Tsutomu Soma, Shiseido Innovative Science Research & Development Center, 2-2-1 Hayabuchi, Tuzuki-ku, Yokohama 224-8555, Japan, Tel.: +81-45-590-6000, Fax: +81-45-590-6087, e-mail: firstname.lastname@example.org
Abstract: In mammalian skin, the existence of stem cells in the dermis is still poorly understood. Previous studies have indicated that mesenchymal stem cells (MSCs) are situated as pericytes in various mammalian tissues. We speculated that the human adult dermis also contains MSC-like cells positive for CD34 at perivascular sites similar to adipose tissue. At first, stromal cells from adult scalp skin tissues showed colony-forming ability and differentiated into mesenchymal lineages (osteogenic, chondrogenic and adipogenic). Three-dimensional analysis of scalp skin with a confocal microscope clearly demonstrated that perivascular cells were positive for not only NG2, but also CD34, immunoreactivity. Perivascular CD34-positive cells were abundant around follicular portions. Furthermore, CD34-positive cell fractions collected with magnetic cell sorting were capable of differentiating into mesenchymal lineages. This study suggests that dermal perivascular sites act as a niche of MSCs in human scalp skin, which are easily accessible and useful in regenerative medicine.
In mammalian skin tissues, stem cells of epidermal, follicular bulge melanocytes have been well characterized by numerous studies (1–3). In contrast, the existence of stem cells in the dermis is still obscure, even though the number of reports describing dermal stem cells has increased in recent years (4–12). Mesenchymal stem cells (MSCs) were first described as fibroblast precursor cells from bone marrow that are capable of differentiating into adipogenic, osteogenic and chondrogenic lineages (13–15). Recent studies have shown that MSCs can be isolated from various tissues and organs (16–20). With regard to skin, follicular dermal sheath cells could differentiate into adipocytes and osteocytes (21). MSC-like cells were also isolated from human skin tissues, sharing a surface antigen profile with BMSCs or adipose-derived stem cells (ADSCs) (4,6). Furthermore, MSCs were also identified as perivascular cells in multiple human organs including foreskin pericytes (22–24).
Bone marrow-Mesenchymal stem cells are a crucial component of the perivascular niche owing to their interaction with hematopoietic stem cells (25). In adipose tissues, the vasculature plays a crucial role by being a progenitor niche of CD34-positive ADSCs (26). In contrast, the exact in vivo nature and localization of MSCs in human skin remain unclear although perivascular CD34-positive cells are well documented to be involved in wound healing (27).
In this study, we speculated that the human adult dermis contains MSC-like cells positive for CD34 at perivascular sites. To prove this hypothesis, stromal cells isolated from adult scalps were examined for their differentiating potency in vivo and in vitro. Confocal microscopic observation was also performed to visualize the sites of MSC-like cells in the human scalp dermis (see Data S1).
Characterization of stromal cells derived from human scalp dermis
To isolate stromal cells from the human scalp, we employed a simple adherent culture method combined with low serum condition. Isolated dermal stromal cells showed colony-forming ability and spindle-shaped morphology (Fig. 1a). They were double-positive for immunofluorescence of pericyte markers, NG2 (magenta pseudo-colour in Fig. 1a) and PDGF receptor β (green colour in Fig. 1a) consistent well with previous reports (22–24). Furthermore, they could differentiate into mesenchymal lineages (osteogenic, chondrogenic and adipogenic) (Figs 1b and S1). We designated these dermal stromal cells as human scalp-derived mesenchymal stem cells (SC-MSCs). Although stem cells were isolated from human and rodent dermis, their in vivo behaviour was not evaluated in these studies (5–9). Recently, Biernaskie et al. (28) showed that mouse SKPs can contribute to dermal regeneration. To investigate the behaviour of SC-MSC, they were subcutaneously injected into the backs of immunodeficient NOD-SCID mice. One week after injection, SC-MSCs visualized by human vimentin immunostaining were still located in the injected sites of the subcutaneous layer (Fig. S2a). Interestingly, 2 weeks after the transplantation, SC-MSCs moved to both the upper and the lower layers from the injected sites (Fig. S2b). They were engrafted into the host muscle layer at the centre and on the periphery of the graft (Fig. S2b). Four weeks after the injection, the transplanted SC-MSCs had spread in the recipient dermis, surprisingly reaching just beneath the epidermis (Fig. S2c). Furthermore, the central area of the recipient dermis was weakly stained by haematoxylin and eosin staining, indicating dermal regeneration compared with the surrounding area (Fig. 1c, star in c1). As a result, the deposition of human type I collagen visualized by immunoperoxidase staining was detected in the extracellular space among cells in the central area (Fig. 1c, brown colour in c2). In contrast, SC-MSCs were predominantly detected near the blood vessels in the peripheral area (Fig. 1c, arrows in c3).
Perivascular localization of CD34-positive cells in human scalp skin
Recent reports demonstrated that MSCs in adipose tissues reside in the perivascular region as CD34-positive cells (24). Although a recent study revealed that MSCs derived from the adult dermis were CD34-positive similar to ADSCs, their in vivo localization was not determined (29). First, we performed immunocytofluorescence staining of SC-MSCs with CD34 and NG2. SC-MSCs were evidently double-positive for CD34 and NG2 immunofluorescence (Fig. 2a). In particular, small, round SC-MSCs were intensely positive (Fig. 2a, arrows). These observations are consistent with a previous report, which showed that very small, round cells isolated from bone marrow stroma are rapidly self-renewing and have great potency for mesenchymal differentiation (15). To investigate whether SC-MSCs are located in the perivascular region of human scalp, we performed immunofluorescence analysis to compare CD34 and NG2 localization. Three-dimensional confocal images clearly demonstrated that perivascular cells in the dermis of scalp skin were double-positive for CD34 and NG2 immunoreactivity (Fig. 2b). In addition, the colony-forming ability of CD34-positive cells was significantly higher than that of CD34-negative cells (Fig. 2c) in contrast to epidermal keratinocytes (30). In addition, CD34-positive cells isolated from human scalp tissues showed a greater ability to differentiate into mesenchymal lineages (Figs 2d and S3). In the subpapillary dermis, CD34-positive cells were abundantly observed along with blood vessels in the longitudinal direction but not in lymph vessels (Fig. 2e, star). To examine CD34 deposition in the perivascular region of human scalp further, we performed whole-mount immunofluorescence staining using isolated hair follicles. Around hair follicles, CD34-positive cells were abundant around the CD31-positive microvessels of perifollicular portions (Fig. 2f). In particular, CD34 immunofluorescence was sequentially observed along with thick microvessels (Fig. 2g). These observations suggest that perivascular sites act as niches of dermal stem/precursor cells in human scalp.
Here, MSC-like stem cells were isolated from adult scalp skin (designated as SC-MSCs). We demonstrated that transplanted SC-MSCs behaved as dermal fibroblasts generating type I collagen in the regenerating sites of the recipient dermis. Furthermore, SC-MSCs were positive for CD34 and pericyte markers. The magnetic sorted CD34-positive cells showed higher colony-forming and mesenchymal differentiation ability. Three-dimensional analysis of scalp skin with a confocal microscope clearly revealed that perivascular cells were positive for not only NG2, but also CD34, immunoreactivity. In summary, our observations suggest that MSC-like stem cells are present as CD34-positive cells at perivascular sites in human scalp, which are easily accessible and may be useful in regenerative medicine.
Ms Haruyo Yamanishi performed the research, Mr Shigeyoshi Fujiwara designed the research and analysed the data, and Dr Tsutomu Soma designed the research study, performed the research and wrote the paper. We thank Dr Yoshinori Ishii for his cooperation in obtaining materials. We also thank Dr Jiro Kishimoto, Dr Yumiko Ishimatsu-Tsuji and Dr Takeshi Hariya for discussions.
Conflicts of interest
The authors have declared no conflicting interests.