Effects of Keratinocytes Differentiated from Embryonic and Adipogenic Stem Cells on Wound Healing in a Diabetic Mouse Model

Objective. The current study aims to assess the molecular effects of keratinocytes derived from embryonic and adipose-derived stem cells (ADSCs) on wound healing in mice with diabetes mellitus. Materials and Methods. Sixty BALB/c mice were randomly allocated into 6 groups of 10. Following diabetes mellitus induction by intraperitoneal injection of streptozocin, wounds were created and covered with gauze dipped in various solutions: isotonic saline, carrier and transfer medium-engineered dermal template, keratinocytes derived from embryonic stem cells (ESCs), keratinocytes differentiated from ADSCs, or ADSC medium alone. Histopathological changes and immunohistochemical alterations in the activities of cytokeratin 8, cytokeratin 14, epidermal growth factor (EGF), interleukin 8 (IL-8), fibroblast growth factor 2 (FGF-2), monocyte chemoattractant protein 1 (MCP-1), and collagen I were compared among the 6 groups. Results. Histopathological analysis revealed that wound healing was accelerated by application of keratinocytes derived from ESCs. Such cells increased the activities of cytokeratin 8 and cytokeratin 14. No significant among-group differences were noted in terms of IL-8, FGF-2, MCP-1, or collagen I production. Conclusions. Keratinocytes derived from ESCs accelerated wound healing in mice with diabetes mellitus. The beneficial effects were evident both histomorphologically and immunohistochemically. Although keratinocytes derived from ADSCs are readily available, such cells did not accelerate wound healing.

Diabetes is a slowly progressing, chronic metabolic disease and may be associated with impaired or delayed wound healing if the disease is ineffectively controlled.¹ Normal wound healing is a complex process featuring hemostasis, inflammation, cell proliferation and maturation, and remodeling.² These overlapping steps are disrupted in patients with diabetes, causing the development of chronic wounds.³ Foot ulcers are the most common of such wounds and are the leading cause of lower extremity amputations associated with high-level economic losses.⁴ The treatment of chronic foot ulcers remains a major challenge, and the standard treatment includes debridement, antibiotic therapy to eliminate infection, reduction of pressure on the wound bed, and optimization of glycemic control. However, in some cases, chronic diabetic ulcers do not respond to these treatments. In such patients, transplantation of autologous skin or a human skin equivalent should be considered. Efforts are underway to use stem cells to treat wounds that are slow to heal.⁵

Stem cells are undifferentiated cells that can self-renew, proliferate, produce differentiated progeny, and aid in tissue regeneration.⁶ Given such therapeutic potential, multilineage stem cells are especially appropriate for tissue engineering. The 2 principal types of stem cells used to this end are embryonic stem cells (ESCs) and adult (autologous) stem cells.⁷ While ESCs are isolated from the inner cell masses of blastocysts, adult stem cells may be harvested from many tissues, including adipose tissue.⁸

Adipose tissue is one of the most prolific sources of mesenchymal stem cells (MSCs). Subcutaneous adipose tissue is derived from the mesodermal germ layer, which consists of several different cell types. The supportive, stromal vascular fraction can be easily isolated from adipose tissue, as can a mixture of cells including mature adipocytes, preadipocytes, fibroblasts, endothelial cells, resident monocytes/macrophages, and lymphocytes.^9-11 Preadipocytes are multipotent stem cells (adipose-derived stem cells [ADSCs]) similar to bone-marrow stem cells.¹² This type is used in cellular therapies because ADSCs may be safely, effectively, and simply harvested.

Keratins are typical intermediate, heteropolymeric, filamentous epithelial proteins that impart mechanical stability and integrity; keratins also play roles in wound healing and apoptosis.¹³ Monocyte chemoattractant protein 1 (MCP-1), interleukin 8 (IL-8), fibroblast growth factor (FGF) 2, and epidermal growth factor (EGF) are involved in angiogenesis, embryonic development, tumor growth, wound healing, epithelialization, keratinocyte migration, cellular proliferation, and differentiation.^14-17 Collagen I is the most common form of human collagen, playing roles in fibrosis, wound healing, and tissue morphogenesis.¹⁸

The present study evaluates the effects of keratinocytes derived from ESCs and ADSCs on wound healing in an experimental mouse model of diabetes mellitus. Both early and long-term outcomes were characterized histopathologically and biochemically.

The current study was performed in the experimental research laboratory of Dokuz Eylül University (Izmir, Turkey) between May 2013 and August 2013, with approval from the local Dokuz Eylül University Institutional Review Board for experimental animal research (74/2012).

This study included 60 BALB/c mice weighing 28 g to 35 g and equally split among sexes. Prior to surgical intervention, 80 mg/kg to 100 mg/kg ketamin (Ketalar, 002038; Eczacibasi Saglik Urunleri A.S, Luleburgaz,
Kirklareli, Turkey) and 5 mg/kg to 10 mg/kg xylazin (Alfazyne, 0804125-11; Alfasan, Woerden, the Netherlands) were administered intraperitoneally. Mice were housed in separate cages at 20ºC under standard conditions: mean humidity varied between 40% and 50%, lighting was established in accordance with a 12-hour light/dark cycle, and access to food and water was provided ad libitum.

Diabetic mouse model
Diabetes mellitus was induced after intraperitoneal administration of streptozotocin (125 mg/kg). After 4 hours, 20% dextrose was orally supplied in addition to standard food in order to avoid the effects of hypoglycemia. A blood glucose level > 200 mg/dL 1 week after administration of streptozotocin confirmed the diagnosis of diabetes mellitus. Mice with a blood glucose < 200 mg/dL were excluded from the study. Confirmation of blood glucose levels was performed again on days 10 and 14.

Stem cell isolation 
Embryonic stem cells (CGR8; European Collection of Authenticated Cell Cultures, Salisbury, UK) were cultured in culture medium of 4500 mg/L glucose and sodium pyruvate containing Dulbecco’s modified Eagle’s medium (F0445; Biochrom GmbH, Berlin, Germany), 15% fetal bovine serum (S0115; Biochrom GmbH), 1% L-glutamine (K0283; Biochrom GmbH), 1% penicillin/streptomycin (A2213; Biochrom GmbH), 0.1 mm nonessential amino acid (K0293; Biochrom GmbH), 10-6 M 2-mercaptoethanol, (M7522; Sigma-Aldrich, St Louis, MO), and 1000 IU/mL leukemia inhibitory factors (LIF; L5158; Sigma-Aldrich) on mitomycin-C-treated STO (Sandos inbred mice thioguanine/ouabain-resistant mouse fibroblast cells; eFigure 1). The previously¹⁹ described culture and differentiation of ESC protocol were used. 

Of the 60 animals, 4 were allocated for isolation of adipogenic MSCs. After cervical dislocation, intraperitoneal adipose tissue was collected into sterile, prewarmed phosphate-buffered saline (PBS; BE17-513F; Lonza Ltd, Basel, Switzerland) with 5% penicillin-streptomycin (DE17-602E; Lonza Ltd). Tissue was minced and transferred into a culture medium (A1049001; Thermo Fisher Scientific, Waltham, MA) supplemented with 15% fetal calf serum (04-127-1B; Lonza Ltd), 1% L-glutamine (BE1/-605E; Lonza Ltd), 1% penicillin-streptomycin (DE17-602E; Lonza Ltd), 1% gentamicin (15710-049; Thermo Fisher Scientific), and 0.1% amphotericin B (15290-18; Thermo Fisher Scientific) containing 0.075% collagenase type I (C1639; Sigma-Aldrich) for 1 hour at 37ºC for dissociation. Tissue then was centrifuged at 1000 rpm for 5 minutes at 24ºC, and cells were cultured. Cells were microscopically observed under inverted microscope (IX71; Olympus, Tokyo, Japan) with a phase-contrast attachment, and photomicrographs were obtained (eFigures 2, 3).

Both ESCs and ADSCs were cultured separately onto basement membrane matrix (BD Matrigel; BD Biosciences, Franklin Lakes, NJ) coated plates for differentiation. Two days after allowing the cells to adhere to the basement membrane matrix, 0.5 nM bone morphogenetic protein 4 (B2680, Sigma-Aldrich) was added into culture media for a 3-day to 10-day period to allow for keratinocyte differentiation. The cells were then incubated with bromo-2’-deoxy-uridine (BrdU; 11296736001; Sigma-Aldrich) at 1:1000 dilution at 37ºC for 1 hour. The BrdU was used in the detection of proliferating cells and was incorporated into newly synthesized deoxyribonucleic acid during the synthesis phase of the differentiated keratinocytes’ cell cycle.  After transferring to a mouse wound, differentiated keratinocytes could be detected by using anti-BrdU (eFigure 4).¹⁹ 

Differentiated cells at a final concentration of 5 x 10³/mL were harvested using trypsin-ethylenediaminetetraacetic acid (Sigma-Aldrich NV) and cultured on to an engineered dermal template for 2 days.

Surgical procedure

The shaved backs of mice were cleansed topically with isotonic saline. This region was then incised with a number 15 scalpel, creating a 1 cm x 1 cm wound including the epidermis and dermis. Surgical wounds were kept moist by topical gauzes dipped in isotonic saline. Wounds were evaluated daily for epithelization, infection, and inflammation. None of the animals received topical or systemic antibiotics for wound infections. MatriDerm (MedSkin Solutions Dr. Suwelack AG, Billerbeck, Germany), an engineered dermal template, was used as a transfer medium. The collagen-elastin matrix in the engineered dermal template was replaced by autologous cells in 6 weeks to serve as a scaffold for tissue reconstruction.²⁰

Mice were randomly allocated into 6 groups of 10 after induction of diabetes mellitus and creation of a surgical wound. The groups allocated according to the gauze content employed on the wound site were as follows:

Group 1 (sham-control):  Wound area was moistened with an isotonic saline-dipped gauze, and no stem cells were administered at the wound site. Biopsies were obtained on postop days 3, 7, and 21.

Group 2 (engineered dermal template):  A gauze contain-ing engineered dermal template and isotonic saline was applied to the wound site. No stem cells were administered and biopsies were obtained on postop days 3, 7, and 21.

Group 3 (engineered dermal template + ESC days 3 and 7): A gauze containing keratinocytes differentiated from ESCs, engineered dermal template, and isotonic saline was applied. Biopsies were obtained on postop days 3 and 7.

Group 4 (engineered dermal template + ESC day 21): A gauze dipped in a solution containing keratinocytes differentiated from ESCs, engineered dermal template, and isotonic saline was applied. Biopsies were obtained on day 21 (in order to eliminate more ESCs without surgical trauma).

Group 5 (engineered dermal template + ADSC):  A gauze containing keratinocytes differentiated from ADSCs, engineered dermal template, and isotonic saline was applied. Biopsies were obtained on days 3, 7, and 21.

Group 6 (engineered dermal template + ADSC culture media): A gauze dipped in a solution containing culture medium of ADSCs and engineered dermal template was applied to the wound site. Biopsies were obtained on days 3, 7, and 21.

At the end of the study, biopsy specimen was fixed in 10% formalin for 24 to 48 hours at 24ºC, and animals were euthanized using pentobarbital (100 mg/kg) intraperitoneally.

Histopathological analyses
Samples from all groups were embedded in paraffin after routine histochemical embedding procedure. Sections (5 µm) were taken and hematoxylin-eosin staining (01560BBE-08954D; Leica Biosystems Inc, Buffalo Grove, IL) was performed to examine the cellular morphology.

Immunohistochemical analyses
The paraffin-embedded sections were washed with PBS for 15 minutes. Endogenous peroxidase activity was quenched by incubation with 3% hydrogen peroxide (K31355100; Merck, Kenilworth, NJ) for 5 minutes at 24ºC. Cells were then washed with PBS and incubated blocking serum (85-9043; Invitrogen, Carlsbad, CA) for 1 hour. Primary antibodies (anticytokeratin 8 [bs-1106R; Bioss Inc, Woburn, MA], anticytokeratin 14 [bs-1729R; Bioss Inc], EGF [bs-4568R; Bioss Inc], anti-FGF [bs-0217R; Bioss Inc], IL-8 [bs-2981R; Bioss Inc], anti-MCP-1 [bs-1955R; Bioss Inc], and anti-collagen I [bs-0578R; Bioss Inc]), diluted 1:100 in a sodium citrate buffer pH6.0 in accordance with the manufacturer’s instructions, were applied at 4ºC overnight. They were then washed with PBS, and the secondary antibodies (KJ1-26 anti-idiotypic monoclonal antibody) (D011.110; Invitrogen) and streptavidin-horseradish peroxidase conjugate (SA10001; Invitrogen) and were incubated for 30 minutes at each step. Samples were then incubated with diaminobenzidine/hydrogen peroxide (00-2020; Invitrogen) for 5 minutes and counterstained with Mayer’s hematoxylin (HMM999; ScyTek Laboratories; Logan, UT). The negative control samples were processed in an identical manner; instead of primary antibodies, the same type IgGs were used.¹⁹ They were mounted with a medium (AML060; ScyTek) and evaluated under a light microscope (Olympus BX40; Olympus).

Following indirect immunohistochemistry, all specimen was assessed by 2 pathologists blinded to study group information, and immunoreactivity was graded as negative (-), mild (+), moderate (++), and severe (+++).

Histopathological results
All groups exhibited notable epithelial loss and granulation tissue formation on day 3 after wounding; no significant difference was apparent among groups. On day 7, the ESC-treated groups (group 3) exhibited epithelialization, but neither the saline (control) nor dermal template-treated groups (group 2) did so (eFigure 5). The ESC group (group 4) exhibited complete wound epithelialization by day 21; the saline-treated group (control) did not. Late biopsies of the wounds on day 21 showed that keratinocyte proliferation was evident in surrounding tissue. Wound healing was 100% complete in mice treated with ESC-derived keratinocytes. There was no significant difference by means of epithelization between groups 1, 2, 5, and 6. 

Immunohistochemical results
The indirect immunoperoxidase method was used to assess production of cytokeratin 8, cytokeratin 14, EGF, FGF-1, FGF-2, IL-8, MCP-1, and collagen I. The results obtained on postop days 3, 7, and 21 are summarized in Tables 1, 2, 3. Weak cytokeratin 8 immunoreactivity was evident in all groups on post-wounding day 3 but not in the saline-treated and dermal template-treated groups 1 and 2 on days 7 or 21. This weak immunoreactivity was sustained to days 7 and 21 in groups 3 and 4 receiving the dermal template combined with stem cells. Thus, stem cells induced differentiation of epithelial cells. Cytokeratin 14, expressed late in wound healing, was not expressed by group 1 animals on day 3.

Group 3 in which stem cells were applied in combination with the dermal template were strongly immunoreactive for cytokeratin 14 on day 3. Group 3 exhibited weak immunoreactivity on day 7, and group 4 also exhibited weak immunoreactivity on day 21. The stronger immunoreactivity of group 3 compared with groups 1 and 2 was attributable to the stem cells delivered to group 3. In groups receiving stem cells (groups 3, 4), the dermis was notably immunoreactive for EGF in all 3 biopsies. Interleukin 8 activity declined remarkably in group 1, commencing on day 3 and continuing to day 21 (eFigure 6). Groups 2 and 3 exhibited similar IL-8 immunoreactivities on days 3 and 7; stem cell application did not seem to affect such expression. The immunoreactivity of FGF-2 decreased gradually in group 1; the activities were similar in groups treated with isotonic saline and the dermal template (groups 1, 2; eFigure 7). In the early wound healing period, FGF-2 expression was more intense in groups 1 and 2 than group 3, associated with a shorter duration of epithelialization in groups treated with stem cells, which decreased FGF-2 activity. Secreted by both keratinocytes and connective tissue cells, MCP-1 was expressed by all groups, especially early in wound healing (day 3; eFigure 8) and then gradually subsided; no significant among-group difference was detected.

Collagen I, the principal connective tissue component secreted by fibroblasts, contributes to scar formation. In the present study, all groups were immunoreactive for collagen I at all times. Collagen I positivity was most obvious in the group receiving isotonic saline only attributable to a higher level of scar formation.

The investigators studied the effects of keratinocytes derived from ESCs, ADSCs, and an engineered dermal template on wound healing in diabetic mice. Stem cells could be safely and effectively delivered using the dermal template; keratinocytes differentiated from ESCs (rather than from ADSCs) seemed to improve wound healing. In the time since the original isolation of stem cells, many clinical and experimental studies have sought to understand the effects of such cells on wound healing, tumor formation, immune system function, and toxicity in transgenic animal models.^21-23

As the incidence and burden of diabetes mellitus is increasing worldwide, the authors focused on diabetic wounds by evaluating the impact of hyperglycemia, the effects of keratinocytes and a dermal implant, and the benefits of stem cells in this context. An understanding of the molecular basis of wound healing will hopefully facilitate the development of new treatment protocols.

Chronic wounds pose challenges to plastic surgeons and are expensive to treat. Complicated mechanisms featuring many elements are involved in wound healing. Cytokines, growth hormones, and signaling molecules must be expressed in a certain sequence to allow for successful wound healing. Keratinocyte proliferation, epithelialization, and granulation tissue formation are important steps in the cascade. Studies on the roles played by stem cells and tissue engineering in this context are currently popular. Human ESCs have been shown to develop into brain cells in infant mice.²⁴ Zhang et al²⁵ showed that human ESC-derived neural precursors differentiated into neurons and astrocytes in the neonatal mouse brain in the absence of tumor or teratoma formation. Keratinocytes derived from stem cells were used in the present study to avoid tumor formation by contaminating undifferentiated precursor cells.

Cianfarani et al²⁶ found that adult ADSCs of diabetic mice exhibited less capacity for proliferation and migration than those of normal animals. Diabetes mellitus has been shown to trigger MSC loss and to inhibit MSC maturation, delaying wound healing.²⁷ As ESCs exhibit unlimited multiplication and differentiation capacities, the authors chose to use keratinocytes derived from such cells.

Cytokeratin 8 expression is an early indicator of keratinocyte differentiation whereas cytokeratin 14 expression is a late indicator; the former more reliably identifies keratinocytes.^28,29  The dermal implant (an artificial extracellular matrix) not only supported the cells but also played a key role in cell proliferation, differentiation, and appropriate functioning.²⁰ In the present study, an engineered dermal template served as both the medium and carrier of keratinocytes that developed from stem cells, thereby facilitating direct stem cell transfer to the wound; no adverse reaction between stem cells and the dermal transplant was observed. The safety and efficacy of this approach has been previously documented by Killat et al.²⁰

Biopsies obtained on postoperative days 3, 7, and 21 revealed hemostatic, inflammatory, and immunological tissue changes. The extents of epithelial defect formation and granulation were similar among all groups early after wounding. However, the group receiving ESCs (group 3) in conjunction with the dermal implant exhibited remarkable epithelialization by day 7. Interestingly, epithelialization was completed by day 7 in the ESC group but not in the group receiving only isotonic saline. Only limited keratinocyte proliferation was evident in the group treated with the dermal implant alone.

Although cytokeratin 8 was detected in all groups on day 3, only the groups receiving stem cells expressed this protein on days 7 and 21. Either keratinocytes developing from stem cells or differentiated microenvironmental cells may have been responsible for this observation. The persistence of cytokeratin 14 to day 21 in the stem cell groups showed that the cells remained viable after addition to wounds. The similarity of the IL-8 expression levels among all groups may reflect increased inflammation in groups lacking stem cells and also increased keratinocyte IL-8 expression in groups treated with differentiated stem cells.³⁰

Similarly, the prolongation of epithelialization in groups treated with isotonic saline and the dermal transplant alone (groups 1, 2) may explain why these groups expressed higher levels of FGF-2.  All groups were positive for MCP-1 and no significant among-group difference was evident, associated with the fact that many cell types secrete MCP-1 during wound healing.³¹

Chronic wounds express reduced levels of EGF, which not only suppresses the release of antibacterial products but also induces keratinocyte proliferation.³² The higher EGF levels of the stem cell-treated groups may reflect EGF secretion by such cells, and the enhanced epithelialization mediated by stem cells may indicate a better response of keratinocytes to EGF.³² Collagen storage under conditions of chronic inflammation and by wounds is linked to induction of collagen release by FGF in response to delayed epithelialization.³³

Keratinocytes differentiated from ADSCs and control cells did not differ in terms of the immunohistochemical indicators studied. However, transfer of ADSC-derived keratinocytes afforded no advantage in terms of wound healing.

The principal limitations of the present study are the relatively few immunohistochemical tests performed and the subjectivity of such analyses. In addition, the results of experimental studies must be repeatedly tested and human clinical trials performed prior to adoption.

Keratinocytes derived from ESCs accelerated wound healing in a diabetic mouse model. The beneficial effects were evident both histopathologically and immunohistochemically. The use of stem cells after differentiation into target cells may afford protection against tumor formation. An engineered dermal template (a prototype dermal transplant) safely enhanced stem cell introduction. Although keratinocytes derived from ADSCs are readily available, such cells did not aid wound healing in a diabetic mouse model.

This paper was granted by the Dokuz Eylül University Research Projects Coordination Office (Project Grant Number 2012/356, Title: Molecular Analysis of the impact of keratinocytes differentiated from embryonic and adipogenic stem cells on wound healing in diabetes mellitus).

Affiliations: Department of Plastic, Reconstructive and Aesthetic Surgery, School of Medicine, Muğla Sitki Kocman University, Muğla, Turkey; Department of Plastic, Reconstructive and Aesthetic Surgery, School of Medicine, Dokuz Eylül University, İzmir, Turkey; Department of Histology and Embryology, School of Medicine, Celal Bayar University, Manisa, Turkey; and Research Center of Experimental Health Science, Near East University, North Nicosia, North Cyprus

Correspondence:
Sukru Kasap, MD
Department of Plastic, Reconstructive and Aesthetic Surgery
School of Medicine
Muğla Sitki Kocman University
48000 Kötekli/Muğla
Muğla, Turkey
drsukrukasap@gmail.com

Disclosure: This study was funded by Dokuz Eylül University Scientific Research Projects 2013.KB.SAG.018 (TUZ, 2012356). The study was presented as a poster at Wound Care Congress 2013, Titanic Lara Otel, Antalya, Turkey (November 28 to December 1).