Cytotherapy, 2015; 0: 1e16

In vitro characterization of human hair follicle dermal sheath mesenchymal stromal cells and their potential in enhancing diabetic wound healing


Department of Plastic, Reconstructive and Aesthetic Surgery, Singapore General Hospital, Singapore, and 2Skin Bank Unit, Singapore General Hospital, Singapore

Abstract Background aims. Little is published on the characterization and therapeutic potential of human mesenchymal cells derived from hair follicle (HF) dermal sheath (DS). In this study, we isolated and characterized HF DS-mesenchymal stromal cells (DS-MSCs) with respect to the bone marrow mesenchymal stromal cells (BM-MSCs). We further tested if DS-MSCeconditioned medium (CM), like what was previously reported for BM-MSC CM, has superior wound-healing properties, in both in vitro and in vivo wound models compared with skin fibroblast CM. Methods. DS-MSCs were isolated from HF and cultured in vitro to assess long-term growth potential, colony-forming efficiency (CFE), expression of CD surface markers and differentiation potential. The cytokine expression of DS-MSC CM was determined through an antibody-based protein array analysis. The wound-healing effects of the CM were tested in vitro with the use of human cell cultures and in vivo with the use of a diabetic mouse wound model. Results. In vitro results revealed that DS-MSCs have high growth capacity and CFE while displaying some phenotypes similar to BM-MSCs. DS-MSCs strongly expressed many surface markers expressed in BM-MSCs and could also differentiate into osteoblasts, chondrocytes and adipocytes. DS-MSCs secreted significantly higher proportions of paracrine factors such as interleukin-6 (IL-6), IL-8 and growth-related oncogene. DS-MSC-CM demonstrated enhanced wound-healing effects on human skin keratinocytes, fibroblasts and endothelial cells in vitro, and the wound-healing time in diabetic mice was found to be shorter, compared with vehicle controls. Conclusions. Human HF DS stromal cells demonstrated MSC-like properties and might be an alternative source for therapeutic use in wound healing. Key Words: dermal sheath, hair follicle, medium, mesenchymal stromal cells, wound healing

Introduction Mesenchymal stromal cells (MSCs) are multipotent cells with the ability to differentiate into diverse cell types such as osteoblasts, chondrocytes and adipocytes [1]. They are characterized by a combination of cell surface markers and functional characteristics such as clonogenicity and plasticity. Besides their extensive proliferation and differentiation capacity, they are also able to suppress the activation and proliferation of immune cells and participate in tissue repair and regeneration through paracrine mechanisms [2]. Therefore, they are attractive candidates for clinical usage, with numerous potential applications in cellular therapy, tissue engineering and wound healing [1,3]. Over the past decade, human MSCs have rapidly

moved from in vitro and animal studies into human trials as an emerging therapeutic modality for the treatment of acute and chronic wounds. Chronic wounds remain a challenging clinical problem as more than 50% of these are refractory to conventional treatments. Current series of clinical reports on the use of bone marrow (BM)-derived and adipose tissueederived MSCs in human wounds have shown promising potential, and this augurs well for developed societies with an increasing incidence of chronic diseases [4]. Apart from BM and adipose tissues, human mesenchymal stromal cells (MSCs) can also be isolated from a variety of other tissues such as the amniotic membrane [5], umbilical cord [6e9], cord

*These authors contributed equally to this work. Correspondence: Alvin Wen Choong Chua, PhD, Singapore General Hospital, Skin Bank Unit c/o Plastics, Reconstructive and Aesthetic Surgery, Block 4 Level 3, Room A15, Outram Road, Singapore 169608. E-mail: [email protected] (Received 9 June 2014; accepted 2 April 2015) ISSN 1465-3249 Copyright Ó 2015, International Society for Cellular Therapy. Published by Elsevier Inc. All rights reserved.


D. Ma et al.

Table I. Profiles of donated human cells: HF-DSs, BM-MSCs, NFs and NKs.

Cell type HF-DS-MSCs DS-MSC0414 DS-MSC1020 DS-MSC0717 DS-MSC0825 BM-MSCs BM-MSC0410 BM-MSC0118 BM-MSC0325 NFs NF2412 NF2907 NF_BK1001 NF1213 NKs NKU05



Age of donor (years)


Chinese Chinese Chinese Chinese

45 28 35 38


Chinese Chinese Chinese

29 32 48


Chinese Malay Malay Chinese

22 25 46 32




blood [10] and dermal skin tissues [11]. Because MSCs were originally isolated from the BM, this source of cells remains one of the most extensively studied stem-cell types, and they are being tested aggressively in many areas of medicine [4]. However, BM aspiration is a painful and invasive procedure with the potential risk of infection [6]. There is also a limit to the supply of BM as well as an observed age-dependent reduction in cell number [12]. In fact, there were suggestions that the long-term growth and differentiation potential of BM-MSCs may be limited [3], with no detectable telomerase activity [13]. Thus, identification and characterization of alternative sources of human MSCs are necessary for the progression of regenerative medicine. Skin is a promising source of MSCs because of its large size and easy accessibility. The hair follicle (HF), one of the main appendages of the skin is a known source of multipotent stem cells [14]. Much research has been focused on its epidermal stem cell lineage, which lies within the bulge region of the HF [15]. As for the dermal component of the human adult HF, there are two known active MSC populations (one derived from the dermal papilla [DP] region and the other, from the dermal sheath [DS]), which demonstrate multi-lineage differentiation potential [11,16]. Currently, DP-derived MSCs are more intensely studied compared with DS cells [16,17] because DP is deemed to have a more direct involvement in the regulation of HF development and growth. To date, little is published on the characterization and therapeutic potential of human MSCs derived from HF DS. In this study, we isolated and characterized these HF DS-MSCs and compared some of their properties against human BM-MSCs in terms

of long-term culture, colony-forming efficiency (CFE) and analysis of cell surface marker expression. BM-MSCs were previously reported to enhance wound healing through the use of their conditioned medium (CM) compared with medium derived from normal skin fibroblasts (NF). Similarly, we evaluated the wound healing-efficacy of CM derived from DS-MSCs on in vitro human primary cell cultures and on an in vivo wound-healing model with the use of full-thickness dermal wounds in leptin receptoredeficient (db/db) mice. Methods Isolation and culture of hair follicle DS-MSCs Informed consent was obtained from donors (see Table I for brief profile), and all studies were conducted with adherence to guidelines of the Institutional Review Board (IRB) at the Singapore General Hospital (SGH) with the approval given (2009/559/D). Human hair follicles were microscopically dissected from excess scalp tissue discarded after surgery. They were treated with 0.1% collagenase  (Invitrogen) dispase (Roche) at 37 C for 30 min, and the hair follicle sheaths were separated from the hair follicle shafts. The sheath tissue was minced and digested with 0.05% trypsin (Invitrogen) and 0.02%  ethylene diamine tetra-acetic acid (JT Baker) at 37 C for another 30 min. These DS-MSCs were collected and seeded at density of 2  105 cells per 10-cm Petri dish, with high-glucose Dulbecco’s modified Eagle’s medium (DMEM) (Invitrogen) medium supplemented with 1% L-glutamine (Invitrogen) and 20% fetal calf serum (FCS; from Invitrogen). For longterm cultures, the cells were trypsinized and subcultured at 3.5  105 cells per 10-cm dish. Culture of human BM-MSCs, NF, keratinocytes and umbilical vein endothelial cells All human BM-MSCs, NFs, keratinocytes (NKs) (donated to SGH for research; Table I) and human umbilical vein endothelial cells (HUVECs) (Lonza) were granted approval by the IRB for the current research (CIRB: 2011/271/E). These cryopreserved cells were thawed and cultured in a humidified 5% carbon dioxide (CO2) incubator at 37 C. BM-MSCs were maintained in high-glucose DMEM supplemented with 1% L-glutamine and 20% FCS; NFs were grown in DMEM with 10% FCS only. NKs were maintained in EpiLife growth medium (Invitrogen) and HUVECs in fully supplemented endothelial cell growth medium-2 (Lonza). All described cell cultures above were treated with 1% penicillin/ streptomycin (Invitrogen) mix. These cells were

Characterization of dermal sheath mesenchymal stromal cells passaged on reaching 80% to 90% confluence, and culture medium was changed every 3 days for maintenance. Flow cytometry analysis of cell surface marker expression Freshly isolated adult human HF DS-MSCs (from adherent cultures) were characterized by means of flow cytometry analysis for their cell surface marker expression. Cells were trypsinized, collected, and washed with phosphate-buffered saline (PBS from Invitrogen) twice. These DS-MSCs were then suspended at a density of 1  106 /mL in PBS and stained with fluorescein isothiocyanate or phycoerythrin-conjugated monoclonal antibodies (BD Pharmingen) for 15 min at room temperature. The cell surface markers include CD34, CD45, CD38, CD14, CD24, CD29, CD31, CD44b, CD49d, CD49e, CD49f, CD73, CD90, CD105, CD271 and GD2. For the control group, cells were stained with corresponding immunoglobulin isotypes. Long-term culture and colony-forming efficiency assay To investigate their long-term growth potential, the DS-MSCs and BM-MSCs were serially sub-cultured until their growth potential was exhausted. These cells were plated onto a Petri dish, and the accumulated population doubling (PD) was calculated according to the following formula: PD ¼ (log N/N0)/log2, where N represents the total number of cells obtained at each passage and N0 represents the number of cells plated at the beginning of the experiment. The CFE assays of DS-MSCs and BM-MSCs were also performed first by plating at a density of 200 cells/plate, followed by 12 days of culture. The cells were fixed with 3.7% formalin for 30 min at room temperature and stained with 2% crystal violet to visualize and enumerate colonies formed under a dissecting microscope. The colony numbers were counted and the CFE determined as percentage relative to the 200 cells seeded. Induction of osteogenic, chondrogenic and adipogenic differentiation In vitro differentiation of DS-MSCs into mesenchymal lineage cells was performed. For adipogenic differentiation, cells were seeded at a density of 5  104/cm2 in six-well culture dishes and cultured in DMEM supplemented with 10% FCS, 0.5 mmol/L isobutyl-methylxanthine, 1 mmol/L dexamethasone, 10 mmol/L insulin and 200 mmol/L indomethacin (all from Sigma). After culture in differentiation medium for 2 weeks, the cells were fixed with 10% formalin,


washed and stained with 2% oil red O reagent (Sigma) for 30 min at room temperature to examine the generation of oil droplets in the cytoplasm. For osteogenic differentiation, cells were cultured with DMEM supplemented with 10% FCS, 0.1 mmol/L dexamethasone, 50 mmol/L ascorbate-2-phosphate (Sigma) and 10 mmol/L b-glycerolphosphate (Sigma). After 2 weeks, osteogenic differentiation was detected by fixing the cells with 10% formalin for 15 min at room temperature and then staining them with Von Kossa (Sigma). For chondrogenic differentiation, cells were cultured with DMEM supplemented with 1% FCS, 6.25 mg/mL insulin (Sigma), 10 ng/mL transforming growth factor-b1 (Sigma) and 50 nmol/ L ascorbate-2-phosphate (Sigma). After 3 weeks of culture, the cells were fixed in 10% formalin and visualized by staining with alcian blue (Sigma). Reverse transcriptionepolymerase chain reaction analysis of differentiated DS-MSCs The messenger RNA expression of differentiation markers for adipogenic, osteogenic and chondrogenic lineages were analyzed through the use of reverse transcriptionepolymerase chain reaction (RT-PCR). Cultivated and differentiated HF DS-MSCs were harvested at various time points adopted from methods described by Zuk et al. [18]. Total RNA (1 mg) was isolated with the use of the RNeasy kit (Qiagen). Complementary DNA was synthesized through the use of the SuperScript III first-strand synthesis system (Invitrogen). Sequences of the PCR primers (sense and antisense) were designed with the use of the PrimerQuest program, and the fidelity of the RT-PCR products was verified by comparing their size with the expected complementary DNA bands visualized on a 1.8% agarose gel. The list of gene and primer sequences is shown in Table II. Telomerase activity assay DS-MSCs were collected, washed once in PBS and re-sedimented by centrifugation (2300g  5 min at  4 C); the pelleted cells were then suspended in 200 mL lysis reagent and left on ice for 30 min. Cell lysates  were centrifuged (10,000g  20 min at 4 C), and the supernatant extracts were aliquoted. Serial dilutions of protein extract of DS-MSCs corresponding to 10, 100, 500, 1000 and 5000 cells; the protein extracts were subjected to PCR. PCR product (2 mL) was evaluated with the use of 10% methylethene bisacrylamide non-denaturing gel electrophoresis. SYBR gold (1:10,000, Invitrogen) was used to stain the gel, and the gel was viewed with the use of UV radiation. He-La cells were used as a positive control.


D. Ma et al.

Table II. PCR primer sequences of adipogenic, chondrogenic and osteogenic genes. Adipogenic markers aP2 Sense: 50 -TGGTTGATTTTCCATCCCAT-30 Antisense: 50 -TACTGGGCCAGGAATTTGAT-30 150 bp Lipoprotein lipase (LPL) Sense: 50 -GAGATTTCTCTGTATGGCACC-30 Antisense: 50 -CTGCAAATGAGACACTTTCTC-30 276 bp Peroxisome-proliferating activated receptor-g-1 (PPAR-g1) Sense: 50 -GCTCTAGAATGACCATGGTTGAC-30 Antisense: 50 -ATAAGGTGGAGATGCAGGCTC-30 250 bp Peroxisome-proliferating activated receptor-g-1 (PPAR-g2) Sense: 50 -GCTGTTATGGGTGAAACTCTG-30 Antisense: 50 -ATAAGGTGGAGATGCAGGTTC-30 325 bp Leptin Sense: 50 -GGCTTTGGCCCTATCTTTTC-30 Antisense: 50 -GCTCTTAGAGAAGGCCAGCA-30 325 bp GLUT4 Sense: 50 -AGCAGCTCTCTGGCATCAAT-30 Antisense: 50 -CAATGGAGACGTAGCACATG-30 275 bp Chondrogenic markers Aggrecan Sense: 50 -GCAGAGACGCATCTAGAAATT-30 Antisense: 50 -GGTAATTGCAGGGAACATCAT-30 505 bp Collagen II Sense: 50 -ATGATTCGCCTCGGGGCTCC-30 Antisense: 50 -TCCCAGGTTCTCCATCTCTG-30 260 bp Collagen X Sense: 50 -TGGAGTGGGAAAAAGAGGTG-30 Antisense: 50 -GTCCTCCAACTCCAGGATCA-30 600 bp Osteogenic markers Osteopontin (OP) Sense: 50 -GCTCTAGAATGAGAATTGCACTG-30 Antisense: 50 -GTCAATGGAGTCCTGGCTGT-30 270 bp Osteocalcin (OC) Sense: 50 -GCTCTAGAATGGCCCTCACACTC-30 Antisense: 50 -GCGATATCCTAGACCGGGCCGTAG-30 302 bp Osteonectin (ON): Sense: 50 -TGTGGGAGCTAATCCTGTCC-30 Reverse: 50 -TCAGGACGTTCTTGAGCCAGT-30 400 bp Core binding factor-a-1 (CBF-a1) Sense: 50 -CTCACTACCACACCTACCTG-30 Antisense: 50 -TCAATATGGTCGCCAAACAGATTC-30 320 bp Alkaline phosphatase (AKP) Sense: 50 -TGAAATATGCCCTGGAGC-30 Antisense: 50 -TCACGTTGTTCCTGTTTAG-30 475 bp Control Glyceraldehyde-3P-dehydrogenase (GAPDH) Sense: 50 -GCCAAGGTCATCCATGACAAC-30 Antisense: 50 -GTCCACCACCCTGTTGCTGTA-30 498 bp

Bincinchoninic acidebased protein quantification Bincinchoninic acid protein quantification assay was performed with the use of a kit (Pierce Thermoscientific), in accordance with manufacturer instructions. The average concentrations of 1, 2 and 5 dilutions of concentrated medium were used to establish the protein concentration of DSMSC-CM and NF-CM. Microplate reading was performed on a spectrophotometer (BIO-RAD, Benchmark Plus) at an absorbance wavelength of 570 nm. Antibody-based cytokine arrays A human cytokine antibody array (RAYBIO Antibody array G series-5; RayBiotech) was used to identify the cytokines present in the concentrated CM from both NFs and DS-MSCs. Each CM (100 mL) was normalized up to 200 mg/mL of protein before loading onto the glass slides for further processing and comparison of cytokine profiles, in accordance with manufacturer instructions. The glass slides were measured at an excitation frequency of 532nm with the use of the Axon GenePix 4000B scanner (Axon Instruments), and the data were extracted with the use of the Axon GenePix Pro Median signal intensity of each cytokine investigated was expressed as fold change with respect to the median signal intensity of the internal positive control. GraphPad InStat software (see Statistical analysis in the Methods section) was further used to analyze the extracted data. In vitro proliferation assay To test the proliferative effects of the above-mentioned concentrated CM, MTT [3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide] metabolic assay was used. Briefly, for the MTT metabolic assay, 0.5 mg/mL of MTT was added to the individual seeded wells at the respective time points and incu bated at 37 C for 2 h. The medium was decanted, and blue formazan was eluted from cells by use of 20% sodium dodecyl sulfate in a solution of dimethyl sulfoxide. The eluted samples were measured directly in a microplate spectrophotometer (BIO-RAD, Benchmark Plus), with a test wavelength of 570 nm.

Preparation, collection and concentration of CM When DS-MSCs or NFs reached 80% to 90% confluence, they were washed with PBS twice before adding serum-free DMEM. After 48 h, CM was collected and membrane-filtered (0.22 mm). The CM was further concentrated 50 times, with the use of centrifugal Amicon Ultra-4 PL 10 or Ultra-15 PL 10 filter units (Millipore) with a 10-kDa cut-off.

Migration assay and live cell time-lapse imaging Migration assay was performed by use of the scratch method and live cell time-lapse imaging. NKs were grown to confluence in a 12-well plate before a scratch wound was manually introduced into each well with the use of a yellow pipette tip. The cultures were washed twice with PBS and replaced with

Characterization of dermal sheath mesenchymal stromal cells


Figure 1. Micro-dissection of HF and culture of HF DS cells. (A) Single human HF was isolated from scalp tissue viewed under inverted microscope. (B) The HF DS was carefully separated by means of micro-dissection away from the main shaft of the hair follicle. This is followed by in vitro primary culture of HF DS MSCs after isolation and plated for (C) 4 to 5 days and (D) 10 to 12 days. Scale bar ¼ 50 mm.

2 mL/well of EpiLife basal medium (Invitrogen) with 100 mL of concentrated (50) DS-MSC-CM or NF-CM. The plate was subsequently transferred to an incubating chamber (5% CO2 37 C) within a Nikon Eclipse C1 Plus Confocal Microscope (Nikon) for time-lapse imaging. Images of cell migration were taken at 30-min intervals for a duration of 21 h, collected and analyzed with the use of NIS-Elements software (Nikon). Cell migration was measured by percent closure of the scratch zone selected, as follows: percent closure (%) ¼ migrated cell surface area/total surface area  100. Matrigel endothelial tube formation assay HUVECs were trypsinized and suspended in endothelial cell growth medium-2 media with the respective concentrated CM (protein concentration between 10e50 ng/mL). The cells were then seeded onto Matrigel-coated (Becton, Dickinson) 24-well plates (Corning) at a concentration of 60,000 cells/well. After HUVECs were seeded, the plate was incubated under normal conditions and monitored at 2-h intervals until 26 h. Photos were taken at 2, 4, 6, 8, 10, 22, 24 and 26 h after incubation. The number of tubes and branch points formed were analyzed at 26-h time point. Wound-healing murine model Mice used in this study were obtained from the Jackson Laboratory ( and consisted of

10-week-old, healing-impaired BKS.Cg-Dock7mþ/ þLeprdb/J (db/db) mice selected from a spontaneous diabetes mutation in the leptin receptor gene (Leprdb). After hair removal from the dorsal surface and anesthesia, two 6-mm full-thickness excisional skin wounds were created on each side of the midline by a biopsy punch under aseptic conditions. A concentration of 250 ng/100 mL of DS-MSC-CM/NF-CM was prepared in 1 mL of serum-free DMEM. The animals were randomly selected for injection of vehicle medium/DS-MSC-CM/NF-CM. Each wound received 100 mL of the prepared media. Serum-free DMEM was used as vehicle medium, which served as control in the study. A silicon splint with a 6-mm hole was placed such that the wound was at the center within the splint. This was done to prevent wound contracture. An adhesive (SuperGlue) was used to fix the splint to the skin followed by sutures to stabilize the position before placing Tegaderm (3M) over the wounds. The animals were housed individually. Wound closure Digital photographs of wounds were taken at days 0, 3, 7, 10, 14, 17, 21 and 24. Time to wound closure was defined as the time at which the wound bed was completely re-epithelized. Wound area was measured by tracing the wound margin and calculated with the use of a public-domain image analysis software, ImageJ 1.43m (National Institutes of Health). The percentage of wound closure was calculated as


D. Ma et al. eosin. These samples were evaluated for the thickness of the epidermis from six randomly selected locations by use of Nis-Elements Software, and the average was determined.

Statistical analysis Data were analyzed with the use of GraphPad InStat, version 3.06. The Student’s unpaired t-test was performed for two-group comparison; one-way analysis of variance (ANOVA) was used for multiple group comparisons with use of the Bonferroni post-test. All values represented in this study represent mean  standard deviation, with n representing the number of independent cultures performed or independent wounds. A probability (P) value

In vitro characterization of human hair follicle dermal sheath mesenchymal stromal cells and their potential in enhancing diabetic wound healing.

Little is published on the characterization and therapeutic potential of human mesenchymal cells derived from hair follicle (HF) dermal sheath (DS). I...
3MB Sizes 1 Downloads 11 Views