For gene transfer involved in therapeutic angiogenesis, Reversine several delivery methods are used. Administration of purified recombinant proteins achieved some positive results, but short half-life of proteins and long-term toxicity with repeated injections may cause system toxicity. Infection with adenoviruses with the target gene is effective in achieving stable and long-term expression, but may induce adverse immune reactions. Davis et al reported that injection of naked plasmid DNA appeared to be better for overexpressing a target gene in skeletal muscles than viral vectors because skeletal muscle cells were able to take plasmid DNA up. Thus, direct injection of plasmid DNA into skeletal muscle tissues has become an alternative to achieve a high level of gene transfer to avoid significant disadvantages with injection of a protein and an adenoviral vector. The data in the present study again showed that direct injection of DNA for an AGGF1 expression plasmid successfully overexpressed AGGF1 in the ischemic tissues. Two recent clinical trials, HGF-STAT and TALISMAN201 utilized a plasmid based angiogenic gene delivery system for HGF and FGF-1, respectively, and achieved some encouraging results. More clinical trials are needed to replicate these findings. Similarly, future clinical studies for therapeutic angiogenesis using AGGF1 with a plasmid-based gene delivery system are needed to unequivocally establish the efficacy of AGGF1 treatment for PAD. Tsurumi et al showed that administration of naked plasmid DNA encoding VEGF increased regional blood flow to the transfected thigh muscle and distal lower limb muscle by 1.5- fold in a rabbit ischemic hindlimb model. Hiraoka et al. reported that in a rat model of hindlimb ischemia, injection of VEGF plasmid DNA increased blood flow by about 30%. Taniyama et al. showed that injection of naked HGF plasmid DNA into skeletal muscle resulted in a 70% increase in blood flow in a rat model of hindlimb ischemia three weeks after the injection. We demonstrated that injection of naked plasmid DNA for an AGGF1 expression construct resulted in a 2.29-fold increase in blood flow. Furthermore, our parallel comparison analysis found that AGGF1 was significantly better than FGF-2 in stimulating blood flow 28 days after gene transfer, although no significant difference was found for day 7 and day 14. Therefore, AGGF1 appears to be an excellent choice for therapeutic angiogenesis for critical limb ischemia. Some side effects were uncovered in previous studies involving therapeutic angiogensis for treating limb ischemia in PAD using VEGF and FGFs. The major adverse effects include increased vascular permeability and transient edema. In contrast to VEGF, AGGF1 is required for maintaining the vascular integrity because adult heterozygous AGGF1+/2 knockout mice showed increased vascular permeability in an assay using Evan’s blue dye. During the AGGF1 treatment in a hindlimb ischemic mouse model for PAD, we did not observe any edema. However, future studies are needed to determine whether a larger dose of AGGF1 DNA injection may result in side effects of edema or other undesirable abnormalities in major organs such as the heart, livers, kidneys, lungs and other organs. A long-standing question about the mode of action of AGGF1 during angiogenesis is whether it acts by an autocrine or paracrine mechanism. Because the AGGF1 protein is expressed and secreted by endothelial cells, we suggested that AGGF1 may act by an autocrine mode.