Purpose: S100A12 and fibroblast growth factor (FGF) 23 are biomarkers for cardiovascular mortality in patients with chronic kidney disease (CKD). The present study was designed to test the hypothesis that elevated serum levels of S100/calgranulin in mice with CKD would accelerate pathological cardiac remodeling, and to discover a potential link between S100A12 and FGF23.
Methods: Human S100/calgranulin is expressed in C57BL6/J mice using a bacterial artificial chromosome containing the genes and regulatory elements for hS1008/9 and hS100A12 (hBAC-S100). hBAC-S100 mice were bred with mice lacking the receptor for advanced glycation endproducts (RAGE). CKD was induced in hBAC-S100 mice and wild type (WT) littermate mice by surgical ligation of the ureters. The heart and aorta were analyzed after 10 weeks of CKD.
Results: S100A12 protein is expressed in myeloid cells and in serum of hBAC-S100 mice (25 ng/ml serum), and increased further in response to CKD. Moreover, serum IL-6 and IL-22 are elevated in hBAC-S100 mice at age 3 and 12 month. hBAC-S100 mice with CKD developed cardiac hypertrophy, diastolic dysfunction and ectopic cardiac calcification upon in-vivo echocardiography, heart weight, histology and gene expression of ANP, β-MHC, TGF-β, CTGF, and Col 1a1. Serum FGF23 was equally increased in WT/CKD and hBAC-S100/CKD compared to sham-operated animals. However, intra-cardiac expression of FGF23 mRNA (6 fold) and protein (2 fold) was increased only in hBAC-S100, and localized to the fibrous tissue of the aortic and mitral valve annulus. Importantly, IL-6, TNF-a, and LPS induced FGF23 mRNA and protein expression in cultured neonatal or adult cardiac fibroblasts while FGF23 mRNA was reduced by 90% in response to calcification-inducing medium (10mM glycerophosphate and 10nM dexamethasone). Recombinant S100A12 did not induce FGF23 mRNA. hBAC-S100 mice lacking RAGE, the receptor for S100/calgranulin, showed attenuation of systemic inflammation (IL-6, Il-22), attenuation of cardiac hypertrophy, diastolic dysfunction, cardiac FGF23 expression, and ectopic cardiac calcification when subjected to CKD, and the cardiac phenotype was similar to WT mice with CKD. Furthermore, aortic valve calcification develops spontaneously in 10-12-month old hBAC/S100 mice, suggesting that CKD accelerates this process, but is not required for S100/calgranulin-medicated aortic valve sclerosis.
Conclusion: Myeloid-derived S100/calgranulin is sufficient to induce a sustained inflammatory state in vivo, which is associated with increased FGF23 expression in cardiac fibroblasts and cardiac remodeling in a RAGE dependent manner.