Impaired cutaneous T‐cell attracting chemokine elevation and adipose‐derived stromal cell migration in a high‐glucose environment cause poor diabetic wound healing

Abstract

Diabetic wound care is a major health care concern. The major cause of non‐healing of wounds in patients with diabetes mellitus (DM) patients mainly involves poor glycemic control, which hinders the migration of progenitor cells including mesenchymal stem cells to the wound site. In this study, we introduced adipose‐derived stromal cells (ADSCs) into wound sites and demonstrated that the local transplantation of ADSCs accelerated DM‐related wound healing. Furthermore, the migration ability of ADSCs, which diminishes in a high‐glucose environment, was partially restored by the exogenous replenishment of the cutaneous T‐cell attracting chemokine (CTACK/CCL27). Our findings suggest that CTACK is a potential novel therapeutic target in DM‐related wound healing.

Introduction

Diabetes mellitus (DM) is one of the most common noncontagious diseases and is currently the leading health concerns worldwide. Since the past decade, the number of patients with DM increased gradually, and DM is the fifth leading cause of death in Taiwan. More than one in ten people in Taiwan are estimated to develop DM during their lifetime. The prevalence of DM is as high as 25% in the elderly population, particularly in men aged >65 years. Diabetic foot is one of the major syndromes of DM [1], which might manifest as intermittent claudication, resting pain, and poor‐healing ulcers. Approximately 25% of patients with DM require major limb amputation during hospitalization for diabetic foot ulcers [2]. Therefore, developing a highly effective therapeutic strategy for DM‐related nonhealing wounds is an urgent priority.

Nonhealing wounds in DM are attributable to multiple factors, such as growth factor deficiency, cell dysfunction, and microvasculopathy, and these wounds are mainly caused by poor glycemic control [[3], [4], [5]], which results in the progression of atherosclerosis and the thickening of the capillary basement membrane. These vascular changes cause hypoperfusion of tissues, narrowing or blockage of blood vessels, and ischemia of the lower limbs [6]. In the typical wound healing process, progenitor cells, such as bone marrow mesenchymal stem cells in the vascular endothelium, migrate from the bone marrow and enter circulation before reaching the wound site and participating in the process of neovascularization and tissue repair [[7], [8]]. However, these events do not occur in patients with DM because progenitor cells are unable to migrate from the bone marrow to the wound site owing to pathological changes in blood vessels.

Adipose‐derived stromal cells (ADSCs) appear to be one of the most promising candidates to address problems related to wound healing. The phenotypic and functional characteristics of ADSCs are highly similar to those of bone marrow‐derived mesenchymal stem cells, which not only exhibit the ability to differentiate into various lineages, including adipogenic, chondrogenic, osteogenic, and myogenic lineages [[9], [10], [11]], but also promote cell growth and wound healing by releasing various growth factors such as vascular endothelial growth factor, hepatocyte growth factor, and basic fibroblast growth factor [[9], [12], [13]]. Furthermore, ADSCs are abundant and can be easily obtained; thus autologous transplantation can be performed without requiring ex vivo expansion. The administration of ADSCs has been shown to promote wound healing in diabetic animal models [[14], [15], [16]].

Chemokine–chemokine receptor interaction plays crucial roles in the migration of mesenchymal stem cells during wound healing [[8], [17]]. During the process of wound healing, various types of cells, including progenitor cells, are recruited to the wound site because of the inflammatory cytokines released from the wound. For instance, cutaneous T‐cell attracting chemokine (CTACK/CCL27) is constitutively expressed in the normal skin and is upregulated in wound sites. A local injection of keratinocyte‐secreted CTACK/CCL27 near wound sites improved wound healing by recruiting circulating progenitor cells [7]. However, the role of CATCK/CCL27 in DM‐related wound healing, which has not been extensively studied, remains unclear. The relationship between the CTACK/CCL27 expression level and ADSC migration during DM‐related wound healing remains unknown. Therefore, we aim to examine whether CTACK/CCL27 promote wound healing in patients with DM.

In this study, we demonstrated that locally transplanted ADSCs improved DM‐related wound healing in a streptozotocin (STZ)‐induced diabetic rat model, the secretion of CTACK/CCL27 by keratinocytes decreased in a high‐glucose environment, and increasing the CTACK/CCL27 level recruited a relative high number of ADSCs even under a high‐glucose condition. Down regulated CTACK/CCL27 expression could be one of the causes of poor wound healing in patients with DM.

Materials and methods

Diabetic rat model

Male Wistar rats (200–250 g) were provided by the National Laboratory Animal Breeding and Research Center (Taipei, Taiwan) and were housed under a constant temperature and controlled illumination. This study was approved by the Animal Care and Use Committee of the Kaohsiung Medical University. The fasting blood glucose levels of the rats were examined before DM induction. Under mild isoflurane anesthesia, a single dose of STZ was administered into the saphenous vein of each rat (50 mg/kg, Sigma–Aldrich) to induce DM as described by Junod et al. [18]. In addition, the rats injected with 0.1 mol L⁻¹ citrate buffer served as the control group. After 1 week, blood samples were collected for determining the glucose level. The rats that had elevated blood glucose levels in the range of 330–480 mg/dL within 1 week of STZ injection were considered as DM rats.

Isolation and expansion of rat ADSCs

Fat tissues were isolated from the groin areas of 6‐7‐week‐old rats. Fat tissues were sliced and washed with phosphate buffered saline (PBS) and centrifuged to remove blood clots. The fragments of the subcutaneous fat were incubated for 3 h at 37 °C with continuous agitation in Dulbecco's Modified Eagle Medium (DMEM) containing collagenase I solution. The tissue samples were then collected in tubes for centrifugation. Five milliliters of DMEM containing 10% fetal bovine serum was then added to re‐suspend the pellet. The isolated cells were seeded in tissue culture flasks with culture medium at 37 °C in a 5% CO₂ incubator.

The next day, the cells were washed with PBS and provided with fresh culture medium at 37 °C in a 5% CO₂ incubator. The isolated ADSCs were refreshed with defined keratinocyte serum‐free medium (keratinocyte‐SFM Gibco, MA, USA) every 2 days until the cell monolayer reached to 95% confluence. After 1 week, the cell monolayer was trypsinized for passages by using Trypsin–EDTA and neutralized using the fresh medium before centrifugation. The ADSCs were expanded ex vivo for up to three passages (P3) in a culture medium. The identity of the surface markers of the isolated ADSCs was confirmed using an LSR II Flow cytometer (BD Biosciences, NJ, USA) with the FlowJo software (FlowJo LLC, OR, USA). CD29, CD45, CD90 (Biolegend, CA, USA), CD31, and CD34 (BD Biosciences) antibodies were used for analysis (Fig. S1A).

Isolation and expansion of human ADSCs

The protocols for obtaining human ADSCs used in this study were approved by the Institutional Review Board of Kaohsiung Medical University Hospital (KMUH‐IRB‐2014‐08‐04). The procedure was explained clearly to patients, and they signed the informed consent forms prior to surgery. Residual subcutaneous adipose tissues were obtained from normal adults who had undergone plastic surgery. ADSCs were further isolated from adipose tissues and cultured in a humidified incubator at 37 °C with 5% CO₂ as described previously. The lineage differentiation ability of the isolated ADSCs was confirmed through induction by using oil red O (adipogenesis), alizarin red S (osteogenesis) or alcain blue (chondrogenesis) staining (Fig. S1B).

STZ‐induced diabetic excision wound model

The excision wound models were used to evaluate the wound healing activity in the rats. All wounding procedures were performed in aseptic conditions. Excision wounds were inflicted on the rats according to the method of Morton and Malone [19].

The rats were anesthetized using 1 mL of intravenous ketamine hydrochloride (50 mg/kg body weight), and their backs were shaved using an electric clipper. Two circular wounds were created in the panniculus carnosus on rats' backs by using a 5‐mm punch. The entire wound was left open. The rats were carefully observed to identify and prevent any possible infection. The rats were randomly arranged into six groups with six rats in each group. Each rat had two wounds, which were injected with either 1 × 10⁶ ADSCs cells/mL or PBS into four points around the wounds. The normal control rats (group 1) were treated with PBS only, the experimental control rats (group 2) were injected with ADSCs, the diabetic control rats (group 3) were treated with PBS only, and the diabetic experimental rats (group 4) were treated with ADSCs. The other rats with DM and non‐DM were treated with cell‐free fresh culture media and served as another control groups (group 5 and 6, respectively). To assess the wound closure in all the groups, the dimensions of each (mm) wound was measured using a ruler and recorded photographically at the following time points: 0, 1, 2, 5, 7, 10, and 12 days. The level of wound healing was evaluated through histological analysis.

Hematoxylin and eosin staining

For histological analysis, the tissues were fixed in 4% paraformaldehyde for 1 h at 4 °C. After washing with PBS, they were incubated overnight in 30% sucrose at 4 °C, followed by 16 h incubation at 4 °C in a 50/50 mixture composed of 30% sucrose and optimal‐cutting‐tissue (OCT) media. Subsequently, the tissues were transferred to 100% OCT and frozen at −80 °C prior to sectioning. Cryosectioning was performed on Hyrax C60 cryostat. Cryosections of 4‐ μm thickness were prepared from the paraffin‐embedded skin tissue of scars after wound healing and the skin tissue surrounding the wound. All sections were stained with hematoxylin and eosin (H&E). The histological tissue sections were used for observing the growth rate of new epithelial cells and micro vessels within the wounds to determine the extent of wound healing.

Culture of keratinocytes

Keratinocytes were purchased from the American Type Culture Collection (USA) and cultured in keratinocyte‐SFM (#10724‐011, Gibco) supplemented with 25 μg/mL bovine pituitary extract and 5 ng/mL recombinant human epidermal growth factor. The culture medium was refreshed every 2–3 days.

CTACK expression level analysis

CTACK expression triggered by tumor necrosis factor (TNF‐α) was analyzed using Quantikine ELISA kit (DCC270, R&D Systems, Minneapolis, MN, USA), and 2.5 × 10⁶ human keratinocytes were plated in advance and incubated overnight to allow attachment. The cells were then treated with TNF‐α for 24 h followed by conditioned medium collection. The level of CTACK/CCL27 was examined using the kit following the instruction manual.

Transwell migration assay

In this study, we examined ADSC migration on Transwell Permeable Supports (3422, CORNING, NY, USA). ADSCs (1 × 10⁴ cells) were plated on the upper layer of the cell culture insert coating with a permeable membrane and incubated overnight. The coating was inserted into a culture well containing 300 ng L⁻¹ CTACK/CCL27 for 6 h. The cells were then washed with 1 × PBS and fixed with 4% paraformaldehyde. The fixed cells were washed four times with 1 × PBS and stained using crystal violet. The cells that had not migrated were wiped off by using cotton swabs. The ADSCs that had migrated were observed and photographed through a microscope.

Statistical analysis

All data reported in this study are presented as the mean ± standard error. Statistical differences between two or more groups were analyzed using Student's t test (wound healing area) or the analysis of variance (maturation of wound and quantification of capillaries surrounding the wounds). A p value of <0.05 was considered statistically significant. The data were analyzed using SPSS (version 16, SPSS Inc., Chicago, IL, USA).

Results

Transplanted ADSCs accelerated wound healing in the rats

To examine the effects of ADSCs on wound healing, particularly DM‐related wound healing, 1 × 10⁶ autologous ADSCs were transplanted into the wound sites in both the normal rats and the rats with STZ‐induced DM. As expected, the DM rats exhibited delayed progression of wound healing. Furthermore, DM‐related wounds exhibited improved healing after the administration of ADSCs. The transplanted ADSCs also accelerated wound healing between day 5 and day 8 in DM rats compared with the non‐DM control rats. By contrast, the wounds treated with PBS alone exhibited healing that was delayed by more than 10 days after the administration of ADSCs in the rats with STZ‐induced DM (Fig. 1). These data revealed that ADSCs exert positive effects on wound healing even under diabetic conditions.

Figure 1.

Wound closure analysis in the normal and DM rats. Both the normal rats and the rats with STZ‐induced DM were treated with PBS or ADSCs around the wound for 12 days (n = 6 in each group). The images indicate the wound sizes at different time points.

Analysis of the wound healing process (wound closure analysis)

To examine the process of wound healing, the STZ‐induced diabetic rat wound model was used for the administration of ADSCs as well as PBS alone controls. The wound healing process was evaluated through the analysis of digital photographs of the wounds on indicated days. Samples measurements were conducted after blinding in all three groups. The time required for wound closure was defined as the number of days required for the wound bed to become completely re‐epithelialized and filled with new tissues. The wound area was measured by tracing the wound margin and calculated using the image analysis program Image J (NIH, USA) [20]. The percentage of wound closure was calculated according to the following equation:

$$\text{wound}\text{closure}\left(\text{\%}\right)=\frac{\text{area}\text{of}\text{original}\text{wound}‐\text{area}\text{of}\text{actual}\text{wound}}{\text{area}\text{of}\text{original}\text{wound}}\mathrm{x}100\text{\%}$$

As shown in Fig. 2A, the rats with STZ‐induced DM exhibited delayed wound healing compared with the non‐DM rats. Although the ADSC‐transplanted group showed a higher level of wound closure than did the other two non‐DM control groups on day 7, the ADSC‐transplanted group exhibited a significantly higher level of wound closure in the rat model of STZ‐induced DM than in the rat models treated with fresh culture media and PBS alone since day 5 after treatment. Furthermore, the wound healing score was also quantified by measuring epidermal regeneration thickness and the number of capillaries surrounding the wounds based on the result of H&E staining. The samples were harvested from the wounded tissues after 3 days of ADSC administration from the DM and non‐DM rats. As shown in Fig. 2B, compared with wounds on the DM rats, wounds on the non‐DM rats showed considerably faster progression of the newly formed skin tissue, a thicker newly formed skin tissue, and a higher mature tissue regeneration level.

Figure 2.

Transplanted ADSCs improve wound healing in the rats with STZ‐induced DM. A) The line graphs represent the percentage of wound healing in either the control rats or the rats with STZ‐induced DM after treatment with of PBS control or ADSCs at different time points. The values represent the mean of six independent individuals ±standard deviation. *p < 0.05 indicates the statistical significance. B) Photomicrograph showing the wound in DM (a, 100 × ; c, 400 × ) or non‐DM (b, 100 × ; d, 400 × ) rats after H&E staining. The arrow indicates the margin of the wound. The star indicates normal skin.

These results strongly highlighted improvements in wound healing due to ADSC transplantation in DM conditions.

Decreased hADSC migration resulted from suppressed secretion of CTACK/CCL27 by keratinocytes in a high‐glucose environment under TNF‐stimulation

Because the blockage of migration of progenitor cells, such as bone marrow mesenchymal stem cells, into the wound sites has been considered the main cause for poor wound healing in DM conditions, we further investigated potential candidates that could enhance hADSC migration. hADSCs have been reported to express several chemokine receptors, including CCR3, CCR4, CCR6, CCR10, CX3CR1, and XCR1, which are responsible for the migration of hADSCs toward chemokines [[21], [22]]. Among these receptors, CCR10 is one of the receptors of CTACK/CCL27. CCR10‐expressing cells were demonstrated to be attracted to CTACK/CCL27, which is exclusively produced by epidermal keratinocytes [23]. To further examine whether replenished CTACK/CCL27 encouraged hADSC migration in the diabetic condition, we seeded 1 × 10⁴ hADSC cells on the permeable upper layer of the cell culture insert with 300 ng L⁻¹ CTACK/CCL27 in the culture medium on the bottom layer of the well. The migrated cells were fixed, stained, and counted. Consistent with the fact that the migration of progenitor cells to wounds is more difficult in patients with DM than in normal adults, hADSCs cultured in a high‐glucose (25 m mol L⁻¹) environment exhibited a lower migration ability than did those cultured in the standard glucose (5 mmol L⁻¹) control medium. Moreover, the decreased migration ability of hADSCs cultured in the high‐glucose environment was partially restored when the medium was exogenously replenished with CTACK/CCL27; however, the migration ability of hADSCs was still lower than that of cells cultured in a standard glucose level (Fig. 3). Furthermore, prolonged inflammation during DM‐related wound healing may result from elevated levels of TNF‐α. TNF‐α is the major regulator that triggers the elevation of the CTACK/CCL27 level in keratinocytes and has been reported that the level of TNF‐α in chronic wound fluids of leg ulcers is around threefold higher than in normal wound fluids [24]. Therefore, we hypothesize that the TNF‐α‐triggered elevation of the CTACK/CCL27 level in keratinocytes is also suppressed in a high‐glucose environment. To examine this hypothesis, 2 × 10⁵ keratinocytes were plated and treated with TNF‐α in a dose‐dependent manner (0–160 ng/mL) for 24 h followed by CTACK/CCL27 analysis. As shown in Fig. 4, the TNF‐α‐triggered expression of CTACK/CCL27 was lower in the high‐glucose environment than in the controlled glucose environment. Increased TNF‐α‐triggered CTACK level elevation in keratinocytes also exhibited significant suppression under a high‐glucose environment than in a standard glucose environment.

Figure 3.

CTACK/CCL27 administration increased ADSC migration under a high‐glucose condition. Representative micrographs depict CTACK/CCL27‐triggered ADSCs migration based on transwell analysis at different glucose levels (5 and 25 m mol L⁻¹ respectively, magnification, 100 ×). Presented values of bar graphs represent the mean of three independent experiments ±s.d. *p < 0.05 compared with control cells. #p < 0.05 between compared groups.

Figure 4.

TNF‐α triggered CTACK secretion in keratinocytes was suppressed under a high‐glucose condition. 5 × 10⁵ primary keratinocytes were cultured in SFM‐keratinocyte medium containing different glucose levels (5 and 25 mmol L⁻¹, respectively) for 24 h and then incubated with TNF‐α in a dose‐dependent manner for another 24 h. The medium was collected, and the CTACK/CCL27 levels were assessed using a Human Quantikine ELISA kit. The values represent the mean of at least three independent experiments ±standard deviation. *p < 0.05 compared with TNF‐α 0 ng/mL under glucose 5 mmol L⁻¹ condition.

In summary, our findings confirmed the role of ADSCs in wound healing. The poor migration ability of progenitor cells, including ADSCs, due to suppressed CTACK/CCL27 elevation in a high‐glucose environment might be the cause of impaired wound healing in patients with DM.

Discussion

Many factors and diseases such as aging, hypoxia, ischemia, chemotherapy, radiotherapy, and topical steroid application cause poor wound healing. Among these factors, DM‐related hyperglycemia is the most common cause of poor wound healing. Because Taiwan is gradually becoming an aged society, the prevalence of DM and DM‐related poor wound healing will increase.

Several animal and human studies have reported that stem cells were not released from the bone marrow in patients with DM [[20], [25]]. This may be the major cause of poor wound healing observed in patients with DM because the classical processes involved in wound healing, including cell proliferation, migration, neovascularization, and mesenchymal formation, rely on the participation of progenitor cells released from the bone marrow. However, the mechanism through which stem cells act in the wound healing process remains unclear. Stem cells not only trigger tissue regeneration because of their capacity for differentiation but also serve as communicators within the immune system [[20], [26], [27], [28]]. Herein, we reported that locally administered ADSCs successfully facilitated the wound healing process in the rats with STZ‐induced DM. Compared with bone marrow stem cells, ADSCs are much more easily attainable, although they are phenotypically and functionally similar to bone marrow stem cells. ADSCs exhibited higher neovascularization capacity and higher expression levels of metalloproteinase (MMP)3 and MMP9 than did bone marrow‐derived stem cells in a rat model of posterior limb ischemia [29]. Furthermore, ADSCs have exhibited immunomodulatory capabilities both in vitro and in vivo [[26], [27], [28]], thus indicating that ADSCs also participate in immune system regulation, which plays a critical role in the process of wound healing. Several in vitro studies have also demonstrated that stem cells derived from the human bone marrow, cord blood, or adipose tissue showed no difference in the morphology and immunological phenotype [30]. These findings suggest that the application of autologous ADSCs may be one of the most promising cell‐based therapeutic strategies in wound healing, particularly in DM‐related wounds.

Precursor cells recruited by inflammatory cytokines play a pivotal role in wound healing [[31], [32]]. Bone marrow‐derived stem cells are believed to migrate to the wound site to regulate the migration and proliferation of keratinocytes in the dermis and epidermis during the early inflammatory phase of healing [8]. Bone marrow‐derived stem cells can be directly grafted into the epidermis and can contribute to the migration of keratinocytes near the wound margin to facilitate wound healing [[7], [8]]. Therefore, a disturbance in the process of homing of bone marrow derived precursor cells has been considered the cause underlying chronic wounds. The homing of bone marrow‐derived precursor cells from blood to wounds is mediated by keratinocytes that exclusively secrete CTACK/CCL27 [7]. Although the inhibition of CTACK/CCL27 by using neutralizing antibodies suppressed the migration of bone marrow‐derived stem cells, intradermal injection of CTACK/CCL27 enhanced the process [7]. These results suggest that chronic wounds in patients with DM are caused by insufficient levels of CTACK/CCL27 in wound sites. Consistent with this hypothesis, our findings demonstrated that TNF‐α‐triggered elevation in CTACK/CCL27 levels in keratinocytes was suppressed in a high‐glucose environment. However, the specific mechanism underlying the suppression of CTACK/CCL27 in chronic wounds in patients of DM warrants further investigations. Furthermore, the results of this study revealed that CTACK/CCL27 enhanced hADSC migration. The CTACK/CCL27‐induced homing of T lymphocytes depends on the expression of the CCR10 chemokine receptor on the surface of lymphocytes [[31], [32]]. However, whether ADSCs share the same mechanism with T lymphocytes remains unclear.

In summary, our data demonstrated that the significant suppression of progenitor cell migration owing to decreased chemokine‐triggered CTACK/CCL27 secretion by keratinocytes under high‐glucose conditions may be the reason for poor wound healing of chronic wounds in patients with DM. In this study, we demonstrated that exogenously supplied CTACK/CCL27 restored suppressed ADSC migration in a high‐glucose environment. Locally administered ADSCs contributed to accelerate wound healing in the rats with STZ‐induced DM. Impaired wound healing in patients with DM remains a formidable challenge in clinical treatments, and our findings indicated that CTACK might play a role in the interaction of keratinocytes and ADSCs. Although ADSCs have showed a great potential in wound healing, whether the therapeutic function of ADSCs is also available in DM‐related poor wound healing is still unclear. Since many reasons result in DM‐related poor wound healing, our findings suggest a novel possibility that the delayed wound healing in DM patients may due to suppressed CTACK/CCL27 expression and stem cell migration. Here we provide evidences that exogenous CTACK/CCL27 promote ADSCs migration in vitro and administrated ADSCs contribute to accelerate diabetic wound healing in vivo.

Acknowledgments

We would like to express our gratitude to the Center for Research Resources and Development (CCRD) of Kaohsiung Medical University for the technical assistance. This study was funded by grants from Ministry of Science and Technology, Taiwan (MOST 103‐2628‐B‐037‐002‐MY3); Kaohsiung Municipal Ta‐Tung Hospital (kmtth105‐011); and Kaohsiung Medical University Hospital (kmuh98‐8G42). This manuscript was edited by Wallace Academic Editing.

Supplementary data

The following is the supplementary data related to this article:

Supplementary Figure S1 A). Surface marker determination of rat ADSCs. Surface markers of isolated rat ADSCs were examined using an LSR II flow cytometer. B). Lineage differentiation assays of human ADSCs. The lineage differentiation ability of human isolated ADSCs was examined by using oil red O (adipogenesis), alizarin red S (osteogenesis) and alcian blue (chondrogenesis) staining, respectively. Magnification: × 100.1

Appendix A. Supplementary data

Supplementary data related to this article can be found at https://doi.org/10.1016/j.kjms.2018.05.002.