Increased biglycan in aortic valve stenosis leads to the overexpression of phospholipid transfer protein via Toll-like receptor 2

Abstract

Aortic stenosis (AS) is the most common valvular heart disease, and it is suspected that atherosclerotic mechanisms are involved in the development of this disorder. Therefore, the retention of lipids within the aortic valve may play a role in the pathobiology of AS. In this study, a gene expression microarray experiment was conducted on human aortic valves with and without AS. The expression levels of transcripts encoding proteoglycans and enzymes involved in lipid retention were compared between the two groups. The microarray results were subsequently replicated in a cohort of 87 AS valves and 36 control valves. In addition, the interaction between proteoglycan and lipid-modifying enzyme was documented in isolated valve interstitial cells (VICs). The microarray results indicated that only biglycan (BGN) and phospholipid transfer protein (PLTP) were overexpressed in the AS valves. These results were then confirmed by quantitative PCR. The immunohistochemical analysis revealed a colocalization of BGN, PLTP, and Toll-like receptor-2 (TLR 2) in AS valves. In vitro, we showed that BGN induces the production of PLTP in VICs via the stimulation of TLR 2. Thus, increased accumulation of BGN in AS valves contributes to the production of PLTP via TLR 2. These results suggest that intricate links between valve matrix proteins, inflammation, and lipid retention are involved in the pathobiology of AS.

Figure 1

Biglycan (BGN) and phospholipid transfer protein (PLTP) mRNA transcript levels, measured with quantitative PCR, in aortic stenosis valves compared with control tissues (A and B). There was a significantly increased number of mRNA copies of BGN (P < 0.0001; A) and PLTP (P < 0.0001; B) in aortic stenosis valves compared with control noncalcified valves.

Figure 2

Immunolocalization of biglycan (BGN) and phospholipid transfer protein (PLTP) in control and aortic stenosis valves (A–D; ×100). In control valves BGN and PLTP were present in some areas but appear as a faint marking with respectively an extra-cellular and cellular localization (A and C). In contrast, in aortic stenosis valves, BGN (B) and PLTP (D) were immunodetected and highly expressed, particularly in the fibrosa layer. Quantitative morphometric analyses revealed that AS valves had a greater percentage area of immunomarking for BGN and an increased cellular density of PLTP positive cells in AS valves when compared with control valves (E and F).

Figure 3

A–D: Immunohistochemistry studies showing the immunolocalization of biglycan (BGN), oxidized-LDL (ox-LDL), phospholipid transfer protein (PLTP), and apolipoprotein AI (Apo AI). BGN was immunolocalized in the fibrosa and ventricularis layers (A). Ox-LDL were detected and present in the fibrosa (B), where it colocalized with PLTP (C) and Apo AI (D; ×200).

Figure 4

Double immununostainings confirmed that biglycan (BGN; red) colocalized with phospholipid transfer protein (PLTP; brown; A). In addition, we documented that BGN (brown) colocalized with smooth muscle α-actin (red; B) along with CD90 (red; arrows; C). Also, PLTP (brown) was found to colocalize with smooth muscle α-actin (red; arrows; D) and with CD90 (red; arrows; E). Furthermore, BGN (brown; F) and PLTP (brown; G) were documented to colocalize with TLR 2 (arrows; red). Of note, TLR 4 (red; arrows) was found to colocalize only with BGN (brown; H) and not PLTP (not shown).

Figure 5

A–D: Correlations between biglycan (BGN), phospholipid transfer protein (PLTP), and nuclear factor kappa B (NF-κB) mRNA transcript levels along with plasma LDL particle size in aortic stenosis valves. BGN transcript level was negatively correlated with LDL particle size (r = −0.18, P = 0.01; A). The mRNA transcript levels of BGN and PLTP were positively correlated (r = 0.48, P < 0.0001; B). BGN and PLTP mRNA transcript levels were positively correlated with NF-κB levels in aortic stenosis valves (r = 0.56, P < 0.0001; r = 0.43, P = 0.001; C and D).

Figure 6

A–E: Expression of phospholipid transfer protein (PLTP) and nuclear factor kappa B (NF-κB) in isolated valve interstitial cells (VICs) on stimulation with biglycan (BGN), bacterial lipopolysaccharide (LPS), and peptidoglycan (PGN). Using immunoblotting we found that after 48 hours of stimulation with BGN (100 ng/ml), LPS (200 ng/ml), and PGN (10 μg/ml), PLTP was produced by VICs (A). Also, mRNA transcript levels of PLTP and NF-κB were significantly increased on stimulation for 48 hours with BGN, LPS, and PGN (B and C). Addition of anti–Toll receptor 2 and anti–Toll receptor 4 antibodies in the medium of stimulated cells with BGN revealed that the BGN-induced production of PLTP and NF-κB were abrogated only by anti–Toll receptor 2 antibody (D and E). *P < 0.01 compared with control (cntrl); BGN + NegCtrl: biglycan and an irrelevant human control antibody.