Impairment in ischemia-induced neovascularization in diabetes: bone marrow mononuclear cell dysfunction and therapeutic potential of placenta growth factor treatment

Abstract

Mechanisms that hinder ischemia-induced neovascularization in diabetes remain poorly understood. We hypothesized that endogenous bone marrow mononuclear cell (BM-MNC) dysfunction may contribute to the abrogated postischemic revascularization reaction associated with diabetes. We first analyzed the effect of diabetes (streptozotocin, 40 mg/kg) on BM-MNC pro-angiogenic potential in a model of surgically induced hindlimb ischemia. In nondiabetic animals, transplantation of BM-MNCs isolated from nondiabetic animals raised the ischemic/nonischemic angiographic score, capillary number, and blood flow recovery by 1.8-, 2.7-, and 2.2-fold, respectively, over that of PBS-injected nondiabetic animals (P < 0.05). Administration of diabetic BM-MNCs also improved the neovascularization reaction in ischemic hindlimbs of nondiabetic mice but to a lesser extent from that observed with nondiabetic BM-MNC transplantation. In diabetic mice, injection of nondiabetic BM-MNCs was still more efficient than that of diabetic BM-MNCs. Such BM-MNC dysfunction was associated with the impairment of diabetic BM-MNC capacity to differentiate into endothelial progenitor cells (EPCs) in vitro and to participate in vascular-like structure formation in a subcutaneous Matrigel plug. Placenta growth factor (PlGF) administration improved by sixfold the number of EPCs differentiated from diabetic BM-MNCs in vitro and enhanced ischemic/nonischemic angiographic score, capillary number and blood flow recovery by 1.9-, 1.5- and 1.6-fold, respectively, over that of untreated diabetic animals (P < 0.01). Endogenous BM-MNC pro-angiogenic potential was affected in diabetes. Therapeutic strategy based on PlGF administration restored such defects and improved postischemic neovascularization in diabetic mice.

Figure 1

A: Representative microangiography of the right ischemic and left nonischemic hindlimbs in nondiabetic mice treated with BM-MNCs isolated from nondiabetic or diabetic mice. B: Ischemic/nonischemic angiographic score at 10 days following ischemic injury in diabetic and nondiabetic mice treated with BM-MNCs. Values are means ± SEM, n = 7 per group. *P < 0.05 versus PBS-injected nondiabetic animals. ^†P < 0.05 versus mice treated with nondiabetic BM-MNCs, #P < 0.05 versus PBS-injected diabetic animals. PBS indicates mice receiving PBS; nondiabetic BM-MNCs, animals treated with BM-MNCs isolated from nondiabetic animals and diabetic BM-MNCs, animals treated with BM-MNCs isolated from diabetic mice.

Figure 2

A: Representative photomicrographs of ischemic muscle sections from nondiabetic mice treated with BM-MNCs isolated from nondiabetic or diabetic mice, hybridized with antibody directed against total fibronectin. Capillaries appear in white and myocytes in black. B: Ischemic and nonischemic capillary density. Values are means ± SEM, n = 7 per group. *P < 0.05, **P < 0.01 versus PBS-injected nondiabetic animals; ^†P < 0.05 versus mice treated with nondiabetic BM-MNCs; #P < 0.05, ##P < 0.01 versus PBS-injected diabetic animals. See legend to Figure 1.

Figure 3

A: Ischemia-induced changes in hindlimb blood flow monitored in vivo by laser Doppler perfusion imaging performed at day 10 after ischemia in nondiabetic mice treated with BM-MNCs isolated from nondiabetic or diabetic mice. In color coded images, normal blood flow is depicted in red. A marked reduction in blood flow of ischemic hindlimb is depicted in blue. B: Quantitative evaluation of blood flow expressed as a ratio of blood flow in ischemic foot to that in nonischemic one. Values are means ± SEM, n = 7 per group. *P < 0.05 versus PBS-injected nondiabetic animals. ^†P < 0.05 versus mice treated with nondiabetic BM-MNCs, #P < 0.05 versus PBS-injected diabetic animals. See legend to Figure 1.

Figure 4

A: Representative images of EPCs isolated from bone marrow of nondiabetic and diabetic mice with femoral artery ligature. EPCs were characterized as adherent cells with double-positive staining for AcLDL-Dil and BS-1 lectin. B: Representative phase contrast image showing spindle-shaped attaching cells from bone marrow of nondiabetic ischemic mice. C: Representative images of EPCs isolated from bone marrow of ischemic nondiabetic mice showing that most of these cells were negative for CD18. D: Quantification of AcLDL-Dil and BS-1 lectin-positive cells in nondiabetic and diabetic mice. Values are means ± SEM, n = 5 per group. ***P < 0.001, versus nondiabetic mice with femoral artery ligature.

Figure 5

A: Top, representative photomicrographs of Matrigel sections from Matrigel treated with BM-MNCs isolated from nondiabetic mice stained with Masson Trichrome at a magnification of 10×, 20×, or 40×. Bottom, representative photomicrographs of Matrigel sections from Matrigel treated with fluorescent-labeled BM-MNCs isolated from nondiabetic mice at a magnification of 10× or 20×. B: Top, representative photomicrographs of Matrigel sections from Matrigel treated with BM-MNCs isolated from diabetic mice stained with Masson Trichrome at a magnification of 10×, 20×, or 40×. Bottom, representative photomicrographs of Matrigel sections from Matrigel treated with fluorescent-labeled BM-MNCs isolated from diabetic mice at a magnification of 10× or 20×.

Figure 6

A: Ischemic/nonischemic angiographic score at 7 days following ischemic injury in diabetic and nondiabetic mice treated with or without PlGF. B: Ischemic/nonischemic capillary number at 7 days following ischemic injury in diabetic and nondiabetic mice treated with or without PlGF. C: Ischemic/nonischemic limb perfusion at 7 days following ischemic injury in diabetic and nondiabetic mice treated or not with PlGF. Values are means ± SEM, n = 6 per group. ***P < 0.01 versus untreated nondiabetic animals. ^†††P < 0.001 versus untreated diabetic animals.