Scale bar represents 100 µm.

Figure 5. Representative 2-D PAGE image of proteins expressed in aortae of Lp(a) mice.

A, Pooled aortic arch protein extracts (n = 12) were separated by 2-D PAGE in triplicate. Comparative analysis of equivalent spots between the averaged gels of the Lp(a) and wildtype mice identified 34 spots with significantly different intensities. Protein spots showing a significant difference in relative abundance (P<0.05) are indicated by numbers and their abbreviated protein names are listed in Table S1. B, Close-up images of selected protein spots showing ≥2 fold difference between wildtype and Lp(a) mice. In a few cases, multiple different spots of the same protein were identified indicative of post translational modifications to the protein. Proteins identified with multiple spots are denoted with (*).

Figure 6. RT-PCR analysis of transcripts of interest.

Quantitative RT-PCR was performed to investigate if proteins of interest showing a ≥2 fold change at the protein level were also regulated at the mRNA transcript level between the Lp(a) and wildtype mice. Total RNA was isolated from aorta samples (n = 6) of wildtype and Lp(a) mice. RNA was reverse transcribed to cDNA and quantitative PCR for transcripts of interest relative to 18S rRNA was performed. Glucose-6-phosphate dehydrogenase (G6pdx) and peroxiredoxin 4 (Prdx4) transcripts were increased in Lp(a) mice compared to wildtype. Dihydrolipopoyllysine succinyltransferase (Dlst), Glycerol-3-phosphate dehydrogenase (Gpd1) and fatty acid-binding protein 4 (Fabp4) transcripts were decreased in Lp(a) mice compared to wildtype. Isocitrate dehydrogenase (Idh3a) transcript showed no significant difference in Lp(a) mice compared to wildtype mice. Results are presented as relative levels of transcript normalized to wildtype. ^*** P<0.001 versus wildtype.