Background-Bone marrow implantation (BMI) was shown to enhance angiogenesis in a rat ischemic heart model. This preclinical study using a swine model was designed to test the safety and therapeutic effectiveness of BMI. Methods and Results-BM-derived mononuclear cells (BM-MNCs) were injected into a zone made ischemic by coronary artery ligation. Three weeks after BMI, regional blood flow and capillary densities were significantly higher (4.6-and 2.8-fold, respectively), and cardiac function was improved. Angiography revealed that there was a marked increase (5.7-fold) in number of visible collateral vessels. Implantation of porcine coronary microvascular endothelial cells (CMECs) did not cause any significant increase in capillary densities. Labeled BM-MNCs were incorporated into Ϸ31% of neocapillaries and corresponded to Ϸ8.7% of macrophages but did not actively survive as myoblasts or fibroblasts.There was no bone formation by osteoblasts or malignant ventricular arrhythmia. Time-dependent changes in plasma levels for cardiac enzymes (troponin I and creatine kinase-MB) did not differ between the BMI, CMEC, and medium-alone implantation groups. BM-MNCs contained 16% of endothelial-lineage cells and expressed basic fibroblast growth factorӷvascular endothelial growth factorϾangiopoietin 1 mRNAs, and their cardiac levels were significantly upregulated by BMI. Cardiac interleukin-1␤ and tumor necrosis factor-␣ mRNA expression were also induced by BMI but not by CMEC implantation. BM-MNCs were actively differentiated to endothelial cells in vitro and formed network structure with human umbilical vein endothelial cells. Conclusions-BMI may constitute a novel safety strategy for achieving optimal therapeutic angiogenesis by the natural ability of the BM cells to secrete potent angiogenic ligands and cytokines as well as to be incorporated into foci of neovascularization.

Angiotensin II (Ang II)-mediated signals are transmitted via heparin binding epidermal growth factor (EGF)-like growth factor (HB-EGF) release followed by transactivation of EGF receptor (EGFR). Although Ang II and HB-EGF induce angiogenesis, their link to the angiopoietin (Ang)-Tie2 system remains undefined. We tested the effects of Ang II on Ang1, Ang2, or Tie2 expression in cardiac microvascular endothelial cells expressing the Ang II receptors AT(1) and AT(2). Ang II significantly induced Ang2 mRNA accumulations without affecting Ang1 or Tie2 expression, which was inhibited by protein kinase C inhibitors and by intracellular Ca(2+) chelating agents. Ang II transactivated EGFR via AT(1), and inhibition of EGFR abolished the induction of Ang2. Ang II caused processing of pro-HB-EGF in a metalloproteinase-dependent manner to stimulate maturation and release of HB-EGF. Neutralizing anti-HB-EGF antibody blocked EGFR phosphorylation by Ang II. Ang II also upregulated vascular endothelial growth factor (VEGF) expression in an HB-EGF/EGFR-dependent manner. AT(2) inhibited AT(1)-mediated Ang2 expression and phosphorylation of EGFR. In an in vivo corneal assay, AT(1) induced angiogenesis in an HB-EGF-dependent manner and enhanced the angiogenic activity of VEGF. Although neither Ang2 nor Ang1 alone induced angiogenesis, soluble Tie2-Fc that binds to angiopoietins attenuated AT(1)-mediated angiogenesis. These findings suggested that (1) Ang II induces Ang2 and VEGF expression without affecting Ang1 or Tie2 and (2) AT(1) stimulates processing of pro-HB-EGF by metalloproteinases, and the released HB-EGF transactivates EGFR to induce angiogenesis via the combined effect of Ang2 and VEGF, whereas AT(2) attenuates them by blocking EGFR phosphorylation. Thus, Ang II is involved in the VEGF-Ang-Tie2 system via HB-EGF-mediated EGFR transactivation, and this link should be considerable in pathological conditions in which collateral blood flow is required.