Serum retinol-binding protein-induced endothelial inflammation is mediated through the activation of toll-like receptor 4

PURPOSE Elevation of serum retinol-binding protein 4 (RBP4) induces inflammation in primary human retinal microvascular endothelial cells (HRECs) via a retinol-independent mechanism; thus, it may play a causative role in the development and progression of vascular lesions in diabetic retinopathy (DR). Since HRECs do not express the classical RBP4 receptor, stimulated by retinoic acid gene 6 (STRA6), this study focuses on identifying the endothelial cell receptor and signaling that mediate RBP4-induced inflammation. METHODS HRECs were treated with a toll-like receptor 4 (TLR4) small molecule inhibitor (Cli95, also known as TAK-242), TLR4 neutralizing antibody, or mitogen-activated protein kinase (MAPK) inhibitors before treatment with purified recombinant RBP4. The HREC inflammatory response was quantified by in vitro leukostasis assays, western blotting, and enzyme-linked immunosorbent assay (ELISA). To understand how the serum binding partner for RBP4, transthyretin (TTR), may affect RBP4 activity, we also measured RBP4 and TTR levels in serum and retinal lysates from RBP4-Tg and wild-type mice. RESULTS TLR4 inhibition significantly reduced RBP4-induced expression of pro-inflammatory proteins and in vitro leukostasis. RBP4 treatment significantly increased phosphoactivation of p38 and c-Jun N-terminal protein kinase (JNK). The p38 inhibitor (SB203580) attenuated RBP4-stimulated vascular cell adhesion molecule 1 (VCAM-1), intracellular adhesion molecule 1 (ICAM-1), monocyte chemoattractant protein (MCP-1), and interleukin 6 (IL-6) production, while the JNK inhibitor (SP600125) reduced RBP4-stimulated sICAM-1, endothelial cell selectin (E-selectin), and MCP-1 production. The MAPK inhibitors only showed partial (50-70%) suppression of the RBP4-stimulated proinflammatory response. Moreover, TLR4 inhibition did not decrease RBP4-induced MAPK phosphoactivation, suggesting that RBP4-mediated MAPK activation is TLR4 independent and occurs through a secondary unknown receptor. We also found that the RBP4/TTR molar ratio was exceptionally high in the retina of RBP4-Tg mice, indicating an abundance of TTR-free RBP4. CONCLUSIONS RBP4-induced inflammation is largely mediated by TLR4, and in part, through JNK and p38 MAPK signaling. The high TTR/RBP4 molar ratio in serum likely protects the endothelium from the proinflammatory effects of RBP4 in vivo, whereas elevation of serum RBP4 causes a significant increase in TTR-free RBP4 in retinal tissue. This offers insight into how RBP4-Tg mice can develop retinal neurodegeneration without coincident retinal microvascular pathology.