Human Cord Blood-Derived Endothelial Progenitor ... - SAGE Journals

0963-6897/10 $90.00 + .00 DOI: 10.3727/096368910X516637 E-ISSN 1555-3892 www.cognizantcommunication.com

Cell Transplantation, Vol. 19, pp. 1635–1644, 2010 Printed in the USA. All rights reserved. Copyright  2010 Cognizant Comm. Corp.

Human Cord Blood-Derived Endothelial Progenitor Cells and Their Conditioned Media Exhibit Therapeutic Equivalence for Diabetic Wound Healing Ji Yeon Kim,*† Sun-Hwa Song,* Koung Li Kim,*§ Jeong-Jae Ko,* Ji-Eun Im,* Se Won Yie,† Young Keun Ahn,‡ Duk-Kyung Kim,§ and Wonhee Suh* *Department of Biomedical Science, College of Life Science, CHA University, Gyeonggi-do, Korea †Department of Molecular Bioscience, Division of Bioscience and Biotechnology, Kangwon National University, Chuncheon, Korea ‡Department of Cardiology, Chonnam National University Medical Center, Gwangju, Korea §Department of Medicine, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul Korea

Transplantation of human cord blood-derived endothelial progenitor cells (EPCs) is reported to contribute to neovascularization in various ischemic diseases. However, the possible beneficial role and underlying mechanisms in diabetes-impaired wound healing have been less well characterized. In this study, EPC transplantation stimulated keratinocyte and fibroblast proliferation substantially as early as 3 days after injury, leading to significantly accelerated wound closure in streptozotocin-induced diabetic nude mice, compared to PBS control. RT-PCR analysis showed that EPCs secreted various wound healing-related growth factors. Among them, keratinocyte growth factor and platelet-derived growth factor were highly expressed in the EPCs and were present at substantial levels in the EPC-injected dermal tissue. Using EPC-conditioned medium (CM), we found that paracrine factors from EPCs directly exerted mitogenic and chemotactic effects on keratinocytes and fibroblasts. Moreover, injection of EPC-CM alone into the same diabetic wound mice promoted wound healing and increased neovascularization to a similar extent as achieved with EPC transplantation. These results indicate that the beneficial effect of EPC transplantation on diabetic wounds was mainly achieved by their direct paracrine action on keratinocytes, fibroblasts, and endothelial cells, rather than through their physical engraftment into host tissues (vasculogenesis). In addition, EPC-CM was shown to be therapeutically equivalent to EPCs, at least for the treatment of diabetic dermal wounds, suggesting that conditioned medium may serve as a novel therapeutic option that is free from allograft-associated immune rejection concern. Key words: Endothelial progenitor cells (EPCs); Diabetes mellitus; Paracrine signaling; Wound healing; Animal model

INTRODUCTION

associated with decreased peripheral blood flow and which remain difficult to heal using current therapeutic approaches. Furthermore, EPCs stimulate endogenous angiogenesis by secreting a variety of angiogenic growth factors, which implies that paracrine factors released from EPCs could directly stimulate keratinocyte and fibroblast proliferation during the wound healing process (7,17). The wound healing repair process is a well-orchestrated integration of the multiple biological and molecular events that occur during inflammation, cell migration/proliferation, and extracellular matrix deposition (19). Repair occurs through the coordinated action of many growth factors and cytokines and through the in-

Endothelial progenitor cells (EPCs) are endothelial precursors involved in the revascularization of injured tissue and tissue repair (1). The therapeutic potential of these EPCs has been reported in animal models and in humans with ischemic diseases, including myocardial infarction, stroke, and peripheral arterial diseases (6, 16,18). We previously demonstrated that the local injection of EPCs into the excisional dermal wounds of normal mice significantly accelerated wound healing and promoted neovascularization of granulation tissues (20). Thus, EPC transplantation could be beneficial for the treatment of wounds that resist healing, which are often

Received October 9, 2009; final acceptance June 14, 2010. Online prepub date: July 15, 2010. Address correspondence to Wonhee Suh, Ph.D., Department of Biomedical Science, College of Life Science, CHA University, CHA Stem Cell Institute, 606-16 Yeoksam1-dong, Kangnam-gu, Seoul, 135-907, Korea. Tel: 82-2-3468-3668; Fax: 82-2-538-4102; E-mail: [email protected]

1635

1636

terplay of different cell types, such as inflammatory cells, fibroblasts, keratinocytes, and endothelial cells (25). Normal wounds without underlying pathophysiological defects heal readily via the orderly progression of the healing process. In contrast, wounds in patients with preexisting pathophysiological abnormalities such as diabetes do not. The wound-healing deficiency of diabetic wounds can be attributed to a number of factors, such as decreased production of growth factors, reduced angiogenesis, and impaired migration and proliferation of keratinocytes and fibroblasts. In addition, a recent study in diabetic mice demonstrated that a reduced number of EPCs in the bloodstream and in wound tissues is another crucial cause of impaired diabetic wound healing (5). Indeed, the peripheral blood of diabetic patients has significantly fewer EPCs than that of normal subjects (4,11). Moreover, EPCs from diabetic mice impaired nevoascularization when injected into an ischemic disease animal model, whereas EPCs from normal mice significantly improved blood flow in injured tissues (21). Based on the above data, a sufficient supply of EPCs isolated from a healthy source such as human cord blood was hypothesized to restore the impaired wound healing process in diabetic wounds. The present study investigated the wound healing role of human cord bloodderived EPCs in streptozotocin (STZ)-induced diabetic mice. Furthermore, this study addressed the major mechanism involved in the EPC-mediated wound healing process by comparing the EPCs-transplanted and EPC-CMinjected wound. MATERIALS AND METHODS Cell Culture EPCs were isolated as described previously by culturing CD34-positive cells from human umbilical cord bloods in the endothelial growth medium (EGM)-2MV Singlequot (Clonetics, Walkersville, MD, USA) for 2–3 weeks (20). These cells were found to express CD31, KDR, and Tie2, and also to exhibit other properties consistent with the existing definitions of an EPC phenotype, such as uptake of acetylated LDL and tube formation on Matrigel (9). The Institutional Review Board at CHA University approved all protocols, and informed consent was obtained from all donors. For experiments performed with conditioned medium (CM) from EPCs (EPC-CM), the CM was prepared by culturing EPCs in serum-free M199 media (Clonetics) at 37°C for 24 h and was cleared out and concentrated by centrifugation using Microcon (MW cutoff = 10K, Amicon Division, Danvers, MA, USA). Pam212 (murine keratinocyte, gift by Kim TY, Catholic University of Korea, Korea), NIH3T3 (murine fibroblast, ATCC, Manassas, VA, USA), and human dermal fibroblast

KIM ET AL.

(HDF, gifted by Kim DS, Chung-Ang University, Korea) cells were cultured in DMEM (Invitrogen, Carlsbad, CA, USA) plus 10% fetal bovine serum. Hek001 (human keratinocyte, gifted by Kim DS, Chung-Ang University, Korea) cells were cultured in keratinocyte serum-free media (Invitrogen) with 5 ng/ml of human recombinant epidermal growth factor (EGF, Invitrogen) and 2 mM of L-glutamine. Diabetic Dermal Wound Animal Model To induce diabetes in immunodeficient mice, a single high dose of STZ (Sigma, St. Louis, MO, USA; 225 mg/ kg) was intraperitoneally injected into Balb/c nude mice (previously fasted for 16 h, 7–8 weeks old, male, body weight 17–23 g, Japan SLC Inc., Shizuoka, Japan) (26). In this study, a low-dose STZ protocol failed to induce hyperglycemia in Balb/c nude mice, as previously reported (15). Every week after STZ administration, serum glucose levels were measured using an Accu-Check Advantage glucometer (Roche, Indianopolis, IN, USA) under the nonfasting status. Mice with a plasma glucose level of >300 mg/dl (in normal Balb/c nude mice, glucose level was around 100 mg/dl) at 3 weeks after injection were regarded as having STZ-induced diabetes (26). To investigate the effect of EPCs on diabetic wound healing, STZ-induced diabetic mice were anesthetized with an intraperitoneal injection of ketamine-xylazine (79.5 mg/kg and 9.1 mg/kg, respectively), after which two full-thickness excisional wounds (upper and lower wounds were separated by 2.5 cm) were created on the dorso-lateral area using a standard skin biopsy punch (0.5 cm in diameter, Acuderm Inc., Fort Lauderdale, FL, USA). Immediately after surgery, EPCs (1 × 106 cells suspended in 60 µl of PBS) and 60 µl of PBS were injected at three different intact dermis sites near the upper and lower (or lower and upper wound to rule out the influence of wound location) wounds created on the same mice to rule out the interanimal variance caused by the differing degrees of hyperglycemia among the STZ-induced mice. To examine the paracrine effect of EPCs, 100 µl of M199 basal medium or concentrated EPC-CM (collected from 1 × 106 EPCs) were injected into wounds with the same experimental method described above. All procedures were performed with the approval of CHA University Institutional Animal Care and Use Committee (IACUC). Wound Analysis The open wound area was measured on days 0, 3, 5, 7, 9, and 12 after treatment using a method described previously (20). Briefly, wounds were documented with a digital camera and their images were analyzed using Image-Pro Plus (Media Cybernetics, Inc., Silver Spring, MD, USA) by tracing the wound margin and calculating

EPCs AND DIABETIC WOUND HEALING

the pixel area. Pixel counts were then normalized to the scale of the metric ruler that was photographed with each wound. Measurement was performed in duplicate, and the open wound area was expressed as a percentage of the wounded area at each time point divided by the area of its original wound at day 0. To measure the granulation tissue area, wound tissues were harvested at days 15 and 20, fixed in 4% paraformaldehyde (Sigma), and embedded in paraffin. Wound sections from the least healed central region (maximal distance between the epithelial edges; considered to be a central region of the wounds) were stained with hematoxylin and eosin (H&E). Then, the granulation tissue area was analyzed with Image-Pro Plus by tracing the wound boundary that was considered as the presence of intact hair follicles and the transition from normal to hypertrophic epidermis, and calculating the pixel area (20). Ten to 12 sections were examined for each wound. Immunohistochemistry Wound sections were treated with 0.3% hydrogen peroxide to quench the endogenous peroxidase and then antigen epitopes were retrieved by heating in Target Retrieval Solution (DakoCytomation, Carpinteria, CA, USA). After sections were blocked in 10% normal goat serum, they were treated with anti-von Willebrand factor (vWF; DakoCytomation), anti-cytokeratin14 (CK14; Abcam, Cambridge, MA, USA), anti-vimentin (DakoCytomation), anti-Ki67 (BD Bioscienece, San Jose, CA, USA), anti-human CD31 (Abcam), anti-mouse CD31 (BD Bioscience), or irrelevant nonspecific IgG. Sections were then incubated with biotinylated or fluorescenceconjugated secondary IgGs (Jackson ImmunoResearch Laboratories Inc., West Grove, PA, USA) and were visualized using an ABC kit (Elite kit, Vector Laboratories, Burlingame, CA, USA) and Vector Blue (Vector Laboratories). For multicolor immunofluorescence analysis, sections were further stained with 4,6-diamidino-2-phenylindole (DAPI) and imaged with a fluorescence microscope (Carl Zeiss, Oberkochen, Germany). Reverse Transcriptase Polymerase Chain Reaction (RT-PCR) Analysis Total RNA (1 µg) was extracted from EPCs or wound tissues harvested at days 3 and 7 after surgery and was processed for cDNA synthesis using the Superscript first-strand synthesis system (Invitrogen). The cDNA was amplified with 30 to 35 cycles of PCR using gene-specific primers as follows: hVEGF, 5′-CAAGGC CAGCACATAGGAGA-3′ and 5′-AGGGAACGCTC CAGGACTTA-3′; hPDGF-α, 5′-CCATTCGGAGGAA GAGAAGC-3′ and 5′-GTATTCCACCTTGGCCACCT3′; hPDGF-β, 5′-TCGAGATTGTGCGGAAGAAG-3′ and 5′-GTGTGCTTGAATTTCCGGTG-3′; hTGF-β1,

1637

5′-GGGACTATCCACCTGCAAGA-3′ and 5′-CGGAG CTCTGATGTGTTGAA-3′; hTGF-β2, 5′-AAAATAG ACATGCCGCCCTT-3′ and 5′-CTTTTGGGTTCTGCA AACGA-3′; hbFGF, 5′-AAGCGGCTCTACTGCAAAA A-3′ and 5′- CTTTCTGCCCAGGTCCTGTT-3′; hKGF, 5′-TTCACATTATCTGTCTAGTGGGT-3′ and 5′-TG GGTCCCTTTTACTTTGCC-3′; hGAPDH, 5′-GAAGG TGAAGGTCGGAGTC-3′ and 5′-GAAGATGGTGAT GGGATTTC-3′; mGAPDH, 5′-ATGACTCCACTCAC GGCAAA-3′ and 5′-ATGATGACCCTTTTGGCTCC-3′. Real-time PCR was performed with a SYBR-Green PCR master mix (Qiagen, Valencia, CA, USA) using Bioneer ExicyclerTM 96 (Bioneer Corporation, Daejeon, Korea). Data analysis was based on the ∆∆Ct method with normalization of the raw data to housekeeping gene, GAPDH, included in the experiment. All reactions were performed in triplicate. Cell Proliferation and Migration Assay For the cell proliferation analysis, Pam212 and Hek001 cells were incubated in EPC-CM, in M199 basal medium with or without human keratinocyte growth factor (KGF; 50 ng/ml, ProSpec-Tany Technogene, Rohovot, Israel) or human platelet-derived growth factor (PDGF-BB; 10 ng/ml, ProSpec-Tany Technogene) at 37°C for 48 h. The cell number was then estimated by measuring the absorbance at 540 nm after using a methylthiazolyldiphenyl-tetrazolium bromide (MTT, Sigma) assay according to the manufacturer’s instructions. A migration assay was performed using a modified Boyden chamber (Costar, Cambridge, MA, USA), where cells (Pam212, Hek001, NIH3T3, HDF) were placed in the upper chamber and the lower chamber was filled with EPC-CM, M199 basal medium, or M199 supplemented with KGF (50 ng/ml) or PDGF-BB (10 ng/ml). After incubation for 24 h, migrated cells that attached to the lower side of the filter were stained with Giemsa solution and counted. Statistical Analysis All data are presented as mean ± SEM. Statistical significance was evaluated using paired t-test or one-way analysis of variance followed by Bonferroni’s post hoc multiple comparison test. A value of p < 0.05 was considered statistically significant and the number of samples examined is indicated by n. RESULTS EPC Transplantation Accelerates Wound Closure in STZ-Induced Diabetic Mice To evaluate the therapeutic effect of xenotransplantation of human cord blood-derived EPCs to heal STZinduced diabetic wounds in athymic nude mice, diabetes was induced in athymic nude mice by injecting STZ. As

1638

KIM ET AL.

expected from previous studies, the STZ-induced diabetic mice displayed the markedly delayed wound healing (13). These mice achieved the 50% wound closure by day 12, whereas normal athymic nude mice exhibited 50% closure by day 7 after wounding (data not shown). When EPCs were injected around full-thickness dermal wounds created on the diabetic nude mice, wound closure was significantly accelerated as early as day 3 after injury compared to PBS (vehicle control)-treated wounds (Fig. 1A, B). This significant reduction in the open wound area was consistently observed until day 12 (EPC, 10.6 ± 5.3%; PBS, 45.7 ± 9.2% on day 12, p < 0.05). In addition, the EPCs-treated wounds had a substantially reduced cross-sectional area of granulation tissues at day 20 compared with PBS-treated controls (Fig. 1C). EPC Transplantation Promotes the Proliferation of Keratinocytes and Fibroblasts in the Early Wound Healing Process To carefully examine the cellular changes that occur with EPC transplantation during the early wound healing

process, wound sections retrieved at days 3 and 7 after injury were immunohistologically analyzed, as shown in Figure 2. From the earliest time points studied (day 3), a number of proliferating keratinocytes (positive for both CK14 and the proliferation marker, Ki67) were found in EPCs-treated wounds, especially in the basal layer adjacent to the thickened wound edge and in the distal epidermis, in contrast with a few proliferating keratinocytes in PBS-treated controls (Fig. 2A). Notably, CK14-positive proliferating cells were also found at the outer root sheath of hair follicles in the neighboring intact skin, area known as keratinocyte reservoirs. Because follicular keratinocytes, whose growth is arrested in normal skin, undergo mitosis and provide progeny to the epidermis after wounding, they have been thought to play a crucial role in initiating rapid reepithelialization (10). CK14/Ki67-positive cells within hair follicles were more frequently observed in EPCs-treated wounds than in PBS-treated controls. In addition, the immunohistochemical analysis of granulation tissue at day 7 after injury revealed that EPC transplantation promoted fibroblast proliferation in the granulation tissue, in that EPCs-

Figure 1. EPC transplantation accelerates excisional wound healing in STZ-induced diabetic mice. (A) Representative images of EPC-injected (suspended in PBS) and PBS-injected (vehicle control) wounds in STZ-induced diabetic Balb/c nude mice at indicated time points after wounding. Scale bar: 2 mm. (B, C) The open wound area expressed as a percentage of its initial wound area (B) and the relative granulation tissue area at day 20 (C) were determined in wounds treated with EPCs or PBS (white bar, EPC; black bar, PBS in both B and C). *p < 0.05, the number of animals (n) = 6. Data are the mean ± SEM.


1639

Figure 2. EPC transplantation promotes the in vivo proliferation of keratinocytes and fibroblasts during the early wound healing process. (A) Representative immunofluorescence images of epidermal tissues from PBS- or EPCs-treated wounds at day 3 after wounding. Proliferating keratinocytes were identified by co-immunofluorescence staining with antibodies against Ki67 (red, nucleus) and cytokeratin 14 (CK14, green, cytoplasm). Nuclei were stained with DAPI (blue). Arrows indicate the epidermal wound edges; e, epithelium; d, dermis; h, hair follicle; and es, eschar. The lower images show a high magnification view of the hyperthickened epithelium at the wound edges, distal epithelium, and hair follicles shown in the upper images. Scale bar: 100 µm. (B) Representative immunofluorescence images of granulation tissues from PBS- or EPCs-treated wounds at day 7. Proliferating fibroblasts were identified by co-immunofluorescence staining with antibodies against Ki67 (red, nucleus) and vimentin (green, cytoplasm). Nuclei were stained with DAPI (blue). Scale bar: 100 µm.

1640

treated wounds exhibited more proliferating fibroblasts (positive for both vimentin and Ki67) than PBS-treated controls (Fig. 2B). The Conditioned Medium From EPCs Exerts Mitogenic and Chemotactic Effects on Keratinocytes and Fibroblasts To examine whether secreted factors from injected EPCs might directly stimulate keratinocyte and fibroblast proliferation, growth factors and cytokines that are potentially relevant in dermal wound healing were analyzed using real-time PCR. The EPCs expressed a wide array of growth factors and cytokines, of which PDGFα, PDGF-β, and KGF were the most strongly expressed (Fig. 3A). Consistently, RT-PCR analysis with wound tissues using human-specific primers revealed that abundant mRNAs of human PDGF-α, PDGF-β, and KGF were detected only in EPCs-treated wound tissues harvested at day 3 after injury (Fig. 3B). Furthermore, these paracrine factors released from human EPCs exhibited mitogenic and chemoattractant properties on keratinocytes and fibroblasts. As shown in Figure 3C and D, EPC-CM significantly enhanced the proliferation of both murine keratinocytes (Pam212) and fibroblasts (NIH3T3) to a comparable extent as their growth with human recombinant PDGF-BB, KGF, or bFGF proteins. Similar mitogenic activity of EPC-CM was observed in the same experiment with human keratinocytes (Hek001) and fibroblasts (HDF) (Fig. 3C, D). In addition, the effect of EPC-CM on the chemotactic migration of fibroblasts and keratinocytes was examined using transwell migration assay. EPC-CM induced significant seven and twofold increases in the migration of murine and human fibroblasts respectively (Fig. 3E), although it had no significant effect on keratinocyte migration (data not shown). EPC-CM Is Therapeutically Equivalent to EPCs in its Ability to Enhance Wound Closure and Neovascularization in STZ-Induced Diabetic Mice We next asked whether paracrine factors from injected EPCs made a contribution to accelerating the wound healing process in diabetic mice and evaluated the extent of their contribution to the enhanced skin regeneration observed in EPCs-treated wounds. Concentrated EPC-CM (prepared from an overnight culture of 1 × 106 EPCs) was intradermally injected into the same diabetic wound murine model and its effect on wound healing was analyzed and compared with that of cell injection (1 × 106 EPCs). Macroscopic observation at indicated time points in Figure 4 showed that wounds treated with EPC-CM had substantially smaller open wound area than M199 basal medium controls from the earliest time point studied. Notably, the paralleled comparison of EPC-CM-treated and EPC-injected groups re-

KIM ET AL.

vealed that the intradermal application of EPC-CM accelerated the wound closure rate to the same extent as EPC injection for the entire duration of study (Fig. 4B). The EPC-CM-treated wounds had significantly smaller granulation tissue areas than those in the M199 controls (Fig. 4C). Furthermore, the extent of this reduction in granulation tissue area was similar to that seen in EPCstransplanted wounds. These data indicate that EPC-CM was therapeutically equivalent to EPCs in accelerating wound closure and reducing the granulation tissue area in STZ-induced diabetic mice. To investigate whether the injection of EPC-CM might increase the neovascularization of wounded tissue as observed in EPCs-injected tissues, the capillary density was estimated by counting the vWF-positive neovessels in the granulation tissues of wound sections harvested at day 15. Representative images and quantitative analysis displayed in Figure 4D and E showed that the capillary density in EPC-CM-treated wounds was substantially greater than that in M199-treated controls. Although slightly less in magnitude, this enhancement was not significantly different from that observed in EPCstreated wounds. Immunohistological analysis with species-specific CD31 antibodies revealed that a few neovessels in the EPCs-treated wounds were lined with both mouse and human CD31-positive cells (Fig. 4F). This result indicates that injected human EPCs contributed to the neovascularization of wounded tissues, in part through physically incorporating into the newly formed murine vasculature (vasculogenesis). However, EPCmediated postnatal vasculogenesis might not be a supportive mechanism to enhance neovascularization after EPC transplantation into diabetic wounds. DISCUSSION The present study revealed that the transplantation of human cord blood-derived EPCs significantly accelerated wound closure in diabetic mice from the earliest time point studied, with enhanced keratinocyte/fibroblast proliferation and neovascularization. In particular, EPC transplantation appeared to stimulate the proliferation of CK14-positive keratinocytes in the basal layer of the epidermis and the outer root sheath of hair follicles near the wound. This proliferation might have contributed to the rapid reepithelialization at early time point by supporting keratinocyte migration into the center of the wound bed where very few Ki67-positive proliferating cells were found. Because the activation of hair follicular keratinocytes is necessary to initiate rapid reepithelialization right after wounding, their progenies constitute a significant portion of the new epidermis formed in wounded skin (10). In this regard, the abundant proliferating follicular keratinocytes observed in the EPC-treated mice implied that the enhanced reepi-


1641

Figure 3. Paracrine factors secreted from EPCs exert mitogenic and chemotactic effects on keratinocytes and fibroblasts. (A) The mRNA profile of wound healing-related growth factors and cytokines expressed in EPCs. The mRNA level in EPCs was analyzed by real-time PCR and normalized as its relative ratio to human GAPDH. The data were obtained with EPCs from three independent preparations and are expressed as the mean ± SEM. (B) The mRNA expression of human growth factors in PBS- or EPCs-injected wounds. RT-PCR analysis with human-specific primers was performed on wound tissues harvested at day 7 and unwounded normal tissues (Non). (C, D) EPC-CM stimulates the proliferation of keratinocytes (C) and fibroblasts (D) (Pam212, murine keratinocyte; Hek001, human keratinocyte; NIH3T3, murine fibroblast; HDF, human dermal fibroblast). Cell proliferation was analyzed using an MTT assay after an overnight incubation with basal medium (M199), EPC-CM, or M199 supplemented with bFGF (10 ng/ml), KGF (50 ng/ml), or PDGF-BB (10 ng/ml). *p < 0.01 versus M199, n = 5. (E) EPC-CM enhances the chemotactic migration of fibroblasts (murine, NIH3T3; human, HDF). Boyden chamber migration assay was performed with M199, EPC-CM, or M119 supplemented with bFGF (10 ng/ml) or PDGF-BB (10 ng/ml). *p < 0.01 versus M199, n = 5.

1642

KIM ET AL.


1643

FACING PAGE Figure 4. EPC-CM is therapeutically equivalent to EPCs in enhancing wound closure and neovascularization in STZ-induced diabetic mice. (A) Representative images of EPC-CM- or M199 basal medium-injected wounds in STZ-induced diabetic Balb/c nude mice. Scale bar: 2 mm. (B) The open wound areas from four experimental groups were quantified at each time point as described before. (Open circle, EPC; filled circle, PBS; open square, EPC-CM; filled square, M199). *p < 0.05 versus M199 or PBS, the number of animals (n) = 7. Data are the mean ± SEM. (C) The granulation tissue area was determined at day 15 after wounding. *p < 0.05 versus M199 or PBS, n = 7. NS, nonsignificant. Data are the mean ± SEM. (D) Representative vWF-stained images of the granulation tissue from M199-, EPC-CM-, PBS-, or EPCs-treated mice. Excisional wound tissues were harvested at day 15 after injury and the murine vasculature in the granulation tissues was immuohistochemically stained with anti-vWF (arrows) or nonspecific IgG. Scale bar: 100 µm. (E) Capillary density in the granulation tissue was determined as the number of vWFpositive vessels per high power field (hpf). *p < 0.05 versus M199 or PBS, n = 5. Data are the mean ± SEM. (F) Injected human EPCs were directly incorporated into newly formed capillaries in granulation tissues. Immunohistochemistry was performed on EPCs- or PBS-injected wounds with human-specific (red) and mouse-specific (green) CD31 antibodies to differentiate the species origin of capillary endothelial cells. Scale bar: 50 µm.

thelialization made a great contribution in accelerating wound closure rate. The finding that the EPC-treated wounds exhibited a number of Ki67-positive cells and rapid wound closure at very early wound healing process, suggested that the EPCs might directly activate the proliferation of keratinocytes and fibroblasts. Indeed, the experiment with EPC-CM revealed that paracrine factors secreted from EPCs directly activated the proliferation of both keratinocytes and fibroblasts, and also promoted the chemotactic migration of fibroblasts. In the normal wound recovery process, growth factors and cytokines are massively produced at the wounded skin sites and contribute to wound healing process, whereas their expressions in diabetic human and animal models are significantly reduced and delayed, thereby disrupting the normal healing process (2,3,24). In particular, impaired wound healing in diabetes is associated with decreased expression levels of PDGF and KGF, indicating that certain expression levels are essential for normal wound repair (12,22). In this regard, PDGF and KGF released from injected human EPCs might stimulate the keratinocytes and fibroblasts to proliferate and migrate. Or a wellorchestrated combination of many paracrine factors secreted from EPCs might more effectively accelerate the wound healing process rather than one or two growth factors. Such a paracrine action of EPCs on keratinocytes and fibroblasts has not yet been studied, although several previous studies have investigated the paracrine effect of EPCs on endothelial cells (7,17). The aforementioned paracrine and vasculogenic actions of EPCs are well-characterized as mechanisms involved in EPC-mediated neovascularization process. However, it remains unclear which mechanism plays the dominant role in EPC-mediated tissue regeneration. Although numerous studies have documented the presence of injected EPCs at newly formed capillaries in injured tissues, the actual magnitude of EPC incorporation into the vasculature substantially varies among such studies.

Some studies have reported that >50% of the newly formed vasculature contains injected EPCs, whereas other reports have shown only occasional EPC-incorporated capillaries, despite noting a substantial increase in capillary density and improved tissue regeneration (23). In this regard, our comparison of EPC-transplanted and EPC-CM-injected wounds clearly addressed which mechanism was responsible for the observed EPC-mediated wound healing effect. The present data suggest that paracrine factors released from EPCs are sufficient for efficient wound healing in diabetic mice and EPC-mediated vasculogenesis might not be responsible for the improved healing speed and enhanced neovascularization at wound sites that were observed in EPC transplantation. Although some of injected human EPCs were physically engrafted into the murine vasculature, the EPCs appeared to increase wound revascularization mainly by stimulating angiogenesis rather than through postnatal vasculogenesis. The neovascularization effect of human EPCs in diabetic mice was recently reported, wherein EPC transplantation ameliorated murine diabetic neuropathy by enhancing vascularization around the peripheral nerves (8,14). However, these studies did not directly address the effect of EPC transplantation on diabetic wound healing. The present study is the first to our knowledge to definitely characterize the therapeutic potential of human cord blood-derived EPCs in diabetic wound healing. Importantly, the present data demonstrate that EPCCM is therapeutically equivalent to EPCs, at least in terms of wound closure rate, granulation tissue formation, and neovascularization. Although further study is required to define its composition, EPC-CM could serve as a potential therapeutic alternative without adverse immunological reaction to allografts that should be carefully considered when injecting any nonautologous cells. ACKNOWLEDGMENTS: This work was supported by Basic Science Research Program and Priority Centers Program through the National Research Foundation (NRF) funded by

1644

KIM ET AL.

the Ministry of Education, Science and Technology (20090093821) (2009-0084090) and a grant of the Korea Healthcare technology R&D Project, Ministry of Health, Welfare & Family Affairs, Republic of Korea (A084072).

REFERENCES 1. Asahara, T.; Murohara, T.; Sullivan, A.; Silver, M.; van der Zee, R.; Li, T.; Witzenbichler, B.; Schatteman, G.; Isner, J. M. Isolation of putative progenitor endothelial cells for angiogenesis. Science 275:964–967; 1997. 2. Castronuovo, Jr., J. J.; Ghobrial, I.; Giusti, A. M.; Rudolph, S.; Smiell, J. M. Effects of chronic wound fluid on the structure and biological activity of becaplermin (rhPDGF-BB) and becaplermin gel. Am. J. Surg. 176: 61S–67S; 1998. 3. Doxey, D. L.; Ng, M. C.; Dill, R. E.; Iacopino, A. M. Platelet-derived growth factor levels in wounds of diabetic rats. Life Sci. 57:1111–1123; 1995. 4. Fadini, G. P.; Sartore, S.; Albiero, M.; Baesso, I.; Murphy, E.; Menegolo, M.; Grego, F.; Vigili de Kreutzenberg, S.; Tiengo, A.; Agostini, C.; Avogaro, A. Number and function of endothelial progenitor cells as a marker of severity for diabetic vasculopathy. Arterioscler. Thromb. Vasc. Biol. 26:2140–2146; 2006. 5. Gallagher, K. A.; Liu, Z. J.; Xiao, M.; Chen, H.; Goldstein, L. J.; Buerk, D. G.; Nedeau, A.; Thom, S. R.; Velazquez, O. C. Diabetic impairments in NO-mediated endothelial progenitor cell mobilization and homing are reversed by hyperoxia and SDF-1 alpha. J. Clin. Invest. 117:1249–1259; 2007. 6. Hung, H. S.; Shyu, W. C.; Tsai, C. H.; Hsu, S. H.; Lin, S. Z. Transplantation of endothelial progenitor cells as therapeutics for cardiovascular diseases. Cell Transplant. 18:1003–1012; 2009. 7. Hur, J.; Yoon, C. H.; Kim, H. S.; Choi, J. H.; Kang, H. J.; Hwang, K. K.; Oh, B. H.; Lee, M. M.; Park, Y. B. Characterization of two types of endothelial progenitor cells and their different contributions to neovasculogenesis. Arterioscler. Thromb. Vasc. Biol. 24:288–293; 2004. 8. Jeong, J. O.; Kim, M. O.; Kim, H.; Lee, M. Y.; Kim, S. W.; Ii, M.; Lee, J. U.; Lee, J.; Choi, Y. J.; Cho, H. J.; Lee, N.; Silver, M.; Wecker, A.; Kim, D. W.; Yoon, Y. S. Dual angiogenic and neurotrophic effects of bone marrowderived endothelial progenitor cells on diabetic neuropathy. Circulation 119:699–708; 2009. 9. Kim, K. L.; Shin, I. S.; Kim, J. M.; Choi, J. H.; Byun, J.; Jeon, E. S.; Suh, W.; Kim, D. K. Interaction between Tie receptors modulates angiogenic activity of angiopoietin2 in endothelial progenitor cells. Cardiovasc. Res. 72:394– 402; 2006. 10. Langton, A. K.; Herrick, S. E.; Headon, D. J. An extended epidermal response heals cutaneous wounds in the absence of a hair follicle stem cell contribution. J. Invest. Dermatol. 128:1311–1318; 2008. 11. Loomans, C. J.; de Koning, E. J.; Staal, F. J.; Rookmaaker, M. B.; Verseyden, C.; de Boer, H. C.; Verhaar, M. C.; Braam, B.; Rabelink, T. J.; van Zonneveld, A. J. Endothelial progenitor cell dysfunction: A novel concept in the pathogenesis of vascular complications of type 1 diabetes. Diabetes 53:195–199; 2004. 12. Marti, G.; Ferguson, M.; Wang, J.; Byrnes, C.; Dieb, R.; Qaiser, R.; Bonde, P.; Duncan, M. D.; Harmon, J. W. Electroporative transfection with KGF-1 DNA improves wound healing in a diabetic mouse model. Gene Ther. 11: 1780–1785; 2004.

13. Michaels, J.; Churgin, S. S.; Blechman, K. M.; Greives, M. R.; Aarabi, S.; Galiano, R. D.; Gurtner, G. C. db/db mice exhibit severe wound-healing impairments compared with other murine diabetic strains in a silicone-splinted excisional wound model. Wound Repair Regen. 15:665– 670; 2007. 14. Naruse, K.; Hamada, Y.; Nakashima, E.; Kato, K.; Mizubayashi, R.; Kamiya, H.; Yuzawa, Y.; Matsuo, S.; Murohara, T.; Matsubara, T.; Oiso, Y.; Nakamura, J. Therapeutic neovascularization using cord blood-derived endothelial progenitor cells for diabetic neuropathy. Diabetes 54: 1823–1828; 2005. 15. Paik, S. G.; Fleischer, N.; Shin, S. I. Insulin-dependent diabetes mellitus induced by subdiabetogenic doses of streptozotocin: Obligatory role of cell-mediated autoimmune processes. Proc. Natl. Acad. Sci. USA 77:6129– 6133; 1980. 16. Park, D. H.; Borlongan, C. V.; Willing, A. E.; Eve, D. J.; Cruz, L. E.; Sanberg, C. D.; Chung, Y. G.; Sanberg, P. R. Human umbilical cord blood cell grafts for brain ischemia. Cell Transplant. 18:985–998; 2009. 17. Rehman, J.; Li, J.; Orschell, C. M.; March, K. L. Peripheral blood “endothelial progenitor cells” are derived from monocyte/macrophages and secrete angiogenic growth factors. Circulation 107:1164–1169; 2003. 18. Shantsila, E.; Watson, T.; Lip, G. Y. Endothelial progenitor cells in cardiovascular disorders. J. Am. Coll. Cardiol. 49:741–752; 2007. 19. Singer, A. J.; Clark, R. A. Cutaneous wound healing. N. Engl. J. Med. 341:738–746; 1999. 20. Suh, W.; Kim, K. L.; Kim, J. M.; Shin, I. S.; Lee, Y. S.; Lee, J. Y.; Jang, H. S.; Lee, J. S.; Byun, J.; Choi, J. H.; Jeon, E. S.; Kim, D. K. Transplantation of endothelial progenitor cells accelerates dermal wound healing with increased recruitment of monocytes/macrophages and neovascularization. Stem Cells 23:1571–1578; 2005. 21. Tamarat, R.; Silvestre, J. S.; Le Ricousse-Roussanne, S.; Barateau, V.; Lecomte-Raclet, L.; Clergue, M.; Duriez, M.; Tobelem, G.; Levy, B. I. Impairment in ischemiainduced neovascularization in diabetes: Bone marrow mononuclear cell dysfunction and therapeutic potential of placenta growth factor treatment. Am. J. Pathol. 164:457– 466; 2004. 22. Tanii, M.; Yonemitsu, Y.; Fujii, T.; Shikada, Y.; Kohno, R.; Onimaru, M.; Okano, S.; Inoue, M.; Hasegawa, M.; Onohara, T.; Maehara, Y.; Sueishi, K. Diabetic microangiopathy in ischemic limb is a disease of disturbance of the platelet-derived growth factor-BB/protein kinase C axis but not of impaired expression of angiogenic factors. Circ. Res. 98:55–62; 2006. 23. Urbich, C.; Dimmeler, S. Endothelial progenitor cells: Characterization and role in vascular biology. Circ. Res. 95:343–353; 2004. 24. Werner, S.; Breeden, M.; Hubner, G.; Greenhalgh, D. G.; Longaker, M. T. Induction of keratinocyte growth factor expression is reduced and delayed during wound healing in the genetically diabetic mouse. J. Invest. Dermatol. 103:469–473; 1994. 25. Werner, S.; Grose, R. Regulation of wound healing by growth factors and cytokines. Physiol. Rev. 83:835–870; 2003. 26. Wu, K. K.; Huan, Y. Streptozotocin-induced diabetic models in mice and rats. Curr. Protoc. Pharmacol. 40: 5.47.1–5.47.14; 2008.