Wound healing is a complex process that requires a well-orchestrated interplay between different tissue structures and a large number of resident and infiltrating cell types. Especially, angiogenesis is an essential step in successful wound healing [1
]. Progenitor cells take part in this orchestration, as they are able to regulate neovascularization and enhance healing by cytokine production [2
]. Thus, it has already been shown that augmentation of local number of progenitor cells ameliorates impaired wound healing [3
One fraction of progenitor cells are endothelial progenitor cells (EPC). EPC arise from the bone marrow and circulate in the blood [6
]. These cells can be identified by uptake of DiLDL after cultivation [7
]. Peripheral injected DiLDL-labeled EPC have been shown to migrate into ischemic tissues [8
]. There, they can adopt endothelial characteristics thus, contributing to neovascularization [10
]. This is at least partially facilitated by direct incorporation into newly formed capillaries [10
]. Furthermore, EPC stimulate endogenous angiogenesis by secreting a variety of angiogenic growth factors [7
]. They also release factors that directly stimulate keratinocyte and fibroblast proliferation during wound healing [7
]. Capillaries can be detected in granulation tissue by CD90 expression, that can be found on activated microvascular endothelial cells and CD31 expression, that is a linage marker for vascular endothelial cells and is also involved in angiogenesis [12
Homing of EPC to injured tissues is triggered by VEGF and SDF-1α that is mainly released by platelets [13
]. For this reason, EPC express vascular endothelial growth factor receptor 2 (VEGFR2) and stromal cell-derived factor 1α (SDF-1α) receptor CXCR4 [13
]. It has been shown that under pathological conditions like diabetes, homing of EPC by SDF-1α expression is impaired [16
Previous studies by us, and others demonstrated that delivering EPC to wounds, either by local injection or systemically significantly accelerates healing and promotes neovascularization in granulation tissues as well as in bone defects [17
]. Suh et al. could also show an improvement in dermal wound healing after local transplantation of human blood derived early EPC into dermal wounds of immunodeficient nude mice [17
Still number and purity of EPC that can be harvested from bone marrow or blood is limited [20
]. Considering this limitation, we wanted to investigate in the following study whether a lower number of EPC locally administered has the same effect on wound healing as a higher number of these cells systemically applied. For this reason, we used a standardized wound model in hairless mice and directly compared systemic versus local administration of cultivated EPC monitoring and epithelialization throughout the healing process as earlier described [21
Materials and methods
Animals care and wound model
All procedures were performed in accordance with the guidelines set by German law for the care and use of laboratory animals. The experimental study was approved by the regulatory authorities (Regierungspräsidium Darmstadt) under the Ethic Approval Number v54-19c2015-F3/12.
Male homozygous hairless mice (SKH-1, 20–30 g, 8–12 weeks, Charles River Laboratories, Sulzfeld, Germany) were housed in separate cages in room temperature (24 °C), light (12 h/day) and airflow regulated rooms. They were fed a balanced rodent diet and water ad libitum.
All procedures were performed with the animals anesthetized with intraperitoneal injection (i.p.) of 100 µL solution containing 2.215 mg of ketamine and 0.175 mg of xylazine hydrochloride. After desinfecting the ears, mice were placed on a plexiglas platform with their ears extended on a microscope slide by placing three permanent loops (9–0, nylon) at opposite poles of their ears. Standardized, circular wounds (2.25 mm in diameter, 125 µm in depth) were created on the dorsum of the ears using a punch. Wounds were positioned between the ears’ anterior and middle principle neurovascular bundles. After the punch incision, a full thickness layer of skin within the punch was dissected away down to the underlying cartilage [21
]. The day of wounding was designated as day 0.
Immediately after surgery, in the groups receiving systemic EPC, 2 × 106
EPC (in 250 µL PBS) or PBS (250 µL) were injected into the tail vein [18
]. In the groups receiving local EPC, 2 × 105
EPC (in 30 µL PBS) or PBS (30 µL) alone was injected directly into the wound [17
]. Wounds were covered with self-adhesive polyurethane foam dressing (Allevyn thin; Smith and Nephew Medical Ltd., Hull UK) and the entire ear was then covered with a bio-adhesive dressing (Opsite; Smith and Nephew Medical Ltd.) to protect the wound from contamination and mechanical irritation.
EPC isolation and culture
EPC were isolated as in our previous work by density gradient centrifugation (20 min, 600 g) with Ficoll (1.077 g/mL, Biochrom, Berlin, Germany) from the spleen of homozygous hairless mice (SKH-1, 20–30 g, 8–12 weeks, Charles River Laboratories, Sulzfeld, Germany) after mechanically mincing using syringe plungers [18
After isolation, total EPC (4 × 106 cells, cell density 2 × 106 cells/cm2) were cultured on fibronectin-coated (10 μg/mL; Sigma, Deisenhofen, Germany) 24-well plates maintained in 0.5 mL endothelial cell basal medium (EBM-2) supplemented with endothelial growth medium SingleQuots (EGM-2 MV; Clonetics, Cambrex, Walkersville, MD) at 37 °C, 5% CO2. Non-adherent cells were removed after 4 days and adherent cells were incubated in medium for another 24 h prior to initiation of the experiments.
To detect EPC in vivo in healing wounds, they (after 5 days of culture) were harvested by Accutase (PAA Laboratories, Pasching, Austria) for 10 min at 37 °C, 5% CO2, and 2 × 106 and cells were re-suspended in 250 µL PBS. To detect EPC incorporation, EPC were pre-labeled with 2.5 µg/mL DiLDL in EBM-2 supplemented with 20% FCS for 1 h at 37 °C, 5% CO2, followed by harvest and administration.
Intravitally, EPC were visualized under a fluorescent microscope and pictures were merged with light microscope image.
Measuring wound reepithelialization and closure
Epithelialization and EPC recruitment were directly visualized and measured using intra-vital microscopy and computerized planimetry. Microscopic area measurements were performed immediately after wounding and every second day thereafter up to complete wound closure. When epithelialization was near completion, the wounds were observed daily to determine the exact day each process was completed. Measurements were performed by placing anesthetized mice with the Plexiglas platform on the stage of an intra-vital microscope (Carl Zeiss, Oberkochen, Germany). The microscope images were captured with a low light camera (DXC-390P, 3CCD color video camera; Sony, Tokyo, Japan) and transmitted through a digital converter (ADVC-100; Canopus, Ruppach-Goldhausen, Germany) to a monitor. Photographic images were analysed by tracing the wound margin and calculating the area using ImageJ software (http://rsb.info.nih.gow/ij/download.html
). The rate of wound closure was expressed as the ratio of the wounded area at each time point divided by the area of the original wound at day 0. The analysis was performed off-line in a blinded fashion by a different investigator not knowing the treatment each animal received.
Measuring wound neovascularization and SDF-1α expression
Tissue samples were taken from the wound area at days 3, 6, 9 and 12 after wounding. The tissues were dehydrated by isopentane, embedded in TissueTek (Sakura Finetaek Europe, Zoeterwoude, Netherlands) and stored at − 80 °C for subsequent examination.
As capillaries can be detected in granulation tissue by CD90 expression and CD31 expression, wounds were stained for these markers as well as for SDF-1α [12
For analysis, 6 µm thick wound sections, prepared as described above, were treated with acetone (− 20 °C, 10 min) and 0.1% hydrogen peroxidase to quench the endogenous peroxidase activity. Sections were stained with primary antibodies (Abcam, Cambridge, UK) directed against CD31 (1:100; ab7388), CD90 (1:150; ab3105), VEGF (1:100; ab1316) and SDF1α (1:100; ab25117) for 1 h at RT. Primary antibodies were detected by HRP-AEC (Abcam) staining according to the guidelines of the manufacturer. All sections were counterstained with hematoxylin and viewed at 100× magnification (Axio Observer; Carl Zeiss, Oberkochen, Germany). The microscope image was captured with a low light camera (AxioCam; Carl Zeiss) and digitized. Photographic images were analyzed by tracing the stained areas and calculating the area, using ImageJ software. The staining of each section was randomized to the mean value of the granulation tissues' area from all groups. The analysis was performed off-line in a blinded fashion by a different investigator not knowing the treatment each animal had received.
Animals were randomly allocated into four treatment groups (n = 10 per group):
PBS sys (control systemic) = Animals receiving systemic PBS (250 μL) alone.
PBS loc (control local) = Animals receiving local PBS (30 µL) alone.
EPC sys = Animals receiving systemic EPC and PBS (2 × 106 cells in 250 μL PBS).
EPC loc = Animals receiving local EPC and PBS (2 × 105 cells in 30 µL PBS).
For immunohistochemical analysis of the wounds were performed on day 3, 6, 9 and 12 after wound creation (n = 8 per group). They were treated in the same manner as the in vivo groups.
Data are presented as the mean ± standard deviation (SD). Statistical evaluation was performed with Kruskal–Wallis test followed by a Dunn post hoc test using a Bonferroni–Holm adjustment with Bias 10.0. Values of p < 0.05 were considered statistically significant. The number of samples examined is indicated by n.
The aim of this study was to determine the effect of local versus systemic EPC treatment on dermal wound epithelialization, neovascularization and closure. With the unique model used, we are able to directly monitor reepithelialization as there is no significant wound contraction as the dermis is connected to the underlying cartilage, thus mimicking the process more accurately to wound healing in humans [21
EPC are already known to enhance tissue regeneration and wound healing [5
]. In response to ischemia or vascular injury, they are released into peripheral circulation. They migrate to damaged tissues and promote endothelial healing and angiogenesis.
We could confirm the ability of EPC to enhance wound healing in this study as it has already been shown by several groups before [17
]. Furthermore, we could demonstrate that local injection of EPC in wounds had almost the same effect as systemic i.v. treatment on wound closure and epithelialization (Figs. 1
). This finding is remarkable, since locally treated wounds received a tenfold lower amount of EPC then their systemically treated counterparts.
Although it is known that delivering cells directly into damaged tissue causes cell loss compared to systemic application, Bonaros et al. and Reinecke et al. also reported that local delivery of cardiomyocytes directly into the heart resulted in a significantly greater number of cells in the infarcted regions [29
]. Regarding this, the effect of local EPC transplantation is even more remarkable as we only applied the cells around the wound margin and not in the local vessels for treatment. Nevertheless, wound healing was ameliorated by local as well as systemic delivery. Thus, our findings show that the cellular effects seem not only to be connected to EPC delivery in local vessels or through blood flow.
Furthermore, it has already been shown that systemic transplantation requires a higher amount of cells due to a poor distribution and low cell survival [31
]. 90% of systemic transplanted EPC are entrapped in undesired organs including the liver, spleen and kidney when cells were intravenously injected [31
]. The positive effects of local EPC transplantation and our conclusions are supported by the findings of Kim et al. They demonstrated that delivering of EPC directly into diabetic wounds not only enhanced neovascularization but also activated the proliferation of local keratinocytes and fibroblasts [32
In cardiovascular disease models, circulating EPC have been shown to preferentially home to ischemic tissue where they are directly incorporated into vessel walls [8
]. We could demonstrate the presence of the systemically as well as locally transplanted pre-labeled EPC at the wound area in our model by intra-vital microscopy. Incorporated endothelial progenitor cells are able to promote neovascularization and cardiac regeneration by releasing growth factors, which act in a paracrine manner to support local angiogenesis and mobilize progenitor cells residing in the local tissue [33
]. They also promote recruitment of monocytes and macrophages which furthermore support neovascularization [17
In our hands, local application of EPC enhanced the expression of CD31 at the wound margin on day 6 after wounding comparing the control groups (Fig. 3
). Systemic treatment also seemed to enhance CD31 expression on day 6 though this finding was not significant. CD31 expression is used to evaluate angiogenesis that in turn leads to a faster wound closure [35
Furthermore, we demonstrated in our study an increase in CD90 expression from day 3 to day 12 after local and systemic EPC transplantation, though this finding was only significant for systemic treatment compared to systemic PBS injection on day 6 after wounding. CD90 is known as a versatile modulator of signalling affecting cellular adhesion, proliferation, survival and cytokine/growth factor responses [36
]. CD90 is used to monitor vascular density in granulation tissue [18
]. Thus, elevated expression of CD31 and CD90 by local as well as systemic EPC transplantation marks enhancement of local vascularization.
In contrast to enhanced expression of CD31 and CD90, there was no significant difference in VEGF expression in wounds after systemic or local EPC transplantation compared to control groups. VEGF is a potent angiogenic growth factor, which induces increased vascular permeability, proliferation, migration and recruitment of EPC from the bone marrow [33
]. EPC attracted to wound site are known to incorporate directly into neovasculature and also augment angiogenesis through the secretion of VEGF [17
]. Therefore, enhanced vascularization after local and systemic EPC application seemed to be independent of VEGF signalling as we found enhanced vascularization by CD31 and CD90 expression. This might be mainly explained by the fact that EPC can be directly incorporated into the neovasculature during the repair [17
]. Furthermore, they promote endogenous angiogenesis by secreting angiogenic growth factors at EPC-incorporated foci which in turn contributes to the development of host-derived neovessels [17
]. Another VGF-independent way of enhancing neovascularization by EPC is the paracrine modulation of endothelial cells by exosomes [37
]. Therefore, we propose that the positive effects of local EPC transplantation on wound healing is partly due to paracrine effects of the EPC and an improved cell–cell interaction between EPC and local endothelial cells.
We also found a significant upregulation of SDF-1α at wound margins on all days investigated after local and systemic EPC transplantation (Fig. 7
). SDF-1α is produced by tissue ischemia or in response to vessel damage [2
]. SDF-1α promotes trans-endothelial migration of progenitor cells toward vascular lesions and in the endothelial damage site, which in turn exposes specific adhesion molecules [38
]. Furthermore, it can also enhance keratinocyte proliferation and migration in vitro helping the process of reepithelialization [39
]. Thus, upregulation of SDF-1α contributes to amelioration of wound healing by local and systemic EPC treatment.
Also under pathological conditions as ischemia or diabetes, transplantation of EPC has been found to be a promising treatment strategy [28
]. Therefore, it will be interesting to investigate whether local treatment can reduce the number of EPC needed in disease models as well to pave the way for clinical trials.