Environmental reprogramming and molecular profiling in reconstitution of human hair follicles
Authors Erin L. Weber and Cheng-Ming Chuong, Department of Pathology, Keck School of Medicine, University of Southern California, Los Angeles, CA 90033
The ability to synthesize an organ de novo is the ultimate goal of regenerative medicine. Given the difficulty, it is reasonable to first attempt tissue regeneration with a miniorgan, such as the hair follicle. Hair follicles (HF) undergo physiologic cyclical regeneration and cellular components exhibit robust regenerative capabilities (1, 2). The molecular biology of HF stem cells and their niche has been well characterized and scientists are poised to make progress (1). In PNAS, Higgins et al. discuss several unique findings regarding de novo human hair folliculogenesis (3). The authors show that human scalp dermal papillae (DP), cultured in 3D, can induce hair from foreskin and identified gene candidates that may contribute to the DP phenotype.
The mature HF undergoes a cyclical pattern of hair growth (anagen), regression (catagen), rest (telogen), loss (exogen), and renewal (4). Upon injury, hair filaments or follicles themselves can be lost. Regeneration of hair miniorgans can be categorized into at least four groups (Fig. 1A) (5). In the first group, plucking of the hair causes minor injury and induces regeneration of the HF structure from epithelial stem cells and the DP. Bulge stem cells undergo apoptosis and hair germ cells become reprogrammed to repopulate the bulge (6). Cells within the HF have robust regenerative ability, which is regulated by both the intrafollicular microenvironment (7) and the extrafollicular macroenvironment (8, 9). If we simply want to enhance hair growth, it may be practical to screen for small molecules that can modulate the macroenvironment. However, this approach will only enhance existing HFs and cannot increase the number of new follicles.
View larger version:
In this window
In a new window
Download as PowerPoint Slide
(A) Cycling hair follicles with four different types of hair regeneration, listed on the side (5). (B) Defining the engineered hair follicle (27).
The second type of regeneration involves de novo formation of new HFs after severe wounding. In mice, wound-induced neogenesis was observed in the center of a large wound (>1 cm after contraction) (10). This phenomenon is also found to occur naturally in African spiny mice, which shed large portions of full-thickness skin when threatened. This shedding is followed by extraordinarily rapid regeneration of a fully functional skin, including HFs, in 1 mo (11). Following injury, repair and regeneration may act as competitive forces, with regeneration favored in large wounds and driven by macroenvironmental signals (12). In the laboratory mouse, FGF9, which is secreted by γδ T cells, can induce a new placode (13). In spiny mice, the extracellular matrix (ECM) appears to be more loosely organized (11). After all, stem cells that give rise to this regenerative wound healing share the same genome as other cells that produce only a reparative response, suggesting that “endogenous reprogramming” has occurred.
The third type of regeneration involves the reconstruction of HFs through the recombination of follicle parts. For example, the DP is implanted beneath the epidermal portion of a different HF from a different organism or body site. Successful HF regeneration from both rat and human DP cells has been demonstrated (14, 15). Here, Higgins et al. use this strategy to implant DP aggregates beneath human foreskin and are able to generate hair (3), although it remains to be determined whether the new hair epithelia is derived from existing epidermal progenitors or a respecification of cellular fates.
The fourth type of regeneration is the reconstitution of HFs from dissociated cells via tissue engineering. The first success with this type of regeneration was shown using a chamber assay (16). The patch assay, in which dissociated murine dermal and epidermal cell populations were injected into murine hypodermis, resulted in the more rapid and efficient formation of cycling HFs, albeit with disoriented, internalized hair shafts (17). Successful HF formation was demonstrated with embryonic cells from several mammals and high conservation of the basic stages of HF regeneration was noted (18). In the planar hair-forming assay (19), dissociated dermal and epidermal cells are applied as a slurry to a full-thickness murine wound and properly oriented, anatomically accurate HFs extrude externally in 3 wk and can cycle. These procedures allow the expansion of cells through culture and it is possible to use embryonic stem cells or induced pluripotent stem cells (20).
To reprogram adult cells to become DP, one needs to know the molecular profiles of inductive DP and fibroblasts. The signature transcriptional profile of the newborn murine DP has been elucidated using microarray (21). However, although this process highlights which genes may be important for the DP phenotype, it does not easily indicate the set of factors that will successfully reprogram a fibroblast into a DP cell. A similar microarray was performed using human adult scalp DP (22). Surprisingly, the transcriptional profile appears to be much different from what would be expected to simply explain a loss of regenerative capacity. Higgins et al. have identified fresh insights into the transcriptional profile of the human DP cell (3). Culture of adult human DP cells in 2D was associated with a rapid alteration in gene expression. Downstream targets of the Wnt signaling pathway were significantly down-regulated. However, 3D organization partially rescued the DP signature, resulting in a gene-expression profile that is more to similar to that of cells of an intact DP. Although only 22% of the altered genes were rescued by 3D culture and other key signature DP genes likely remain underexpressed, this study provides unique information for reprogramming.
In addition to gene expression, the spatial organization of cells is also important. This group has shown that growth of DP cells in 3D prolongs the phenotype (23). Here, sandwiching of human DP cells grown in 2D culture between the epidermal and dermal layers of human neonatal foreskin did not produce HFs (3). However, HFs were generated when intact DP, or DP cells grown in 3D hanging-drop cultures, were placed between the foreskin layers. Interestingly, some ECM and signaling molecules were up-regulated upon spherical culture, and Higgins et al. identify FLI1 (friend leukemia virus integration 1) as a potential master transcriptional regulator of many rescued genes (3).
Although the murine model is helpful for research, folliculogenesis is only clinically useful if achieved with human cells. The most accessible source of human skin is neonatal foreskin. Early studies demonstrated that, although the foreskin epidermis is able to form HFs in combination with a murine DP, the foreskin dermis lacks the necessary cell populations or signals (24). Until recently, only fetal dermal and epidermal cells from the second trimester were known to generate hair follicles (18). Here, a naturally hairless human epidermis (the foreskin) and the human DP are shown to generate hairs, a finding that has also recently been noted by other groups (3, 25, 26).
The engineering procedure may sometimes generate aberrant architecture. It is important that engineered structures meet the criteria that define an HF (Fig. 1B) (27). A true follicle should possess normal structure, including distally located DP and transit-amplifying cells, proximally located differentiating cells arranged in concentric layers, and a sebaceous gland. In addition, the structure should be able to cycle repeatedly, generating a new hair shaft while extruding the old one, and must undergo extrafollicular regulation appropriate to age and body location.
Here, we briefly reviewed recent progress toward the regeneration, reconstruction, and reconstitution of HFs, highlighting the Higgins et al. (3) report. Much work remains to be done before a clinically useful human HF is generated. Research points to three requirements for HF formation: (i) a competent epidermis, (ii) DP cells that retain hair inductivity, and (iii) epithelial-mesenchymal signaling. Because formation of the DP is felt to be one of the initial steps in folliculogenesis, molecular characterization of the DP is a main focus. The DP phenotype requires a papilla-specific transcriptional profile, paracrine signals from the epidermis, interaction with the ECM, and 3D organization. As more details emerge, reprogramming of the DP cell should be possible. Step by step, scientists are moving toward rebuilding hair follicles and providing new treatment for severe burn and trauma patients. We hope that success in HF neogenesis can pave the way for de novo organogenesis on a greater scale.
This work was supported in part by National Institutes of Health/National Institute of Arthritis and Musculoskeletal and Skin Diseases Grants AR 60306 and 43177 (to C.-M.C.) and by a California Institute for Regenerative Medicine Clinical Fellow Training Grant and an American College of Surgeons Resident Research Scholarship (to E.L.W.).
Feature Article - Biological Sciences - Developmental Biology:
Claire A. Higgins, James C. Chen, Jane E. Cerise, Colin A. B. Jahoda, and Angela M. Christiano
Microenvironmental reprogramming by three-dimensional culture enables dermal papilla cells to induce de novo human hair-follicle growth PNAS 2013 110 (49) 19679-19688; published ahead of print October 21, 2013, doi:10.1073/pnas.1309970110