Diverse cellular players orchestrate regeneration after wounding

Abstract

Fibrosis is one of the largest sources of human morbidity. The skin is a complex organ where interplay between diverse cell types and signalling pathways is essential both in homeostasis and wound repair, which can result in fibrosis or regeneration. This makes skin a useful model to study fibrosis and regeneration. While fibrosis often occurs postinjury, both clinical and laboratory observations suggest skin regeneration, complete with reconstituted cell diversity and de novo hair follicles, is possible. Extensive research performed in pursuit of skin regeneration has elucidated the key players, both cellular and molecular. Interestingly, some cells known for their homeostatic function are not implicated in regeneration or wound-induced hair neogenesis (WIHN), suggesting regeneration harnesses separate functional pathways from embryogenesis or other non-homeostatic mechanisms. For example, classic bulge cells, noted for their role in normally cycling hair follicles, do not finally contribute to long-lived cells in the regenerated tissue. During healing, multiple populations of cells, among them specific epithelial lineages, mesenchymal cells, and immune cells promote regenerative outcomes in the wounded skin. Ultimately, targeting specific populations of cells will be essential in manipulating a postwound environment to favour regeneration in lieu of fibrosis.

1 |. INTRODUCTION: SKIN REPAIR OUTCOMES

The ability to target human stem cells of the skin has a wide variety of applications. A spectrum of diseases and conditions ranging from autoimmune disease to amputation to scarring after a burn injury could benefit in some way by influencing human skin stem cells. In the case of burns, wounded skin typically follows a multistep process of repair that often concludes with fibrosis and scar formation.^1–3 The resultant scar does not have the same complexity or functionality as normal skin.² The underlying dermis has excess collagen fibrils deposited parallel to the skin surface. Hair follicles and their associated glands are absent.⁴ Additionally, the scar tissue does not have characteristic dermal papillae or rete ridges.⁵ Targeting stem cells to promote functional healing is one potential important future therapy.

2 |. FIBROSIS IS ANTI-REGENERATION

While effective at quickly healing damaged tissue, fibrosis yields undesirable outcomes. Fibrosis as a response to damage and a precursor to pathology is a theme throughout the body. In the lung, idiopathic pulmonary fibrosis is lethal without a lung transplant.⁶ Any chronic liver disease can progress to fibrosis, cirrhosis and organ failure.⁷ In the skin, aberrant fibrosis may limit range of motion of joints, as seen in Scleroderma and CREST syndrome.^8–10 Deep burns can often create contractures that are functionally restrictive and distressing.^11–13 While most skin wounds are repaired via fibrosis, studies in some loose-skinned laboratory animals, including mice, have shown a regenerative process postwounding that forgoes traditional fibrosis in favour of complete skin repair complete with hair follicles and organizational complexity.^14–18 The process of regenerating complete functional hair follicles in this manner is termed wound-induced hair neogenesis (WIHN)^19,20 and is a productive avenue to study the targeting of human skin stem cells; WIHN in mice may be contrasted to the lack of rete ridge regeneration in Lanyu pigs, a model for glabrous, hairless skin wounding. In those models, full thickness wounds scar over with fibrosis.²¹ WIHN is comparatively complex, involves many cell types and populations (Figure 1), and does not result in a fibrosis scar. The hair follicle is known to contain multiple populations of skin stem cells^22,23 and has been successfully targeted with various drugs and compounds.^24–26 WIHN may therefore be used to explore potential strategies to influence the outcome of wounded skin away from a scar and towards a fully regenerated result.

FIGURE 1.

Wound-induced hair neogenesis (WIHN) is one of the only examples of adult mammal organogenesis and a well-studied model of skin regeneration. The complex interplay between multiple cell types is essential in regenerating skin while avoiding fibrosis

3 |. THE CELL POPULATIONS OF HAIR FOLLICLES HAVE DISTINCT ROLES

The hair follicle contains many distinct populations of stem cells, some of which are implicated in normal hair cycling, such as bulge cells,^27,28 and some that are also implicated in WIHN, such as Lgr6 + cells.^29,30 While bulge cells have an established role in normal hair follicle cycling and early wound healing,²⁷ they do not always contribute to regenerated hair follicles in WIHN.³¹ Earlier studies showed bulge cells contribute to both keratinocytes and new hair follicles in the context of wounding.³² However, Ito et al showed bulge cells that create an initial wound bed epidermis regress entirely by 20 days postwound and do not typically form any lasting portion of the long-lived regenerated epidermis.³³ In contrast, Page et al found Lrig + cells from the bulge can contribute to regrown epidermis after wounding and persist up to a year later.³⁴ Investigation of the exact contribution and persistence of bulge stem cell derived cells in wounding and regeneration is ongoing, and their role in early wounding at all has been called into question.²⁸

Given these findings, Bulge cells are perhaps not a promising target to pharmacologically influence the final outcome of wounded or damaged skin. The stem cells to target for regeneration might better be an undifferentiated population subset of the hair follicle in vivo that provides long-term repopulation. For example, various parts of the prenatal hair follicle, including the sebaceous gland, hair follicle and interfollicular epidermis, are LGR6+ and have been shown to contribute to wound healing and WIHN postnatally.³⁰ Snippert et al showed that transplanted LGR6 + stem cells isolated during telogen onto the backs of nude mice could completely reconstitute hair follicles.³⁰ Unlike some studies on bulge cells, LGR6 + cell progeny persist past 3 months postwounding and are believed to operate via a non-canonical wnt independent pathway.³⁰ LGR6 + seeded constructs also accelerate healing when transplanted onto wound beds, further emphasizing their importance in regeneration.³⁵ These experiments present LGR6 + cells as a promising target with which to influence skin postwounding.

While LGR6 + cells are of epidermal origin, mesenchymal cells originating in the skin or hypodermis are also known to be involved in regeneration and WIHN. Mesenchymal dermal stem cells of the hair follicle have been shown to make contributions to neogenic hair follicles during WIHN in vivo.³⁶ Also of mesenchymal origin, multiple distinct populations of fibroblasts have been studied for their role in inhibiting WIHN,^37–39 underscoring fibrosis as the foil to regeneration. However, some populations of myofibroblasts associate with regeneration. These include a distinct population of myeloid-derived stem cells that may terminate as myofibroblasts or rare regenerated adipocytes, critical to full functionality of normal human skin.³⁸ Adipocytes are known to be helpful in chronic or non-healing wounds and thus pose a promising target in regenerative therapy.⁴⁰ Shook et al found that adipocyte ablation caused deficient wound healing. They also elucidated a mechanism where dermal adipocytes undergo lipolysis within 24 h of injury and deploy fatty acids to the extracellular matrix, which then recruit macrophages to begin the regenerative process.⁴¹ Moreover, adipocytes do not regenerate in hairless skin, underpinning the tissue-wide nature of skin regeneration during WIHN, as opposed to a piecemeal process wherein only some components may regenerate.⁴² Targeting these specific populations of myofibroblasts, among other cells of dermal/epidermal origin, may allow scientists to influence the outcome of the final healed wound.

4 |. THE IMMUNE SYSTEM AND MICROBIOTA HAVE PRO-REGENERATIVE EFFECTS

Similarly, immune cells of mesenchymal origin also contribute to the wound environment both directly and indirectly. Multiple immune pathways and cell types have been implicated as key players in skin regeneration and WIHN.^1,3 LGR5 + cells (bulge cells) may contribute to WIHN when driven by myeloid-macrophage-secreted TNF.⁴³ Macrophages and neutrophils, classically associated with acute inflammation, have been identified as part of WIHN in its earliest stages.^2,19,43,44 In axolotls, aquatic salamanders capable of regrowing entire limbs, ablation of macrophages completely eradicates limb regrowth, but wound healing—of a fibrosis nature—remains intact.⁴⁵ Macrophages were further elucidated as key players in hair follicle regeneration via quorum sensing in hair-plucked skin,⁴⁶ and previously discussed research highlighted an interaction between dermal adipocytes and macrophages in initiating wound healing.⁴¹ In this vein, γδ T cells are also required for WIHN.⁴⁷ The TLR3 receptor, part of the superfamily of toll-like receptor proteins essential to innate immune recognition and eradication of microbes, has been studied as a primary sensing and effector pathway for WIHN.^48–50 The cellular and signalling complexity of WIHN and regeneration are underscored by these and other studies and emphasize the need to identify key players in the regeneration of skin to reliably affect wounding outcomes.

Host cells are not the only essential component for WIHN. New studies involving the microbiota shed light on the delicate interplay between resident microorganisms, tissue maintenance and wound healing. For example, disorders of the scalp and hair such as seborrhoeic dermatitis and dandruff are associated with disarray in the hair follicle microbiome.⁵¹ Furthermore, WIHN is lowest in germ-free mice, higher in conventionally housed mice and highest in pathogenically infected mice. Parallel studies in human subjects show delayed wound healing when wounds are treated with standard antibiotic cream compared to a simple occlusive medium (manuscript in submission). This is reminiscent of immune pathways known to be active in WIHN, such as TLR3 and its substrate, dsRNA.^48,49,52 This pro-healing role of the microbiome is in line with other studies showing probiotics as a reliable and safe alternative to antibiotics in some wounds.⁵³ Clearly, immune signalling is necessary for WIHN; the presence of bacteria, even at pathological levels, is fundamental for regenerative signalling pathways to operate. Furthermore, experiments highlighting the importance of dsRNA and TLR3 signalling in ablating keratinocyte positional identity also enforce its connection to stemness and regeneration.⁵²

5 |. HOW TO IDENTIFY SKIN STEM CELLS

Evidently, identifying the individual mediators of WIHN and regeneration will help inform approaches for future targeted therapy. The importance of single-cell sequencing technology on elucidating the ancestry and interplay of cells involved in WIHN and regeneration cannot be understated. Distinct populations of fibroblasts, human pluripotent stem cells (HPSCs), keratinocytes and other epidermal residents have been identified through use of single-cell sequencing (scRNA seq) technology.^29,37,38 In the past few years, single-cell sequencing technology has flourished. Abbasi et al lineage traced the fibroblasts responsible for WIHN using single-cell sequencing.³⁶ Researchers at the University of Washington and The Fred Hutchinson Cancer Research Center traced the lineage of every cell in a single developmental stage in a nematode,⁵⁴ with some cell populations identified having as little as two individual members. Such wide and deep sequencing could provide enormous benefit towards understanding the complex interplay between cell types and their progeny in WIHN and regeneration of the skin. The authors reported a reliable method that costs 2–3 cents per cell sequenced and was compatible with fixed tissue, which could also elucidate the temporal sequence of events along with the final product of regeneration. A similar experiment in humans or mice may cost more—scRNA seq can cost anywhere from a few to a few hundred dollars per cell, depending on variables such as the depth of coverage and commercial company used.⁵⁵ Despite the potential expense, such data would be invaluable to furthering our understanding of regenerative biology in mammals.

Temporal and positional sequence of events in biology is extremely interesting but difficult to study. Most of our experimental techniques involve lysing or fixing followed by analysis, destroying this information. However, taking data for single-cell RNA sequence (scRNAseq) analysis along a timeline has proved effective and informative in tracing the ability of cells, stem and otherwise, to undergo reprogramming. Zhao et al tracked the inducibility of a pluripotent phenotype in cells and were able to successfully identify the gene network involved with producing chemically induced pluripotent stem cells (CiPSCs). Because only a small population of cells were successfully driven towards a pluripotent stem cell path, scRNAseq was essential in establishing this paradigm.⁵⁶ In a similar series of experiments using scRNAseq, Schiebinger et al discovered the spectrum of developmental programs activated during cellular reprogramming to different terminal states, including pluripotent stem cells, was much wider than previously thought.⁵⁷ They also identified transcription factors associated with reprogramming efficiency. Such studies showcase the ability of scRNAseq to identify reprogramming cells and the robustness of directing cell fate based on that information. WIHN and regeneration rely heavily on reprogramming cells, and scRNAseq can help identify those cells and their changing epigenetic profile. Aside from scRNAseq, fluorescence in situ hybridization (FISH) can elucidate some positional and temporal importance, as was used when identifying spread of native Lgr5+ and Lgr6 + cells to the wound bed.³⁰ 2-photon intravital imaging of cellular behaviours in vivo is especially promising for important insights. Park et al elegantly displayed the speed and distribution of cellular migration of fluorescently labelled epithelial cells into a wound bed.⁵⁸ Similarly, Heitman et al created a model of how dermal sheath cells travel along the hair follicle with fluorescence imaging.⁵⁹

6 |. PROMOTING STEMNESS WHILE AVOIDING TUMORIGENESIS

Attempting to alter the regenerative potential of wounded skin will inevitably include targeting pathways involved in embryogenesis and tumorigenesis. The spectrum from quiescence to neoplasm is wide and variable, and minor alterations can have catastrophic consequences. However, nature seems quite accomplished at allowing regeneration without devolving into cancer. Interestingly, axolotls, capable of regrowing limbs, are also known to develop cancer less often than other organisms—despite their regenerative potential stemming from their ability to re-access their embryologic capacity for organ and limb formation.⁶⁰ Some argue cancer is as much a stem cell disease as a genetic disease,⁶¹ and delicate caution must be taken when attempting to modulate these pathways in cells behaving as stem cells, such as in previously discussed WIHN. The hair follicle complex in both human and mouse is known to exhibit a strictly coordinated circadian-like rhythm to its cycling that continues with high fidelity even without CNS input,⁶² suggesting strict control over stem cells originating in this area. Furthermore, bulge cells are posited to avoid contribution to WIHN and long-term skin regeneration as a preventative measure against cancer development.²⁸ Furthermore, recent studies have shown a remarkable capability of skin and its cell populations to control regeneration without progressing to neoplasm. For example, Hras mutant hair follicle stem cells were incorporated into normal cycling hair follicles and did not progress into neoplasm.⁶³ Various Ras genes are mutated in upwards of 30% of human cancer, so the hair follicle’s ability to regulate a Rasmutant population is remarkable.

In this regard, tissue environments are known to have a remarkable ability to restrain even driver oncogene mutations.⁶⁴ This effect has been studied as early as 1984, when Dolberg and Bissell showed an apparent inability of injected RSV in chicken embryos to cause sarcoma.⁶⁵ Martincorena et al showed that approximately a quarter of sun-aged skin harbours oncogenic driver mutation without evolving to neoplasm.⁶⁶ P53^wt/− cells harboured in the epidermis show slowed proliferation after 6 months, suggesting the surrounding tissue reacts to the mutant cells and alters its own biology to hold them in check.⁶⁷

Despite these controls and balancing mechanisms, neoplasm does sometimes occur under particular conditions. A Nrf2 activating mutation in fibroblasts accelerated wound closure but resulted in neoplasm.⁶⁸ Perhaps the dichotomy of tumorigenesis and regeneration depends upon specific signalling pathway targeted, perhaps the cell type and its niche, or most likely a combination of both. Regardless, the advent of CRISPR-Cas9 and other genetic editing technologies opens the door to modifying the outcome of wounding via alteration of specific genes while avoiding tumorigenesis. For example, Jackow et al created iPSCs from patients with recessive dystrophic epidermolysis bullosa (RDEB), a disease caused by mutation of the COL7A1 gene. Using the CRISPR system, the authors corrected point mutations of COL7A1 in the iPSCs, which then produced wild type collagen 7 protein when implanted as organoids on the backs of nude immunodeficient mice.⁶⁹ The authors reported no off-target nuclease activity. Of further interest is the discovery that lineage infidelity markers rise both under conditions of tumorigenesis and wound healing. Ge et al reported this is a result of stress-induced transcription factors.⁷⁰ However, during wound healing, these factors only increase transiently, and instead remain high indefinitely in cancer. When combined with certain lineage markers, this differentiates cancer from regeneration, suggesting an mechanism by which to control the tumorigenesis-regeneration axis.

7 |. TOWARDS THREAPEUTICS

Pharmacologically influencing skin regeneration and WIHN may begin with removing cells or tissue from a patient, modifying it in a laboratory setting, then returning it. This process is not unprecedented. Of note, recent advancements in cancer therapy involve removing T cells, “training” them to detect specific epitopes of the patient’s cancer, and returning them to the body to seek, destroy and proliferate.⁷¹ Similar methodology may be used in modifying the healing process of skin after solving hurdles of costs and logistics. Studies evaluating positional identity of keratinocytes also suggest important considerations for cell therapy logistics. Epidermal keratinocytes and fibroblasts maintain a memory of their positional identity through the expression of various markers, including Krt9 in volar epidermis. dsRNA sensing functions to decrease positional identity, increasing the stem-like nature of the expressing cells.^48,52 After passage 8–9 Krt-9 and other differential marker expression is lost.⁵² This in vitro loss of cellular identity is common and confirms the need for careful phenotypic maintenance of cell types when expanded in vitro. The positional identity of epidermal cells has been shown to be influenced by both intrinsic keratinocyte signalling and local fibroblast cytokine expression.⁷² Understanding these effects at both an individual cell and population level will help scientists and doctors influence wound healing for specific clinical applications. Unpublished results in the Garza laboratory have shown promising advances to modify skin identity with cell therapy. A clinical trial investigating the ability of injected volar fibroblasts to alter skin identity at stump sites of amputees to transform the skin to a volar phenotype is ongoing (clinicaltrials.gov # NCT03947450). This study is an early step in employing cell therapy to alter the differentiation status of a tissue and is promising in allowing amputees to better utilize prosthetics with reduced discomfort and dermatologic sequelae.

Enormous progress has already been made in replacing large swathes of whole epidermis. Hirsch et al demonstrated the ability of patient derived transgenic primary keratinocytes to replace almost the entire epidermis in a severe paediatric case of junctional epidermolysis bullosa.⁷³ The long-term survival of the graft remained successful. This example is a clear demonstration of the possibilities of using laboratory-grown or derived tissue in the use of clinical treatment of dermatologic disorders.

8 |. MODELLING SKIN IN THE LABORATORY

There are further complications with creating an in vitro environment capable of reproducing in vivo conditions. A hybrid between the two may produce the most realistic results while still presenting as a reliable experimental vector in the laboratory. For example, mouse skin organoids, commonly referred to as ex vivo, spontaneously produce hair follicles.²³ Lee et al created an organoid system mimicking complex human skin, including dermis, epidermis, hair follicles and neural circuitry, from human pluripotent stem cells.⁷⁴ The authors used scRNA seq to determine these complex organoids are equivalent to second trimester foetal face skin. Since foetal skin is known to regenerate without fibrosis, this model system could be of particular interest in WIHN and skin regeneration. The organoid system could expand rare cellular populations or allow a robust environment to test drugs to alter wounding and WIHN. As discussed previously, keratinocytes are not the only important cell type in regeneration and WIHN. Figure 1 illustrates a small subset of contributing populations and signalling involved with WIHN. Co-culturing systems are well known to modify identity and growth in other systems, and growing keratinocyte organoids alongside mesenchymal tissue will be essential in recapitulating in vivo conditions to test new drugs and techniques for regeneration. Furthermore, even non-host cells might be critical; the importance of the microbiome in skin wounding and healing has been shown (manuscript in submission), and steps should be taken to address this fundamental component of human skin. There are innumerable species and varieties of microbes with niches in human skin, and further experiments will be needed to address which microbes are essential for regeneration. Nevertheless, based on recent experiments, quantity of organisms is more important than type (manuscript in submission). Overall, a true mimic of the skin in a laboratory setting would need to include, at the very least, human keratinocytes, fibroblasts, adipocytes, macrophages, neutrophils, smooth muscle cells, sympathetic neurons and T cells according to currently published data, as well as a healthy microbiome. This makes understanding the signalling mechanisms and players of WIHN/regeneration in vivo an important asset to dissect these cellular relationships to further recapitulate the delicate balance of participating entities in the regenerative process.

An important caveat in defining these relationships is the differences between human and mouse skin physiology and wound healing. Like many other fields of study, homology exists between mice and humans, but is imperfect. For example, human skin is markedly thicker, and mice do not routinely form hypertrophic scars. This may be particularly important when studying topical drug delivery or other research where skin thickness may alter outcomes. In addition, human skin is adhered to lower tissues, where mouse skin is loose and free from adhesions. Differences between species can make elucidation challenging, but mouse studies are still essential towards the goal of eventual cellular therapy to study WIHN, treat disease and avoid fibrosis.

9 |. CONCLUSION

The potential of skin regeneration to overcome fibrosis and scarring was observed in laboratory animals as early as the 1940s. Research has progressed to a more thorough understanding of the key players in the process. More recent advents in single-cell sequencing technology and imaging experimentation have and will continue to provide more nuanced understanding. Finally, the adaptation of new methods of cell culturing, particularly organoids, may advance understanding even further while allowing scientists to discover next generation cell therapies. Recent and historical studies will allow clinicians and scientists to specifically and iteratively target key players in the regenerative process.