Recombinant Human Erythropoietin Stimulates Angiogenesis and Wound Healing in the Genetically Diabetic Mouse Mariarosaria Galeano,1 Domenica Altavilla,2 Domenico Cucinotta,3 Giuseppina T. Russo,3 Margherita Calo`,4 Alessandra Bitto,2 Herbert Marini,2 Rolando Marini,2 Elena B. Adamo,2 Paolo Seminara,2 Letteria Minutoli,2 Valerio Torre,5 and Francesco Squadrito2 The effects of recombinant human erythropoietin (rHuEPO) in diabetes-related healing defects were in- vestigated by using an incisional skin-wound model he wound-healing process involves a complex produced on the back of female diabetic C57BL/KsJ- interplay of cells, mediators, growth factors, and m؉/؉Leptdb mice (dbϩ/dbϩ) and their normoglycemic cytokines (1). The cascade of events begins with littermates (dbϩ/ϩm). Animals were treated with Tclotting and recruitment of inflammatory cells rHuEPO (400 units/kg in 100 ␮l s.c.) or its vehicle alone and then proceeds to a highly proliferative state. During (100 ␮l). Mice were killed on different days (3, 6, and 12 this proliferative phase, fibroblasts are involved in synthe- days after skin injury) for measurement of vascular sis and remodeling of the collagen matrix, keratinocytes endothelial growth factor (VEGF) mRNA expression spread across the wound to form a new epithelial layer, and protein synthesis, for monitoring angiogenesis by and angiogenesis occurs. This latter step is crucial in the CD31 expression, and for evaluating histological healing process (2). During neovascularization, endothe- changes. Furthermore, we evaluated wound-breaking lial cells change their genetic program and express an strength at day 12. At day 6, rHuEPO injection in diabetic mice resulted in an increase in VEGF mRNA angiogenic phenotype that includes the production of relative amount of proteases, cell migration, and proliferation, followed by 0.1 ؎ 0.33 ؍ expression (vehicle relative amount of mRNA; dedifferentiation, thus resulting in the formation of new 0.09 ؎ 0.9 ؍ mRNA; rHuEPO blood vessels (3). The formation of new blood vessels 5 ؎ 23 ؍ P < 0.05) and protein wound content (vehicle pg/wound; P < 0.05) and provides a route for oxygen and nutrient delivery, as well 12 ؎ 92 ؍ pg/wound; rHuEPO caused a marked increase in CD31 gene expression as a conduit for components of the inflammatory re- relative amount of mRNA; sponse. Healing is concomitant with an increasing release 0.05 ؎ 0.18 ؍ vehicle) -relative amount of mRNA; P < of angiogenic growth factors from macrophages and ker 0.21 ؎ 0.98 ؍ rHuEPO 0.05) and protein synthesis. Furthermore, rHuEPO in- atinocytes, such as vascular endothelial growth factor jection improved the impaired wound healing and, at (VEGF), fibroblast growth factor, and platelet-derived day 12, increased the wound-breaking strength in dia- growth factor, and its impairment leads to a delay in skin ؎ ؎ ؍ betic mice (vehicle 12 2 g/mm; rHuEPO 21 5 g/mm; repair (4–7). P < 0.05). Erythropoietin may have a potential applica- Impaired wound healing may be a consequence of a tion in diabetes-related wound disorders. Diabetes 53: metabolic derangement such as diabetes (8). The diabetic 2509–2517, 2004 foot ulcer is a chronic wound that does not exhibit the orderly cascade of events that characterizes normal wound healing. More specifically, the chronic diabetic foot ulcer is stalled in the inflammation phase of the normal wound-healing process (9). Impaired wound healing in diabetic patients is not completely understood but is From the 1Department of Surgical Sciences, Section of Plastic Surgery, believed to include abnormalities in inflammation cells in University of Messina, Messina, Italy; the 2Department of Clinical and Exper- diabetic patients, impaired neovascularization, decreased imental Medicine and Pharmacology, Section of Pharmacology, University of synthesis of collagen, and increased levels of proteinases. Messina, Messina, Italy; the 3Department of Internal Medicine, University of Messina, Messina, Italy; the 4Department of Veterinary Public Health, Section Genetically diabetic mice are useful as an animal model of Veterinary Pharmacology and Toxicology, University of Messina, Messina, for wound-healing studies because wound healing in these Italy; and the 5Unit of Pathology, F. Veneziale Hospital, Isernia, Italy. animals is markedly delayed when compared with nondi- Address correspondence and reprint requests to Prof. Francesco Squadrito, MD, Department of Clinical and Experimental Medicine, Section of Pharma- abetic littermates (10). Healing impairment is character- cology, Azienda Ospedaliera Universitaria “G Martino,” Torre Biologica, 5th ized by delayed cellular infiltration and granulation tissue Floor, Via Consolare Valeria Gazzi, 98125 Messina, Italy. E-mail: francesco. [email protected]. formation, reduced angiogenesis, and decreased collagen Received for publication 28 November 2003 and accepted in revised form 3 synthesis and organization (11–14). One of the mecha- June 2004. nisms underlying the healing impairment in diabetic mice EPO, erythropoietin; rHuEPO, recombinant human EPO; TBS, Tris-buffered saline; VEGF, vascular endothelial growth factor. is thought to result from a defect in VEGF regulation at the © 2004 by the American Diabetes Association. gene expression level (15). DIABETES, VOL. 53, SEPTEMBER 2004 2509 ERYTHROPOIETIN IN DIABETIC WOUND HEALING TABLE 1 Criteria to evaluate histological scores of wound healing Score Epidermal and dermal regeneration Granulation tissue thickness Angiogenesis 1 Little epidermal and dermal Thin granulation layer Altered angiogenesis (one to two vessels per organization site) characterized by a high degree of edema, hemorrhage, occasional congestion, and thrombosis 2 Moderate epidermal and dermal Moderate granulation layer Few newly formed capillary vessels (three to organization four per site), moderate degree of edema and hemorrhage, occasional congestion and intravascular fibrin deposition, absence of thrombosis 3 Complete remodeling of epidermis and Thick granulation layer Newly formed capillary vessels (five to six dermis per site), moderate degree of perivascular and interstitial edema and congestion, absence of thrombosis and hemorrhage 4 — Very thick granulation layer Newly formed and well-structured capillary vessels (more than seven per site) vertically disposed toward the epithelium and at the wound margins, slight degree of perivascular edema Erythropoietin (EPO) is a hematopoietic factor regulat- RESEARCH DESIGN AND METHODS ing the proliferation and differentiation of erythroid pre- All animal procedures were in accordance with the Declaration of Helsinki cursor cells (16). The biological effects of EPO are and with the Guide for the Care and Use of Laboratory Animals. ϩ/ϩ db ϩ ϩ mediated by its specific interaction with its cell surface Genetically diabetic female C57BL/KsJ-m Lept mice (db /db ) and their normal littermates (dbϩ/ϩm) were obtained from The Jackson Labora- receptor EPOR, a type 1 cytokine receptor that is present tory (Bar Harbor, ME). Animals were 14 weeks old at the start of the in erythroid progenitor cells as well as in several nonhe- experiments. Diabetic mice were obese, weighing 40–50 g, whereas nondia- matopoietic cell types. betic littermates weighed 25–32 g. During the experiments, the animals were Recombinant human EPO (rHuEPO) is therapeutically housed one per cage, maintained under controlled environmental conditions applied for stimulation of the erythroid lineage. Experi- (12-h light/dark cycle, temperature ϳ23°C), and provided with standard laboratory food and water ad libitum. mental evidence has demonstrated that this hormone can The animals were divided into four groups of 21 animals each. The first and stimulate mitosis and induce differentiation and activation second groups, consisting of normoglycemic mice, were given, respectively, of numerous cell lines, such as those of endothelium, rHuEPO at a dose of 400 IU/kg s.c. in 100 ␮l or rHuEPO vehicle (100 ␮lofa myocardium, and smooth muscle (17,18). The presence of NaCl 0.9% saline solution) for 12 days. The third and the fourth groups, EPO receptors on endothelial cells prompted several in consisting of diabetic mice, were treated, respectively, with rHuEPO at a dose ␮ ␮ vitro and in vivo investigations aimed to determine of 400 IU/kg s.c. in 100 l or with rHuEPO vehicle (100 l of a NaCl 0.9% saline solution) for 12 days. After general anesthesia with ketamine hydrochloride whether the hormone may directly affect some of the (110 mg/kg), the hair on the back was shaved and the skin washed with functions of these cells (19), and, more specifically, atten- povidone-iodine solution and wiped with sterile water. Two full-thickness tion has been paid to the complex network of cytokines longitudinal incisions (4 cm) were made on the dorsum of the mice, and the and growth factors involved in both the maturation of wound edges were closed with skin clips placed at 1-cm intervals. Seven erythrocytes and the proliferation of endothelial cells (20). animals for each group were killed after 3, 6, and 12 days, respectively, and the wounds were divided into three segments (0.8 cm wide and 1.2 cm long). The The hypothesis that hematopoietic and endothelial cells caudal and cranial strip was used for histology, whereas the central one was share a common hemangioblast progenitor is based on the divided into two segments. The first one was used for molecular analysis and finding (3) that both cell lineages express cell surface the second one for wound-breaking strength measurements