RAGE ectodomain shedding (sRAGE) and splice variant (esRAGE) under hyperglycemic and inflammatory conditions.
L. Ashall1, S. A. Thomas1, J. Hu2, C. K. Inman1, H. Li2, R. Ramasamy1&3* and A. M. Schmidt1&3*.
1 Diabetes Research Center, NYUAD, Abu Dhabi, UAE.  2 Department of Population Health, NYU School of Medicine, NY, USA. 3 Diabetes Research Program, Division of Endocrinology, NYU School of Medicine, NYU, USA. 
*Co-senior authors.
Background: Obesity, diabetes and its complications represent an increasing worldwide health epidemic, however little is known about the specific cellular and molecular mechanisms that drive these complications. Central features of obesity and diabetes are hyperlipidemia and hyperglycemia respectively, which create a cellular environment conducive to increased formation of advanced glycation end products (AGEs). AGEs bind to their chief cell surface receptor for AGE (RAGE), thereby activating downstream signaling pathways, including NF-κB, which leads to the transcription of genes that in part create an inflammatory environment and are hypothesized to contribute to the development of diabetic complications. RAGE has various splice variants and isoforms, a full length (fl) membrane bound RAGE (flRAGE), which can be cleaved from the cell surface by metalloprotease (MMP-9) and ADAM10 to produce soluble RAGE (sRAGE). A second form of soluble RAGE is alternatively spliced and secreted by the cell known as endogenous secretory RAGE (esRAGE or RAGEv1). Both sRAGE and esRAGE act as decoys to RAGE as they can bind to and sequester RAGE ligands, potentially preventing ligand binding to flRAGE and accompanying downstream signal transduction.
Aims: The objective of this study is to investigate RAGE ectodomain shedding and the splice variants of RAGE. Although studies have examined the regulation of RAGE shedding, little investigation into both sRAGE and esRAGE in cells exposed to stresses that provoke vascular damage in diabetes, such as hyperglycemia with the addition of a prototypic inflammatory mediator, has yet to be explored. We hypothesized that the glucose/AGE-stimulated upregulation of inflammatory mechanisms might contribute to the generation of soluble RAGEs.
Method: RAGE-expressing cells were exposed to either normal glucose (NG, 5.5 mM D-glucose) or diabetes-relevant levels of high glucose (HG, three to five weeks 25 mM D-glucose) and to the prototypic inflammatory stimulus TNFα to test potential modulation of esRAGE and sRAGE levels. Using monocyte/macrophage-like human THP-1 cells, we measured endogenous mRNA levels of esRAGE and AGER, the gene encoding RAGE. As the levels of sRAGEs produced by these cells were below the limit of detection in the ELISAs, we employed HEK293T cells expressing either esRAGE or AGER, which facilitated the measurement of the sRAGEs.  
Results: The production of esRAGE remained unchanged when esRAGE expressing HEK293T cells were exposed to HG. However, TNFα treatment alone resulted in a significant (p<0.05) increase in RAGE splice variant levels under NG conditions measured by esRAGE ELISA, which was also observed at the mRNA and protein level in both NG and HG conditions. Further mechanistic analyses in THP-1 cells showed that BAY 11-7082 inhibition of NF-κB reduced the mRNA expression of esRAGE following TNFα treatment. Phosphorylation of p65 was also observed prior to esRAGE expression. Exposure of RAGE expressing HEK293T cells to HG was also not sufficient to induce changes in RAGE ectodomain shedding, whereas HG together with TNFα treatment resulted in a significant (p<0.05) increase in RAGE ectodomain shedding as determined by ELISA. Under the same conditions, RAGE protein levels in the cell lysate remained unchanged indicating that the increase in ectodomain shedding was not solely due to increased endogenous flRAGE levels. No change in ADAM10 protein level was observed under HG or HG with TNFα treatment compared to NG conditions, however ADAM10 activity increased in cells exposed to HG alone. At the mRNA level, an increase in MMP-9 transcripts following TNFα treatment in both NG and HG conditions was noted. Knockdown of ADAM10 or MMP-9 using siRNAs resulted in >80 % reduction in sRAGE levels, compared to control scrambled siRNA. 
Discussion: Both isoforms of soluble RAGE increased in the presence of conditions that mimic diabetic complications, with the addition of TNFα treatment being a key factor. NF-κB plays an important role in the regulation of esRAGE, with ADAM10 and MMP-9 being critical to the regulation of shedding of sRAGE in both hyperglycemic and inflammatory conditions. With a better understanding of the mechanisms regulating the production of esRAGE and sRAGE under stresses that mimic vascular damage in diabetes, novel therapeutic strategies may be identified to block adverse AGE-RAGE consequences.