High-mobility group box-1 protein promotes angiogenesis after peripheral ischemia in diabetic mice through a VEGF-dependent mechanism

Abstract

Objective: High-mobility group box-1 (HMGB1) protein is a nuclear DNA-binding protein released from necrotic cells, inducing inflammatory responses and promoting tissue repair and angiogenesis. Diabetic human and mouse tissues contain lower levels of HMGB1 than their normoglycemic counterparts. Deficient angiogenesis after ischemia contributes to worse outcomes of peripheral arterial disease in patients with diabetes. To test the hypothesis that HMGB1 enhances ischemia-induced angiogenesis in diabetes, we administered HMGB1 protein in a mouse hind limb ischemia model using diabetic mice.

Research design and methods: After the induction of diabetes by streptozotocin, we studied ischemia-induced neovascularization in the ischemic hind limb of normoglycemic, diabetic, and HMGB1-treated diabetic mice.

Results: We found that the perfusion recovery was significantly attenuated in diabetic mice compared with normoglycemic control mice. Interestingly, HMGB1 protein expression was lower in the ischemic tissue of diabetic mice than in normoglycemic mice. Furthermore, we observed that HMGB1 administration restored the blood flow recovery and capillary density in the ischemic muscle of diabetic mice, that this process was associated with the increased expression of vascular endothelial growth factor (VEGF), and that HMGB1-induced angiogenesis was significantly reduced by inhibiting VEGF activity.

Conclusions: The results of this study show that endogenous HMGB1 is crucial for ischemia-induced angiogenesis in diabetic mice and that HMGB1 protein administration enhances collateral blood flow in the ischemic hind limbs of diabetic mice through a VEGF-dependent mechanism.

FIG. 1.

A: Foot blood flow monitored in vivo by LDPI in control normoglycemic and diabetic mice. Representative evaluation of the ischemic (right) and nonischemic (left) hind limbs, immediately after, and on days 7, 14, 21, and 28 after surgery. In color-coded images, red indicates normal perfusion and blue indicates a marked reduction in blood flow in the ischemic hind limb. Blood flow recovery is impaired in diabetic mice compared with normoglycemic mice. The blood flow of the ischemic hind limb is expressed as the ratio between the perfusion of the ischemic limb and the uninjured limb. P < 0.05 and P < 0.01 vs. diabetic mice. B: Representative photomicrographs of ischemic muscle sections from control normoglycemic and diabetic mice stained with antibody directed against VEGF, 7 days after surgery, and against CD31, 28 days after surgery. Positive staining appears in brown. Magnification ×20. C: Number of vessels per cross section is significantly reduced in diabetic mice with respect to normoglycemic mice. P < 0.05 vs. diabetic mice. D: Representative Western blot of VEGF protein content in the ischemic legs of control and diabetic mice on postoperative day 7. VEGF expression is reduced in the ischemic tissue of diabetic mice compared with control mice. ns, not significant. (A high-quality digital representation of this figure is available in the online issue.)

FIG. 2.

A: Representative photomicrographs of nonischemic and ischemic muscle sections from control normoglycemic and diabetic mice stained with antibody directed against HMGB1 7 days after surgery. Magnification ×20. Positive staining appears in brown. B: Representative Western blot of HMGB1 protein content in the ischemic legs of control and diabetic mice on postoperative day 7. HMGB1 expression is reduced in the ischemic tissue of diabetic mice compared with control mice. P < 0.05 vs. diabetic mice. C: Representative photomicrographs of ischemic muscle sections from control normoglycemic and diabetic mice stained with antibody directed against CD45, with hematoxylin-eosin, and with transferase-mediated dUTP nick-end labeling, 7 days after surgery. Magnification ×20. For CD45 (leukocyte infiltration) and transferase-mediated dUTP nick-end labeling assay (apoptotic cells) positive staining appears in brown. D: Evaluation of leukocyte infiltration, apoptosis, and necrosis in the ischemic muscle sections from control normoglycemic and diabetic mice. There are no differences between control and diabetic mice according to all three aspects. ns vs. diabetic mice. (A high-quality digital representation of this figure is available in the online issue.)

FIG. 3.

A: LDPI ratio in control and BoxA-treated mice. Representative evaluation of the ischemic (right) and nonischemic (left) hind limbs immediately after and on days 7, 14, 21, and 28 after surgery. Blood flow recovery is impaired in BoxA-treated mice compared with vehicle-treated mice. The blood flow of the ischemic hind limb is expressed as the ratio between the perfusion of the ischemic limb and the uninjured limb. P < 0.05 and P < 0.01 vs. BoxA-treated mice. B: Representative photomicrographs of ischemic muscle sections from control and BoxA-treated mice stained with antibody directed against CD31 28 days after surgery. Positive staining appears in brown. Magnification ×20. C: The number of vessels per cross section is significantly reduced in BoxA-treated mice compared with vehicle-treated mice. P < 0.05 vs. BoxA-treated mice. (A high-quality digital representation of this figure is available in the online issue.)

FIG. 4.

LDPI ratio in diabetic mice treated with 200, 400, 600, and 800 ng of HMGB1 and with PBS (control group). Representative evaluation of LDPI ratio immediately after and on days 7, 14, 21, and 28 after surgery. HMGB1 administration restored blood flow recovery in diabetic mice compared with PBS-treated diabetic mice. The blood flow of the ischemic hind limb is expressed as the ratio between the perfusion of the ischemic limb and the uninjured limb. P < 0.05 and P < 0.01 vs. HMGB1-treated mice. (A high-quality digital color representation of this figure is available in the online issue.)

FIG. 5.

A: Representative photomicrographs of ischemic muscle sections from diabetic mice treated with 200, 400, 600, and 800 ng of HMGB1 and PBS (control group), stained with antibody directed against CD31 28 days after surgery. Positive staining appears in brown. Magnification ×20. B: The number of vessels per cross section is significantly increased in HMGB1-treated mice compared with untreated mice. P < 0.05 vs. PBS-treated control mice. C: Representative Western blot evaluation of VEGF protein content, 3 days after surgery in the ischemic legs of HMGB1-treated and PBS-treated mice. P < 0.05 vs. mice treated with PBS. (A high-quality digital color representation of this figure is available in the online issue.)

FIG. 6.

A: Representative Western blot evaluation of VEGFR Flt-1 and Flk-1 protein content and their phosphorylated/activated isoforms (p-Flt-1 and p-Flk-1), 7 days after surgery, in the ischemic legs of diabetic mice previously treated with sFlt-1 or empty vector (control group). sFlt-1 treatment strongly reduced VEGFR phosphorylation, confirming the inhibition of the VEGF pathway. B: LDPI ratio in normoglycemic or diabetic mice previously treated with sFlt-1 or empty vector. Representative evaluation of LDPI ratio immediately after and on days 7, 14, 21, and 28 after surgery. VEGF inhibition attenuates postischemic angiogenesis in nondiabetic mice, but this group showed a better angiogenic response compared with diabetic animals. The blood flow of the ischemic hind limb is expressed as the ratio between the perfusion of the ischemic limb and the uninjured limb. P < 0.01 and P < 0.05 vs. sFlt-1-treated nondiabetic mice or vs. diabetic mice. C: LDPI ratio of the diabetic mice treated with 200, 400, 600, and 800 ng of HMGB1 previously treated with sFlt-1 or empty vector (control group). Representative evaluation of LDPI ratio immediately after and on days 7, 14, 21, and 28 after surgery. HMGB1-induced blood flow recovery in the diabetic mice is impaired when VEGF activity is inhibited. The blood flow of the ischemic hind limb is expressed as the ratio between the perfusion of the ischemic limb and the uninjured limb. P < 0.05 and P < 0.01 vs. sFlt-1-treated mice. D: The number of vessels per cross section is significantly reduced in HMGB1 + sFlt-1–treated mice compared with the HMGB1-treated mice that received the empty vector. P < 0.05 vs. sFlt-1–treated mice.