The in vitro calcification potential of uremic serum increases with advancing chronic kidney disease

A Patidar1, DK Singh2, P Winocour3, K Farrington2, A Baydoun1. 1University of Hertfordshire, School of Life Sciences, AL10 9AB, UK, 2Lister Hospital, Renal Unit, SG1 4AB, UK, 3QEII Hospital, Department of Diabetes and Endocrinology, AL7 4HQ, UK

Vascular calcification strongly correlates with declining renal function in chronic kidney disease (CKD) and is closely regulated by circulating factors which are yet to be fully defined. Moreover, it is not clear whether serum can induce calcification outside disease settings or even whether the event correlates with disease vintage. We have therefore examined the ability of serum to induce calcification of smooth muscle cells using samples from age- and sex-matched subjects with and without advancing renal impairment. In addition, we have also analysed biomarker profiles in samples and correlated these with calcification in vitro.

Serum was obtained from controls (20 subjects) and from patient with moderate CKD (stage 3; 21 subjects), advanced CKD (stage 4 and 5; 19 subjects) and on haemodialysis (5D; 30 subjects). All subjects were Caucasian with no significant differences in age or blood pressure. Calcification was induced in rat cultured aortic smooth muscle cells (RASMCs) by incubating the latter with 10% serum. In further studies, cells were co-incubated with serum (10%) and CaCl2 (7mM) plus β-glycerophosphate (β-GP; 7mM). Calcium was quantified using the DICA-500 Ca+2 assay kit and the presence of calcific plaques on cell monolayers confirmed microscopically. The data is the mean±SD for normally distributed, and as median (interquartile range) if not normally distributed. Multivariate modelling was used to determine the independent predictors of calcification.

Sera from the CKD setting induced calcification of RASMC in vitro and this was significantly higher in cells treated with the 5D samples (µg Ca2+/µg protein: 1.2±0.08 vs 0.6±0.06 for CKD4/5 or 0.5±0.03 for CKD3). Addition of CaCl2 and β-GP enhanced the calcification caused by sera from CKD3 (52% increase above CKD3 control; p<0.001) and from CKD 4/5 (51% increase above CKD4/5 control; p<0.001) but not from 5D. Within the later, half the samples (referred to here as responders) induced lower calcification (0.9±0.4 µg Ca2+/µg protein) which increased by 30% in the presence of CaCl2 and β-GP while the other half (non-responders) caused higher calcification (1.5±0.4 µg Ca2+/µg protein) which was not enhanced by CaCl2 and β-GP. The responders had lower phosphate (1.5±0.4 vs 1.8±0.5 mmol/l in non-responders) but high levels of Matrix Gla protein (2.1±0.9 vs 1.8±0.5 ng/l in non-responders), high density lipoprotein (1.5±0.7 vs 1.0±0.3 mg/l in non-responders) and nuclear factor kappa B ligand (RANKL; (0.20±0.05 vs 0.15±0.07 pmol/l in non-responders) which all regulate calcification in vivo. In the whole group, the calcification potential of serum correlated positively with blood urea, serum creatinine, phosphate, parathyroid hormone and osteoprotegerin (OPG), and negatively with haemoglobin and albumin. In stepwise multivariate analysis the significant predictor of log transformed calcification potential was OPG (p=0.010). This factor however only explain around 50% of the variation in calcification potential (r2=0.506).

These findings confirm that serum from CKD can induce graded calcification outside of the disease milieu that reflects the degree of kidney impairment in vivo. Moreover, this may be regulated by the presence of biomarkers such as OPG, and in part by other regulators that are as yet unidentified.