A. D. Morris1, H. Li1, L. Brewster1,2 1Emory University Hospital,Division of Vascular Surgery,Atlanta, GA, USA 2Atlanta VA Medical Center,Division of Vascular Surgery,Atlanta, GA, USA
Introduction: Diabetes increases the risk of cardiovascular disease and major amputation. Mesenchymal Stem Cells(MSCs) have robust regenerative potential, but concerns remain that MSCs from diabetic patients or under diabetic conditions may limit therapeutic effect. The objective of this project was to compare healthy MSCs(hMSCs) and diabetic MSCs(dMSCs) angiogenic activity under standard and high glucose conditions.
Methods: MSCs were harvested from 4 patients (2 healthy and 2 diabetic). Human aortic endothelial cells (HAECs) (20,000 cells), HAECs with hMSCs(10,000/10,000 cells) or HAEC with dMSC(10,000/10,000 cells) were formed into pellets, suspended in fibrin gel, and cultured in 5mM, 20mM, or 40mM glucose media. Cell invasion was captured by daily microscopy for 3 days. To obtain secretome, hMSCs and dMSCs were cultured under 5mM glucose, 20mM glucose, or 40mM glucose in 10% FBS media for 24 hours. Secretome analysis was performed using a commercial microarray for 20 individual angiogenic factors. Densitometry using positive control as reference allowed comparison of relative secretion.
Results: High glucose environments did not significantly inhibit HAEC invasion at any time point. At day 3, co-culture with hMSCs improved HAEC invasion under high glucose conditions(20mM and 40mM, p= 0.004 and p=0.02) but not in 5mM glucose media(p=.24). dMSC co-culture also improved HAEC invasion at day 3 (p<0.001 for all conditions). On day 3, dMSCs encouraged HAEC invasion more than hMSCs in 5mM glucose and 40mM glucose, and equivalently in 20mM glucose (p<0.01,p<0.01,p=0.14). hMSC co-culture improved HAEC invasion in 20mM glucose compared to 5mM and 40mM glucose conditions(Day 2:p=0.034,p<0.001, Day 3:p=0.029,p=0.013). On day 1, 40mM glucose inhibited HAEC invasion in dMSC co-culture compared to 5mM glucose(p=0.002), but that difference did not persist on day 2 or day 3(p=0.366, p=0.176).
Compared to hMSCs, dMSCs secreted increased levels of CXCL-1(p=0.002). Secretion of CXCL-1 was an average of 6.5 times greater by dMSCs versus hMSCs. Additionally, 20mM glucose environments stimulated secretion of CXCL-1 compared to 5mM and 40mM glucose in dMSCs(p=0.043, p=0.021).
Conclusions: We have shown that addition of hMSCs and dMSCs are capable of improving cell invasion. Diabetic MSCs are not inferior to hMSCs and may have superior function in certain instances. While dMSC and hMSC are functionally similar, dMSCs secrete the unusual angiogen CXCL-1 at higher levels. We theorize that CXCL-1 may serve as an alternate pathway for encouraging endothelial function under diabetic conditions. With future in vivo experiments, we hope to show dMSCs retain their function by alternate mechanisms and dMSCs can be used in clinical trials with equal efficacy.
