- Email: [email protected]
Understanding fibroblast heterogeneity in the skin
TICB-1095; No. of Pages 8
Review
Understanding fibroblast heterogeneity in the skin Ryan R. Driskell and Fiona M. Watt Centre for Stem Cells and Regenerative Medicine, King’s College London, 28th Floor, Tower Wing, Guy’s Hospital Campus, London SE1 9RT, UK
Fibroblasts are found in most tissues, yet they remain poorly characterised. Different fibroblast subpopulations with distinct functions have been identified in the skin. This functional heterogeneity reflects the varied fibroblast lineages that arise from a common embryonic precursor. In addition to autocrine signals, fibroblasts are highly responsive to Wnt-regulated signals from the overlying epidermis, which can act both locally, via extracellular matrix (ECM) deposition, and via secreted factors that impact the behaviour of fibroblasts in different dermal locations. These findings may explain some of the changes that occur in connective tissue during wound healing and cancer progression. Varied origins and functions of fibroblasts Fibroblasts are mesenchymal cells that deposit collagen and elastic fibres of the ECM in connective tissue [1–4]. This simple operational definition, however, masks the considerable heterogeneity of fibroblasts found in different tissues (healthy or diseased) and at different stages of development. Indeed, fibroblasts in different body sites arise from different embryonic origins [5]. For example, fibroblasts in face skin arise from the neural crest; those in ventral body skin derive from the lateral plate mesoderm; and those in back skin originate from the dermomyotome [6–10] (Figure 1). While genetic tools for lineage tracing have led to considerable progress in elucidating stem cells and lineage relationships in most cell types and tissues [11], characterisation of fibroblasts has lagged. This is, in part, because of the challenge of defining clonal relationships when the progeny of an individual cell do not remain attached to one another, and in part because of a lack of defined markers to distinguish different fibroblast subsets. It is important to consider the lineage relationships of fibroblasts because this may influence their differentiation potential [7]. For example, mesenchymal stem cells from the bone marrow share similar properties with tissue resident fibroblasts; however, just because different cell populations express common markers does not mean they have a common origin [12]. Nevertheless, recent advances in marker identification, functional assays, and lineage tracing have led to new insights into fibroblasts in the skin. These findings reveal that functional heterogeneity Corresponding author: Watt, F.M. ([email protected]). Keywords: dermis; connective tissue; stem cells; lineage; Wnt; adipocytes. 0962-8924/ ß 2014 Published by Elsevier Ltd. http://dx.doi.org/10.1016/j.tcb.2014.10.001
reflects, at least in part, the existence of different fibroblast lineages, and that fibroblasts respond readily to signals from the overlying epidermis and thereby exhibit surprising plasticity. These observations provide new ways of interpreting the dynamic changes in fibroblast behaviour – such as proliferation and migration – observed in tissue repair and disease. Mesenchymal cells of skin connective tissue The dermis, the connective tissue underlying the epidermis, provides a good example of how even within one tissue, in a single body site, and at a single developmental stage, fibroblasts are remarkably diverse [13–15] (Figure 2; Tables 1 and 2). In neonatal (P2) mouse back skin, the upper (papillary) dermis is distinguished from the lower (reticular) dermis by differences in fibroblast density and by the maturity of the collagenous ECM [16]. Signals from epidermal stem cells in the hair follicle bulge induce adjacent fibroblasts to form the smooth muscle known as the arrector pili muscle (APM), which controls piloerection [17]. A cluster of fibroblasts at the base of the hair follicle, known as the dermal papilla (DP), has specialized signalling properties required for hair follicle morphogenesis and coordination of the hair cycle [18,19]. The dermal papillae that are associated with different types of hair follicles are distinct [20] and there is evidence that some dermal papilla cells have the ability to differentiate into a wide range of cell types, including nerve and cartilage [21,22]. Further heterogeneity is evident in the region underlying the reticular dermis, known as the hypodermis or dermal white adipose tissue, since this contains a mixture of pre-adipocytes and mature adipocytes [23]. Two other skin mesenchymal cell types that warrant consideration are perivascular smooth muscle cells (pericytes) and mesenchymal stem cells (MSCs), both of which express some markers in common with fibroblasts (Table 1). MSCs form part of the perivascular stromal compartment of the bone marrow and can be isolated from bone marrow cell preparations as the cells that adhere to tissue culture plastic, in contrast to the non-adherent haemopoietic cells [12]. Although there have been some reports that MSCs contribute to the dermis [24–26], more recent studies, involving lineage tracing and bone marrow transplantation, suggest that this is not the case [27–29]. Perivascular smooth muscle cells surround blood vessel endothelial cells and have a contractile function that regulates endothelial cell homeostasis [30]. There are no Trends in Cell Biology xx (2014) 1–8
1
TICB-1095; No. of Pages 8
Review
Trends in Cell Biology xxx xxxx, Vol. xxx, No. x
ers Whisk
blast F i b r o aon ren diffe lineage and itment m com
Neural crest fibroblast precursor
Upper lineage Lower lineage
Facial skin last Fibrob on na differe eage and lin ent itm comm
Somite fibroblast precursor
Hair follic
Upper lineage Lower lineage
les
Back skin
TRENDS in Cell Biology
Figure 1. Developmental origins of dermal fibroblasts. Fibroblasts arise from different developmental origins such as the neural crest, dermomyotome, and lateral plate mesoderm. Regardless of the developmental origin of fibroblasts, they undergo differentiation and lineage commitment to give rise to both upper and lower lineages. Redrawn from [5].
definitive markers of pericytes, but the field has utilized asmooth muscle actin (a-SMA), desmin, and neuron–glial antigen 2 [NG2; also known as chondroitin sulfate proteoglycan 4 (CSPG4)] as markers [31]. In skeletal muscle, pericytes can differentiate into muscle fibres [32]. In skin, pericytes have been identified in the upper dermis and can be isolated with an HD-1 antibody [31]. However, there is currently no evidence that they contribute to the skin fibroblast compartment. When fibroblasts are isolated from adult skin and placed in culture or transplanted to a new location they exhibit
Papillary dermis
positional memory, which is reflected in Homeobox (Hox) gene expression [33,34]. Classic tissue recombination experiments have shown differences in the ability of the dermis from hair-bearing and non-hair-bearing regions of mouse skin to induce hair follicle formation [35]. While positional memory likely accounts for the different types of hair follicles in different body sites, fibroblasts arising from the neural crest, lateral plate mesoderm, and dermomyotome are all competent to support hair follicle formation. Indeed, lineage tracing and cell isolation experiments have shown that dermal cells can exhibit characteristics of
Key: r1 Lg 6 ig Lr
Recular dermis
Upper lineage Dermal papilla Papillary fibroblast Arrector pili
r5 Lg
Dermal sheath
Lower lineage Recular fibroblast Hypodermis
Pre-adipocyte Adipocyte
TRENDS in Cell Biology
Figure 2. Heterogeneity of epidermal stem cells and mesenchymal cells in skin. The locations of three epidermal stem cell populations (Lrig1+, Lgr6+, and Lgr5+) in the hair follicle are shown. The three dermal layers are also indicated: the papillary dermis, reticular dermis and hypodermis/white adipose layer. The specialised fibroblasts of the dermal papilla (DP) and arrector pili muscle (AP) are derived from the same lineage as the papillary dermis, while reticular fibroblasts, pre-adipocytes and adipocytes share a common origin.
2
TICB-1095; No. of Pages 8
Review
Trends in Cell Biology xxx xxxx, Vol. xxx, No. x
Table 1. Mesenchymal cell subpopulations in the skin Cell type Papillary fibroblast
Location Papillary dermis, adjacent to epidermis
Reticular fibroblast Intradermal pre-adipocyte Intradermal adipocyte
Reticular dermis Reticular dermis
Myofibroblast Arrector pili muscle Dermal papilla Mesenchymal stem cell Perivascular cell
Wound/scar Attachment to hair follicle bulge and IFE Attachment to hair follicle bulb Bone marrow Attachment to blood vessels of skin
Dermal white adipose tissue (hypodermis)
neural crest cell derivatives even though they have a nonneural crest origin [7]. However, the extent to which Hox genes regulate local fibroblast positional identity, such as the different characteristics of fibroblasts underlying the scale and interscale interfollicular epidermis in mouse tail dermis, remains to be determined [36]. Markers and lineage relationships of dermal fibroblast subpopulations In order to understand fibroblast heterogeneity in the skin, it is vital to have a set of markers that can be used to define various subpopulations of cells. Several pan-fibroblast markers have been well studied, such as the intermediate filament protein vimentin, the collagen 1a2 chain and Pdgfra, the cellular receptor for platelet-derived growth factor subunit A (PDGFA) [1]. However, these markers are also expressed by other cells in the body, such as endothelial and myoepithelial cells in the case of vimentin, and osteoblasts and chondroblasts in the case of collagen 1a2, making it difficult to accurately identify and trace fibroblast populations through development. Recent lineage tracing of the dermis at different stages of development has provided new insight into fibroblast origins. Specifically, the analysis of mouse back skin at different stages of development, from E12.5, when the dermis appears as a homogeneous tissue, to P2, when all the dermal regions have formed [37], has revealed a remarkably dynamic pattern of marker expression. Recent lineage tracing has revealed that dermal fibroblasts arise from a multipotent mesenchymal cell population at E12.5, which expresses Pdgfra, delta-like homolog 1 (Dlk1), and the leucine-rich repeat protein Lrig1 (Figure 3). Upon differentiation, the fibroblasts begin to express
Function Source of AP, IFE support, ECM production ECM production ECM production, source of adipocytes, long distance signalling Long distance epidermal stem cell signalling, insulation Wound contraction, crosstalk with immune cells Piloerection Hair follicle morphogenesis, hair cycle Not detected consistently in the dermis Vascular constriction and homeostasis
Refs [13,37] [13] [23,76] [23,76] [3] [17] [19,55] [82] [30]
different markers distinguishing common progenitor fibroblasts from papillary dermal fibroblasts of the upper dermis and reticular dermal fibroblasts of the lower dermis. This lineage commitment between upper and lower dermal fibroblasts was observed to occur at approximately E16.5. Indeed, the a8 integrin subunit, cluster of differentiation 26 [CD26; or dipeptidyl peptidase 4 (Dpp4)], Lrig1, and B lymphocyte-induced maturation protein 1 (Blimp1) were expressed in the papillary dermis during early development, which was later confirmed by lineage tracing using Blimp1Cre and Lrig1CreER to label the upper dermal lineage. Furthermore, lineage tracing using Dlk1CreER to label dermal fibroblasts at E12.5 revealed that, by E16.5, Dlk1 (Pref1) was selective for the lower dermis. Dlk1 and Sca1 together were dynamically expressed in the reticular and hypodermis during early skin morphogenesis [37]. Several markers are also differentially expressed in upper (papillary) and lower (reticular) dermis of the human skin [13;37–39]. In addition to the lower and upper dermis, common progenitor fibroblasts can also differentiate into dermal papilla or give rise to papillary dermal fibroblasts and arrector pili muscle cells. DP cells express a number of distinct markers [9,20,40,41]. Gene expression profiling indicates DP markers change according to hair follicle type, stage of hair growth cycle, and stage of development. Alkaline phosphatase and cellular retinoic-acid-binding protein 1 (CRABP1) are expressed in DP cells throughout the hair growth cycle [42], although alkaline phosphatase expression is not confined to fibroblasts with hair inductive properties [40]. Sox2 expression is specifically expressed in the DP of guard, awl, and auchene hairs [20,21]. In addition, lineage analysis of subsets of dermal papillae from
Table 2. Stable and dynamic markers of dermal papilla, papillary and reticular dermis, fat and arrector pili muscle at E16.5, P2, and adult Cell type Pan-fibroblast Dermal papilla Papillary dermis Reticular dermis/hypodermis Arrector Pili Dermal white pre-adipocytes Dermal white adipocytes
Stable marker Col1a2, PDGFRa Crabp1, AlkPhos
Dynamic marker All markers listed below Corin, Sox2, Sox18 Blimp1, Itga8, CD26, Lrig1, EphB2, Trps1 Sca1, Podoplanin, Dlk1
Acta2, Itga8, Sm22a, Npnt CD34, Sca1, PDGFRa
Refs [16] [20,41,42,83] [37;57,84] [23,37,38] [17] [23,37]
Perilipin, adiponectin
[23] 3
TICB-1095; No. of Pages 8
Review
Trends in Cell Biology xxx xxxx, Vol. xxx, No. x
E14.5
Dermal papilla PDGFRa+ CD26– Sox2+
E12.5
Common fibroblast progenitor PDGFRa+ Dlk1+ lrig1+
P2
E16.5
Papillary dermal mal fibroblast progenitor PDGFRa+ Blimp1+ Dlk1– lrig1+ E16.5
Recular dermal mal fibroblast progenitor PDGFRa+ Blimp1– Dlk1+
Dermal papilla D PDGFRa+ CD26– Sox2– P2 Papillar dermal fibroblast PDGFRa+ CD26+ Blimp1– lrig1+
Recular dermal mal fibroblast progenitor PDGFRa+ Blimp1– Dlk1+
APM PDGFRa+ Itga8+ CD26–
Recular fibroblast R oblast PDGFRa+ Dlk1+ Sca1–
Hypodermal fibroblast PDGFRa+ Dlk1+/– Sca1+
Hypodermal adipocyte Hy PDGFRa– Dlk1– lipid+
TRENDS in Cell Biology
Figure 3. Temporal origins and lineage relationships of fibroblasts in mouse back skin. Fibroblasts arise from a common progenitor and progressively differentiate into upper fibroblasts lineages (green) and lower fibroblast lineages (red). Lineage markers are transient and labelled at each time point. Re-drawn from [37].
guard/awl/auchene hair with Sox2CreER revealed that Sox2 fibroblasts do not significantly contribute to wound repair in skin [43]. Additional insights into mesenchymal subpopulations in the skin have been obtained by performing lineage tracing of DP cells. Lineage tracing using Corin–Cre has shown that the progeny of cells that express Corin remain in the dermal papilla and are not found elsewhere in the dermis [41]. Studies in chimeric mice and skin reconstitution assays have shown that dermal papillae have a polyclonal origin and that DP cells can participate in forming new hair follicles without proliferating [44]. Furthermore, lineage tracing of pre-adipocytes using adiponectin–Cre to efficiently label intradermal adipocyte lineages has revealed that adipocyte precursor cells proliferate, while mature adipocytes repopulate skin wounds in parallel with fibroblast migration [45]. As well as providing information about lineage relationships within the dermis, lineage tracing experiments showed that epidermal keratinocytes do not undergo conversion into dermal cells, even under conditions in which keratinocytes are induced to express the epithelial-to-mesenchymal (EMT) marker Slug1 [46]. Similarly, bone marrow derived mesenchymal cells were not found to contribute to the dermis [29,37]. Despite the many advances that have been made by using specific cell surface markers to isolate and characterise fibroblast subpopulations, it is important to note 4
that markers that are specific for one fibroblast population at one time point can be either downregulated or expressed on a different population a few days later. In addition, the lineage tracing approach has caveats, in particular efficiency of labelling and leakiness. Functional heterogeneity of dermal fibroblasts Dermal fibroblasts not only differ in location and gene expression, but also in function. This has been demonstrated in a number of different ways: in cell culture, in cell transplantation experiments and by genetic gain and loss of function approaches. As in the case of epidermal cells, fibroblast behaviour and differentiation can be measured both on the population level and by characterising clones derived from single cells [47,48]. In bulk cultures, it is possible to measure overall proliferation rates, production and responsiveness to growth factors, and differences in cell morphology [13]. Clonal growth of DP cells in hanging drops or captured in hydrogels has shown that hair induction ability can be retained, both in mouse and human, at least at early passages [40,49–55]. Indeed, Sox2-positive cells retain their identity in culture, suggesting that they represent a distinct differentiated lineage [40]. By contrast, upper dermal fibroblasts isolated from E16.5 are competent to differentiate into adipocytes in the presence of adipogenic medium in hydrogel cultures. By P2, however, upper dermal fibroblasts
TICB-1095; No. of Pages 8
Review do not differentiate into adipocytes in culture because lineage restriction has occurred [37]. Mouse skin can be reconstituted from disaggregated epidermal and dermal cells by injecting them into a chamber implanted into a full thickness wound on the back of a syngeneic or immunodeficient mouse [31,37,47]. By sorting specific subpopulations of cells from Pdgfra–Histone-2Benhanced green fluorescent protein (EGFP) expressing dermis, it is possible to follow their fate when combined with an excess of unlabelled and unfractionated epidermal and dermal cells [44]. Transplantation studies have shown that fibroblast proliferation is necessary for new hair follicle formation, but cells that have been irradiated to inhibit proliferation can still contribute to new DP. Furthermore, although new hair follicles do not normally develop postnatally, adult dermal fibroblasts retain the ability to contribute to DP during hair follicle formation in transplantation assays [44] and in response to epidermal b-catenin activation [56]. In skin reconstitution assays, P2 upper dermal fibroblasts (CD26+Sca1–) only give rise to papillary fibroblasts, APM and DP, whereas lower dermal fibroblasts that express Sca1 are restricted to forming the reticular dermis and adipocytes. A multipotent population that is Dlk1+Sca1– can give rise to all the dermal mesenchymal cell types, but cells that express this combination of markers do not persist in adult skin [37]. Lin-CD34+CD29+Sca1+ adipocyte precursors stimulate local hair follicle cycling when injected intradermally [23]. When E14.5 skin was implanted onto kidney capsules it was found that adipocytes formed independently of hair follicle morphogenesis [57]. The availability of Cre mouse lines that target different dermal cell populations has facilitated the analysis of fibroblast function via genetic modification. The manipulation of gene expression via a variety of promoters that express noninducible forms of Cre-recombinase results in phenotypes of the skin, hair, wound repair, and malignancy [8,58–61]. It is more difficult to use transgenics that express tamoxifen-inducible CreER transgenes to induce skin phenotypes, in part because of variable efficiency of tamoxifen-induced activation, and in part because of the dynamic changes in expression of individual fibroblast marker genes [62]. Nonetheless, a thoroughly understood CreER line with specifically timed inductions with tamoxifen treatments can yield valuable information [8,37,63]. Indeed, Cre-mediated induction of the diptheria toxin suicide gene showed that reducing the number of fibroblasts during wounding can inhibit scar formation [64]. Other studies have shown that genetic ablation of Sox2 in the DP leads to a postnatal reduction in the number of awl/auchene hairs and a corresponding increase in zigzag hairs [65], as well as to a reduction in hair shaft length [58]. By contrast, DP expression of SOX18 is required for the formation of zigzag hair follicles [66]. Activation of b-catenin in the DP via Corin–Cre has been shown to regulate hair follicle morphogenesis and the hair follicle growth cycle, in addition to controlling hair pigmentation [59,60]. Although expression of Tbx18Cre and Prx1Cre is not confined to the developing DP, and also occurs in additional fibroblast populations, both Cre models have been used to manipulate genes that are specifically expressed in the dermal papilla
Trends in Cell Biology xxx xxxx, Vol. xxx, No. x
[63,67]. Such studies have shown that Tbx18 is not required for normal skin development and homeostasis. A key concept to emerge from studies of epidermal stem cell populations is that the properties they exhibit in undamaged epidermis may be different from those that they display in culture and following transplantation [11,68]. To some extent, this idea holds true for fibroblast subpopulations, although it is clear that different cell types can retain their identity, at least initially, in culture or following transplantation. Wnt-mediated epidermal regulation of dermal fibroblasts Fibroblasts can signal through various pathways to modulate their environment, and the environment itself can activate signalling pathways within fibroblasts to maintain skin homeostasis, stimulate wound healing, or promote tumorigenesis. Indeed, fibroblast derived activation of Yap, the key transcription factor of Hippo signalling, regulates the bulk stiffness of the ECM [69]. In addition, Notch signalling in fibroblasts regulates immune infiltration of connective tissue, as well as inflammation and ECM remodelling, which in turn regulates hair follicle morphogenesis and cancer progression, as shown by ablating immunoglobulin kappa J region recombination signal binding protein 1 (RBP-Jk) in fibroblasts in the skin [61,70]. In wound healing and scar formation, ECM deposition and contractility of myofibroblasts are mediated by transforming growth factor-beta (TGF-b) signalling [3]. Bmp/Noggin signals originating from keratinocytes, dermal papilla cells and adipocytes regulate hair follicle size and cycling [58,71,72]. In addition, reciprocal Wnt signalling between the epidermis and dermis has long been known to control hair follicle development and the hair growth cycle [51,53,55,73,74]. Epidermal Wnt ligands are required for hair follicle growth and adult wound-induced de novo hair formation [51,75]. In addition, recent studies have provided new insights into the profound effect of epidermal Wnt signalling on the adult dermis. A key effector of this pathway is bcatenin. When Wnt binds to cell surface receptors, b-catenin accumulates in the cytoplasm and translocates to the nucleus. There b-catenin activates transcription by binding to Lef/Tcf transcription factors. When b-catenin is activated via the keratin 14 promoter in all epidermal stem cell compartments, stem cell numbers increase and are accompanied by new hair follicle formation [56]. In addition to inducing new Sox2-negative DP [40], there is a general increase in fibroblast proliferation and extensive remodelling of the dermal ECM, such that the adult dermis acquires characteristics of the neonatal dermis [16]. The effects of epidermal b-catenin activation involve both local and long-range signalling, with an expansion of both the upper and lower dermal lineages [37]. An example of local signalling is deposition of the ECM protein nephronectin by epidermal stem cells in the hair follicle bulge. Fibroblasts expressing a8b1 integrin bind to nephronectin in the epidermal basement membrane. Specifically, as a Wnt target gene, when b-catenin is activated throughout the epidermal basal layer there is a corresponding upregulation of nephronectin and accumulation 5
TICB-1095; No. of Pages 8
Review of a8 integrin-positive cells in the dermis [17]. In addition, nephronectin promotes the differentiation of mesenchymal cells into APM, although the underlying mechanism remains to be elucidated. By contrast, the interaction between the epidermis and the hypodermis occurs over a longer range and is mediated, at least in part, by secreted factors such as bone morphogenetic proteins and insulin-like growth factors [23,71, 76]. The hair follicle growth cycle is synchronized with changes in the thickness of the adipocyte layer. Inhibition of epidermal b-catenin signalling reduces the size of the adipocyte layer, whereas its activation stimulates adipogenesis independent of hair follicle formation. Keratinocytederived adipogenic factors are induced by b-catenin activation and include ligands for the Bmp and Insulin pathways. Not only does the epidermis regulate the skin adipocyte layer, but adipocytes regulate the hair cycle. Intradermal pre-adipocyte precursors express high levels of PDGFA that stimulates hair follicle cycling by activating the dermal papilla [23]. In conclusion, activation of Wnt signalling in the epidermis leads to major effects in all dermal compartments and mediates those effects by a combination of short and long range signals. Wound repair, ageing, and cancer The functional heterogeneity of fibroblasts is not only important for skin homeostasis but also for wound healing. In addition it contributes to changes in the skin during ageing and disease. Lineage tracing studies have shown that the initial wave of dermal repair following wounding is mediated by lower lineage fibroblasts, which express myofibroblast markers such as a-SMA [37]. These cells secrete large amounts of ECM proteins such as collagen, and abnormal collagen deposition is a feature of scarring. By contrast, upper dermal fibroblasts are recruited during subsequent wound re-epithelialisation, during which they are required for hair follicle formation. The different responses between the upper and lower dermal fibroblast lineages towards wounding provide a powerful explanation for why scars are rich in fibrillar ECM and devoid of hair follicles. However, the activation of epidermal b-catenin can cause expansion of the upper dermal lineage prior to wound healing, ultimately resulting in the formation of hair follicles in the wound [37]. This finding is consistent with the observation that activation of Wnt in stem cells of the nail is required for regeneration of the nail or digit in neonatal mice [29,77]. It is of interest to discover the relative contributions of the different fibroblast lineages to changes in the dermis that occur during skin aging [78] and tumour formation in epithelia [79]. With age, the dermis becomes less elastic, and one potential reason, yet to be examined, is that it reflects a change in fibroblast types that populate aged skin versus young skin [69,78]. Indeed, in aged dermis, there is an expansion of the adipocyte layer, which might reflect a change in the relative abundance of reticular fibroblasts and pre-adipocytes. By contrast, increased ECM deposition and stiffness are common features of tumour stroma. One hallmark of cancer-associated fibroblasts (CAFs) is the upregulation of a-SMA, which could potentially reflect an expansion of the lower dermal fibroblast lineages in the skin [1]. 6
Trends in Cell Biology xxx xxxx, Vol. xxx, No. x
Concluding remarks In skin, the answer to the question ‘what is a fibroblast?’ is that a fibroblast represents a number of distinct differentiated mesenchymal cell types that have different origins, locations and functions. Key questions for the future are what pathways control selection of the different dermal lineages and whether our new knowledge of fibroblast heterogeneity can be of use clinically. Autologous fibroblast transplantation is already being evaluated to promote healing of acute wounds to stimulate healing of venous leg ulcers and to increase joint mobility that has been impaired by burn scars (www.clinicaltrials.gov). Allogeneic fibroblasts are being evaluated to repair scar contracture and as a therapy for epidermolysis bullosa (EB) [80]. Tumour stromal fibroblasts strongly influence tumour progression and, therefore, finding a way to target the tumour stroma could potentially decrease tumour growth [69,79]. The available evidence suggests that delivery of upper dermal fibroblasts could be beneficial in resolving scar formation and promoting scar-free wound healing, because they produce less fibrillar collagen than lower dermal fibroblasts. Conversely, lower dermal fibroblasts could have applications in breast reconstruction in mastectomy patients because of their ability to differentiate into adipocytes. There is also potential to explore the immunomodulatory properties of allogeneic fibroblast subpopulations, particularly given the large number of clinical trials that are underway in which bone marrow-derived mesenchymal stem/stromal cells (MSC) are being used in various indications [81]. References 1 Kalluri, R. and Zeisberg, M. (2006) Fibroblasts in cancer. Nat. Rev. Cancer 6, 392–401 2 Parsonage, G. et al. (2005) A stromal address code defined by fibroblasts. Trends Immunol. 26, 150–156 3 Tomasek, J.J. et al. (2002) Myofibroblasts and mechano-regulation of connective tissue remodelling. Nat. Rev. Mol. Cell Biol. 3, 349–363 4 Watt, F.M. and Fujiwara, H. (2011) Cell–extracellular matrix interactions in normal and diseased skin. Cold Spring Harb. Perspect. Biol. 1, 2011 Published online April 1, 2011, http://dx.doi.org/10.1101/ cshperspect.a005124 5 Driskell, R.R. et al. (2011) Hair follicle dermal papilla cells at a glance. J. Cell Sci. 124, 1179–1182 6 Fernandes, K.J. et al. (2004) A dermal niche for multipotent adult skinderived precursor cells. Nat. Cell Biol. 6, 1082–1093 7 Jinno, H. et al. (2010) Convergent genesis of an adult neural crest-like dermal stem cell from distinct developmental origins. Stem Cells 28, 2027–2040 8 Ohtola, J. et al. (2008) Beta-catenin has sequential roles in the survival and specification of ventral dermis. Development 135, 2321–2329 9 Rendl, M. et al. (2005) Molecular dissection of mesenchymal–epithelial interactions in the hair follicle. PLoS Biol. 3, e331 10 Wong, C.E. et al. (2006) Neural crest-derived cells with stem cell features can be traced back to multiple lineages in the adult skin. J. Cell Biol. 175, 1005–1015 11 Kretzschmar, K. and Watt, F.M. (2012) Lineage tracing. Cell 148, 33–45 12 Bianco, P. (2013) Reply to MSCs: science and trials. Nat. Med. 19, 813– 814 13 Sorrell, J.M. and Caplan, A.I. (2004) Fibroblast heterogeneity: more than skin deep. J. Cell Sci. 117, 667–675 14 Van Exan, R.J. and Hardy, M.H. (1984) The differentiation of the dermis in the laboratory mouse. Am. J. Anat. 169, 149–164 15 Bayreuther, K. et al. (1988) Human skin fibroblasts in vitro differentiate along a terminal cell lineage. Proc. Natl. Acad. Sci. U.S.A. 85, 5112–5116
TICB-1095; No. of Pages 8
Review 16 Collins, C.A. et al. (2011) Reprogramming adult dermis to a neonatal state through epidermal activation of beta-catenin. Development 138, 5189–5199 17 Fujiwara, H. et al. (2011) The basement membrane of hair follicle stem cells is a muscle cell niche. Cell 144, 577–589 18 Lee, J. and Tumbar, T. (2012) Hairy tale of signaling in hair follicle development and cycling. Semin. Cell Dev. Biol. 23, 906–916 19 Sennett, R. and Rendl, M. (2012) Mesenchymal–epithelial interactions during hair follicle morphogenesis and cycling. Semin. Cell Dev. Biol. 23, 917–927 20 Driskell, R.R. et al. (2009) Sox2-positive dermal papilla cells specify hair follicle type in mammalian epidermis. Development 136, 2815–2823 21 Biernaskie, J. et al. (2009) SKPs derive from hair follicle precursors and exhibit properties of adult dermal stem cells. Cell Stem Cell 5, 610–623 22 Toma, J.G. et al. (2001) Isolation of multipotent adult stem cells from the dermis of mammalian skin. Nat. Cell Biol. 3, 778–784 23 Festa, E. et al. (2011) Adipocyte lineage cells contribute to the skin stem cell niche to drive hair cycling. Cell 146, 761–771 24 Badiavas, E.V. et al. (2003) Participation of bone marrow derived cells in cutaneous wound healing. J. Cell. Physiol. 196, 245–250 25 Ishii, M. et al. (2005) Molecular markers distinguish bone marrow mesenchymal stem cells from fibroblasts. Biochem. Biophys. Res. Commun. 332, 297–303 26 Kataoka, K. et al. (2003) Participation of adult mouse bone marrow cells in reconstitution of skin. Am. J. Pathol. 163, 1227–1231 27 Barisic-Dujmovic, T. et al. (2010) Fibroblasts/myofibroblasts that participate in cutaneous wound healing are not derived from circulating progenitor cells. J. Cell. Physiol. 222, 703–712 28 Higashiyama, R. et al. (2011) Differential contribution of dermal resident and bone marrow-derived cells to collagen production during wound healing and fibrogenesis in mice. J. Invest. Dermatol. 131, 529–536 29 Rinkevich, Y. et al. (2011) Germ-layer and lineage-restricted stem/ progenitors regenerate the mouse digit tip. Nature 476, 409–413 30 Hall, A.P. (2006) Review of the pericyte during angiogenesis and its role in cancer and diabetic retinopathy. Toxicol. Pathol. 34, 763–775 31 Paquet-Fifield, S. et al. (2009) A role for pericytes as microenvironmental regulators of human skin tissue regeneration. J. Clin. Invest. 119, 2795–2806 32 Dellavalle, A. et al. (2011) Pericytes resident in postnatal skeletal muscle differentiate into muscle fibres and generate satellite cells. Nat. Commun. 2, 499 33 Rinn, J.L. et al. (2006) Anatomic demarcation by positional variation in fibroblast gene expression programs. PLoS Genet. 2, e119 34 Rinn, J.L. et al. (2008) A dermal HOX transcriptional program regulates site-specific epidermal fate. Genes Dev. 22, 303–307 35 Dhouailly, D. (1973) Dermo–epidermal interactions between birds and mammals: differentiation of cutaneous appendages. J. Embryol. Exp. Morphol. 30, 587–603 36 Gomez, C. et al. (2013) The interfollicular epidermis of adult mouse tail comprises two distinct cell lineages that are differentially regulated by Wnt, Edaradd, and Lrig1. Stem Cell Rep. 1, 19–27 37 Driskell, R.R. et al. (2013) Distinct fibroblast lineages determine dermal architecture in skin development and repair. Nature 504, 277–281 38 Janson, D.G. et al. (2012) Different gene expression patterns in human papillary and reticular fibroblasts. J. Invest. Dermatol. 132, 2565–2572 39 Harper, R.A. and Grove, G. (1979) Human skin fibroblasts derived from papillary and reticular dermis: differences in growth potential in vitro. Science 204, 526–527 40 Driskell, R.R. et al. (2012) Clonal growth of dermal papilla cells in hydrogels reveals intrinsic differences between Sox2-positive and negative cells in vitro and in vivo. J. Invest. Dermatol. 132, 1084–1093 41 Enshell-Seijffers, D. et al. (2008) The serine protease Corin is a novel modifier of the Agouti pathway. Development 135, 217–225 42 Collins, C.A. and Watt, F.M. (2008) Dynamic regulation of retinoic acidbinding proteins in developing, adult and neoplastic skin reveals roles for beta-catenin and Notch signalling. Dev. Biol. 324, 55–67 43 Johnston, A.P. et al. (2013) Sox2-mediated regulation of adult neural crest precursors and skin repair. Stem Cell Rep. 1, 38–45 44 Collins, C.A. et al. (2012) Polyclonal origin and hair induction ability of dermal papillae in neonatal and adult mouse back skin. Dev. Biol. 366, 290–297
Trends in Cell Biology xxx xxxx, Vol. xxx, No. x
45 Schmidt, B.A. and Horsley, V. (2013) Intradermal adipocytes mediate fibroblast recruitment during skin wound healing. Development 140, 1517–1527 46 Ambler, C.A. and Watt, F.M. (2010) Adult epidermal Notch activity induces dermal accumulation of T cells and neural crest derivatives through upregulation of jagged 1. Development 137, 3569–3579 47 Jensen, K.B. et al. (2010) Assaying proliferation and differentiation capacity of stem cells using disaggregated adult mouse epidermis. Nat. Protoc. 5, 898–911 48 Mulder, K.W. et al. (2012) Diverse epigenetic strategies interact to control epidermal differentiation. Nat. Cell Biol. 14, 753–763 49 Higgins, C.A. et al. (2013) Microenvironmental reprogramming by three-dimensional culture enables dermal papilla cells to induce de novo human hair-follicle growth. Proc. Natl. Acad. Sci. U.S.A. 110, 19679–19688 50 Horne, K.A. et al. (1986) Whisker growth induced by implantation of cultured vibrissa dermal papilla cells in the adult rat. J. Embryol. Exp. Morphol. 97, 111–124 51 Kishimoto, J. et al. (2000) Wnt signaling maintains the hair-inducing activity of the dermal papilla. Genes Dev. 14, 1181–1185 52 Lichti, U. et al. (1993) In vivo regulation of murine hair growth: insights from grafting defined cell populations onto nude mice. J. Invest. Dermatol. 101, 124S–129S 53 Ohyama, M. et al. (2010) The mesenchymal component of hair follicle neogenesis: background, methods and molecular characterization. Exp. Dermatol. 19, 89–99 54 Rendl, M. et al. (2008) BMP signaling in dermal papilla cells is required for their hair follicle-inductive properties. Genes Dev. 22, 543–557 55 Yang, C.C. and Cotsarelis, G. (2010) Review of hair follicle dermal cells. J. Dermatol. Sci. 57, 2–11 56 Silva-Vargas, V. et al. (2005) Beta-catenin and Hedgehog signal strength can specify number and location of hair follicles in adult epidermis without recruitment of bulge stem cells. Dev. Cell 9, 121–131 57 Wojciechowicz, K. et al. (2013) Development of the mouse dermal adipose layer occurs independently of subcutaneous adipose tissue and is marked by restricted early expression of FABP4. PLoS ONE 8, e59811 58 Clavel, C. et al. (2012) Sox2 in the dermal papilla niche controls hair growth by fine-tuning BMP signaling in differentiating hair shaft progenitors. Dev. Cell 23, 981–994 59 Enshell-Seijffers, D. et al. (2010) Beta-catenin activity in the dermal papilla regulates morphogenesis and regeneration of hair. Dev. Cell 18, 633–642 60 Enshell-Seijffers, D. et al. (2010) Beta-catenin activity in the dermal papilla of the hair follicle regulates pigment-type switching. Proc. Natl. Acad. Sci. U.S.A. 107, 21564–21569 61 Hu, B. et al. (2010) Control of hair follicle cell fate by underlying mesenchyme through a CSL–Wnt5a–FoxN1 regulatory axis. Genes Dev. 24, 1519–1532 62 Jeffery, E. et al. (2014) Characterization of Cre recombinase models for the study of adipose tissue. Adipocyte 3, 206–211 63 Grisanti, L. et al. (2013) Tbx18 targets dermal condensates for labeling, isolation, and gene ablation during embryonic hair follicle formation. J. Invest. Dermatol. 133, 344–353 64 Dulauroy, S. et al. (2012) Lineage tracing and genetic ablation of ADAM12(+) perivascular cells identify a major source of profibrotic cells during acute tissue injury. Nat. Med. 18, 1262–1270 65 Lesko, M.H. et al. (2013) Sox2 modulates the function of two distinct cell lineages in mouse skin. Dev. Biol. 382, 15–26 66 James, K. et al. (2003) Sox18 mutations in the ragged mouse alleles ragged-like and opossum. Genesis 36, 1–6 67 Woo, W.M. et al. (2012) Shh maintains dermal papilla identity and hair morphogenesis via a Noggin–Shh regulatory loop. Genes Dev. 26, 1235–1246 68 Watt, F.M. and Jensen, K.B. (2009) Epidermal stem cell diversity and quiescence. EMBO Mol. Med. 1, 260–267 69 Calvo, F. et al. (2013) Mechanotransduction and YAP-dependent matrix remodelling is required for the generation and maintenance of cancer-associated fibroblasts. Nat. Cell Biol. 15, 637–646 70 Hu, B. et al. (2012) Multifocal epithelial tumors and field cancerization from loss of mesenchymal CSL signaling. Cell 149, 1207–1220 71 Plikus, M.V. et al. (2011) Self-organizing and stochastic behaviors during the regeneration of hair stem cells. Science 332, 586–589 7
TICB-1095; No. of Pages 8
Review 72 Sharov, A.A. et al. (2006) Bone morphogenetic protein signaling regulates the size of hair follicles and modulates the expression of cell cycle-associated genes. Proc. Natl. Acad. Sci. U.S.A. 103, 18166– 18171 73 Millar, S.E. (2002) Molecular mechanisms regulating hair follicle development. J. Invest. Dermatol. 118, 216–225 74 Schneider, M.R. et al. (2009) The hair follicle as a dynamic miniorgan. Curr. Biol. 19, R132–R142 75 Myung, P.S. et al. (2013) Epithelial Wnt ligand secretion is required for adult hair follicle growth and regeneration. J. Invest. Dermatol. 133, 31–41 76 Donati, G. et al. (2014) Epidermal Wnt/beta-catenin signaling regulates adipocyte differentiation via secretion of adipogenic factors. Proc. Natl. Acad. Sci. U.S.A. 111, E1501–E1509 77 Takeo, M. et al. (2013) Wnt activation in nail epithelium couples nail growth to digit regeneration. Nature 499, 228–232
8
Trends in Cell Biology xxx xxxx, Vol. xxx, No. x
78 Giangreco, A. et al. (2008) Epidermal stem cells are retained in vivo throughout skin aging. Aging Cell 7, 250–259 79 Hanahan, D. and Weinberg, R.A. (2011) Hallmarks of cancer: the next generation. Cell 144, 646–674 80 Petrof, G. et al. (2013) Fibroblast cell therapy enhances initial healing in recessive dystrophic epidermolysis bullosa wounds: results of a randomized, vehicle-controlled trial. Br. J. Dermatol. 169, 1025–1033 81 Jones, E. and McGonagle, D. (2008) Human bone marrow mesenchymal stem cells in vivo. Rheumatology 47, 126–131 82 Pittenger, M.F. et al. (1999) Multilineage potential of adult human mesenchymal stem cells. Science 284, 143–147 83 Handjiski, B.K. et al. (1994) Alkaline phosphatase activity and localization during the murine hair cycle. Br. J. Dermatol. 131, 303–310 84 Egawa, G. et al. (2009) Transient expression of ephrin b2 in perinatal skin is required for maintenance of keratinocyte homeostasis. J. Invest. Dermatol. 129, 2386–2395
Review
Understanding fibroblast heterogeneity in the skin Ryan R. Driskell and Fiona M. Watt Centre for Stem Cells and Regenerative Medicine, King’s College London, 28th Floor, Tower Wing, Guy’s Hospital Campus, London SE1 9RT, UK
Fibroblasts are found in most tissues, yet they remain poorly characterised. Different fibroblast subpopulations with distinct functions have been identified in the skin. This functional heterogeneity reflects the varied fibroblast lineages that arise from a common embryonic precursor. In addition to autocrine signals, fibroblasts are highly responsive to Wnt-regulated signals from the overlying epidermis, which can act both locally, via extracellular matrix (ECM) deposition, and via secreted factors that impact the behaviour of fibroblasts in different dermal locations. These findings may explain some of the changes that occur in connective tissue during wound healing and cancer progression. Varied origins and functions of fibroblasts Fibroblasts are mesenchymal cells that deposit collagen and elastic fibres of the ECM in connective tissue [1–4]. This simple operational definition, however, masks the considerable heterogeneity of fibroblasts found in different tissues (healthy or diseased) and at different stages of development. Indeed, fibroblasts in different body sites arise from different embryonic origins [5]. For example, fibroblasts in face skin arise from the neural crest; those in ventral body skin derive from the lateral plate mesoderm; and those in back skin originate from the dermomyotome [6–10] (Figure 1). While genetic tools for lineage tracing have led to considerable progress in elucidating stem cells and lineage relationships in most cell types and tissues [11], characterisation of fibroblasts has lagged. This is, in part, because of the challenge of defining clonal relationships when the progeny of an individual cell do not remain attached to one another, and in part because of a lack of defined markers to distinguish different fibroblast subsets. It is important to consider the lineage relationships of fibroblasts because this may influence their differentiation potential [7]. For example, mesenchymal stem cells from the bone marrow share similar properties with tissue resident fibroblasts; however, just because different cell populations express common markers does not mean they have a common origin [12]. Nevertheless, recent advances in marker identification, functional assays, and lineage tracing have led to new insights into fibroblasts in the skin. These findings reveal that functional heterogeneity Corresponding author: Watt, F.M. ([email protected]). Keywords: dermis; connective tissue; stem cells; lineage; Wnt; adipocytes. 0962-8924/ ß 2014 Published by Elsevier Ltd. http://dx.doi.org/10.1016/j.tcb.2014.10.001
reflects, at least in part, the existence of different fibroblast lineages, and that fibroblasts respond readily to signals from the overlying epidermis and thereby exhibit surprising plasticity. These observations provide new ways of interpreting the dynamic changes in fibroblast behaviour – such as proliferation and migration – observed in tissue repair and disease. Mesenchymal cells of skin connective tissue The dermis, the connective tissue underlying the epidermis, provides a good example of how even within one tissue, in a single body site, and at a single developmental stage, fibroblasts are remarkably diverse [13–15] (Figure 2; Tables 1 and 2). In neonatal (P2) mouse back skin, the upper (papillary) dermis is distinguished from the lower (reticular) dermis by differences in fibroblast density and by the maturity of the collagenous ECM [16]. Signals from epidermal stem cells in the hair follicle bulge induce adjacent fibroblasts to form the smooth muscle known as the arrector pili muscle (APM), which controls piloerection [17]. A cluster of fibroblasts at the base of the hair follicle, known as the dermal papilla (DP), has specialized signalling properties required for hair follicle morphogenesis and coordination of the hair cycle [18,19]. The dermal papillae that are associated with different types of hair follicles are distinct [20] and there is evidence that some dermal papilla cells have the ability to differentiate into a wide range of cell types, including nerve and cartilage [21,22]. Further heterogeneity is evident in the region underlying the reticular dermis, known as the hypodermis or dermal white adipose tissue, since this contains a mixture of pre-adipocytes and mature adipocytes [23]. Two other skin mesenchymal cell types that warrant consideration are perivascular smooth muscle cells (pericytes) and mesenchymal stem cells (MSCs), both of which express some markers in common with fibroblasts (Table 1). MSCs form part of the perivascular stromal compartment of the bone marrow and can be isolated from bone marrow cell preparations as the cells that adhere to tissue culture plastic, in contrast to the non-adherent haemopoietic cells [12]. Although there have been some reports that MSCs contribute to the dermis [24–26], more recent studies, involving lineage tracing and bone marrow transplantation, suggest that this is not the case [27–29]. Perivascular smooth muscle cells surround blood vessel endothelial cells and have a contractile function that regulates endothelial cell homeostasis [30]. There are no Trends in Cell Biology xx (2014) 1–8
1
TICB-1095; No. of Pages 8
Review
Trends in Cell Biology xxx xxxx, Vol. xxx, No. x
ers Whisk
blast F i b r o aon ren diffe lineage and itment m com
Neural crest fibroblast precursor
Upper lineage Lower lineage
Facial skin last Fibrob on na differe eage and lin ent itm comm
Somite fibroblast precursor
Hair follic
Upper lineage Lower lineage
les
Back skin
TRENDS in Cell Biology
Figure 1. Developmental origins of dermal fibroblasts. Fibroblasts arise from different developmental origins such as the neural crest, dermomyotome, and lateral plate mesoderm. Regardless of the developmental origin of fibroblasts, they undergo differentiation and lineage commitment to give rise to both upper and lower lineages. Redrawn from [5].
definitive markers of pericytes, but the field has utilized asmooth muscle actin (a-SMA), desmin, and neuron–glial antigen 2 [NG2; also known as chondroitin sulfate proteoglycan 4 (CSPG4)] as markers [31]. In skeletal muscle, pericytes can differentiate into muscle fibres [32]. In skin, pericytes have been identified in the upper dermis and can be isolated with an HD-1 antibody [31]. However, there is currently no evidence that they contribute to the skin fibroblast compartment. When fibroblasts are isolated from adult skin and placed in culture or transplanted to a new location they exhibit
Papillary dermis
positional memory, which is reflected in Homeobox (Hox) gene expression [33,34]. Classic tissue recombination experiments have shown differences in the ability of the dermis from hair-bearing and non-hair-bearing regions of mouse skin to induce hair follicle formation [35]. While positional memory likely accounts for the different types of hair follicles in different body sites, fibroblasts arising from the neural crest, lateral plate mesoderm, and dermomyotome are all competent to support hair follicle formation. Indeed, lineage tracing and cell isolation experiments have shown that dermal cells can exhibit characteristics of
Key: r1 Lg 6 ig Lr
Recular dermis
Upper lineage Dermal papilla Papillary fibroblast Arrector pili
r5 Lg
Dermal sheath
Lower lineage Recular fibroblast Hypodermis
Pre-adipocyte Adipocyte
TRENDS in Cell Biology
Figure 2. Heterogeneity of epidermal stem cells and mesenchymal cells in skin. The locations of three epidermal stem cell populations (Lrig1+, Lgr6+, and Lgr5+) in the hair follicle are shown. The three dermal layers are also indicated: the papillary dermis, reticular dermis and hypodermis/white adipose layer. The specialised fibroblasts of the dermal papilla (DP) and arrector pili muscle (AP) are derived from the same lineage as the papillary dermis, while reticular fibroblasts, pre-adipocytes and adipocytes share a common origin.
2
TICB-1095; No. of Pages 8
Review
Trends in Cell Biology xxx xxxx, Vol. xxx, No. x
Table 1. Mesenchymal cell subpopulations in the skin Cell type Papillary fibroblast
Location Papillary dermis, adjacent to epidermis
Reticular fibroblast Intradermal pre-adipocyte Intradermal adipocyte
Reticular dermis Reticular dermis
Myofibroblast Arrector pili muscle Dermal papilla Mesenchymal stem cell Perivascular cell
Wound/scar Attachment to hair follicle bulge and IFE Attachment to hair follicle bulb Bone marrow Attachment to blood vessels of skin
Dermal white adipose tissue (hypodermis)
neural crest cell derivatives even though they have a nonneural crest origin [7]. However, the extent to which Hox genes regulate local fibroblast positional identity, such as the different characteristics of fibroblasts underlying the scale and interscale interfollicular epidermis in mouse tail dermis, remains to be determined [36]. Markers and lineage relationships of dermal fibroblast subpopulations In order to understand fibroblast heterogeneity in the skin, it is vital to have a set of markers that can be used to define various subpopulations of cells. Several pan-fibroblast markers have been well studied, such as the intermediate filament protein vimentin, the collagen 1a2 chain and Pdgfra, the cellular receptor for platelet-derived growth factor subunit A (PDGFA) [1]. However, these markers are also expressed by other cells in the body, such as endothelial and myoepithelial cells in the case of vimentin, and osteoblasts and chondroblasts in the case of collagen 1a2, making it difficult to accurately identify and trace fibroblast populations through development. Recent lineage tracing of the dermis at different stages of development has provided new insight into fibroblast origins. Specifically, the analysis of mouse back skin at different stages of development, from E12.5, when the dermis appears as a homogeneous tissue, to P2, when all the dermal regions have formed [37], has revealed a remarkably dynamic pattern of marker expression. Recent lineage tracing has revealed that dermal fibroblasts arise from a multipotent mesenchymal cell population at E12.5, which expresses Pdgfra, delta-like homolog 1 (Dlk1), and the leucine-rich repeat protein Lrig1 (Figure 3). Upon differentiation, the fibroblasts begin to express
Function Source of AP, IFE support, ECM production ECM production ECM production, source of adipocytes, long distance signalling Long distance epidermal stem cell signalling, insulation Wound contraction, crosstalk with immune cells Piloerection Hair follicle morphogenesis, hair cycle Not detected consistently in the dermis Vascular constriction and homeostasis
Refs [13,37] [13] [23,76] [23,76] [3] [17] [19,55] [82] [30]
different markers distinguishing common progenitor fibroblasts from papillary dermal fibroblasts of the upper dermis and reticular dermal fibroblasts of the lower dermis. This lineage commitment between upper and lower dermal fibroblasts was observed to occur at approximately E16.5. Indeed, the a8 integrin subunit, cluster of differentiation 26 [CD26; or dipeptidyl peptidase 4 (Dpp4)], Lrig1, and B lymphocyte-induced maturation protein 1 (Blimp1) were expressed in the papillary dermis during early development, which was later confirmed by lineage tracing using Blimp1Cre and Lrig1CreER to label the upper dermal lineage. Furthermore, lineage tracing using Dlk1CreER to label dermal fibroblasts at E12.5 revealed that, by E16.5, Dlk1 (Pref1) was selective for the lower dermis. Dlk1 and Sca1 together were dynamically expressed in the reticular and hypodermis during early skin morphogenesis [37]. Several markers are also differentially expressed in upper (papillary) and lower (reticular) dermis of the human skin [13;37–39]. In addition to the lower and upper dermis, common progenitor fibroblasts can also differentiate into dermal papilla or give rise to papillary dermal fibroblasts and arrector pili muscle cells. DP cells express a number of distinct markers [9,20,40,41]. Gene expression profiling indicates DP markers change according to hair follicle type, stage of hair growth cycle, and stage of development. Alkaline phosphatase and cellular retinoic-acid-binding protein 1 (CRABP1) are expressed in DP cells throughout the hair growth cycle [42], although alkaline phosphatase expression is not confined to fibroblasts with hair inductive properties [40]. Sox2 expression is specifically expressed in the DP of guard, awl, and auchene hairs [20,21]. In addition, lineage analysis of subsets of dermal papillae from
Table 2. Stable and dynamic markers of dermal papilla, papillary and reticular dermis, fat and arrector pili muscle at E16.5, P2, and adult Cell type Pan-fibroblast Dermal papilla Papillary dermis Reticular dermis/hypodermis Arrector Pili Dermal white pre-adipocytes Dermal white adipocytes
Stable marker Col1a2, PDGFRa Crabp1, AlkPhos
Dynamic marker All markers listed below Corin, Sox2, Sox18 Blimp1, Itga8, CD26, Lrig1, EphB2, Trps1 Sca1, Podoplanin, Dlk1
Acta2, Itga8, Sm22a, Npnt CD34, Sca1, PDGFRa
Refs [16] [20,41,42,83] [37;57,84] [23,37,38] [17] [23,37]
Perilipin, adiponectin
[23] 3
TICB-1095; No. of Pages 8
Review
Trends in Cell Biology xxx xxxx, Vol. xxx, No. x
E14.5
Dermal papilla PDGFRa+ CD26– Sox2+
E12.5
Common fibroblast progenitor PDGFRa+ Dlk1+ lrig1+
P2
E16.5
Papillary dermal mal fibroblast progenitor PDGFRa+ Blimp1+ Dlk1– lrig1+ E16.5
Recular dermal mal fibroblast progenitor PDGFRa+ Blimp1– Dlk1+
Dermal papilla D PDGFRa+ CD26– Sox2– P2 Papillar dermal fibroblast PDGFRa+ CD26+ Blimp1– lrig1+
Recular dermal mal fibroblast progenitor PDGFRa+ Blimp1– Dlk1+
APM PDGFRa+ Itga8+ CD26–
Recular fibroblast R oblast PDGFRa+ Dlk1+ Sca1–
Hypodermal fibroblast PDGFRa+ Dlk1+/– Sca1+
Hypodermal adipocyte Hy PDGFRa– Dlk1– lipid+
TRENDS in Cell Biology
Figure 3. Temporal origins and lineage relationships of fibroblasts in mouse back skin. Fibroblasts arise from a common progenitor and progressively differentiate into upper fibroblasts lineages (green) and lower fibroblast lineages (red). Lineage markers are transient and labelled at each time point. Re-drawn from [37].
guard/awl/auchene hair with Sox2CreER revealed that Sox2 fibroblasts do not significantly contribute to wound repair in skin [43]. Additional insights into mesenchymal subpopulations in the skin have been obtained by performing lineage tracing of DP cells. Lineage tracing using Corin–Cre has shown that the progeny of cells that express Corin remain in the dermal papilla and are not found elsewhere in the dermis [41]. Studies in chimeric mice and skin reconstitution assays have shown that dermal papillae have a polyclonal origin and that DP cells can participate in forming new hair follicles without proliferating [44]. Furthermore, lineage tracing of pre-adipocytes using adiponectin–Cre to efficiently label intradermal adipocyte lineages has revealed that adipocyte precursor cells proliferate, while mature adipocytes repopulate skin wounds in parallel with fibroblast migration [45]. As well as providing information about lineage relationships within the dermis, lineage tracing experiments showed that epidermal keratinocytes do not undergo conversion into dermal cells, even under conditions in which keratinocytes are induced to express the epithelial-to-mesenchymal (EMT) marker Slug1 [46]. Similarly, bone marrow derived mesenchymal cells were not found to contribute to the dermis [29,37]. Despite the many advances that have been made by using specific cell surface markers to isolate and characterise fibroblast subpopulations, it is important to note 4
that markers that are specific for one fibroblast population at one time point can be either downregulated or expressed on a different population a few days later. In addition, the lineage tracing approach has caveats, in particular efficiency of labelling and leakiness. Functional heterogeneity of dermal fibroblasts Dermal fibroblasts not only differ in location and gene expression, but also in function. This has been demonstrated in a number of different ways: in cell culture, in cell transplantation experiments and by genetic gain and loss of function approaches. As in the case of epidermal cells, fibroblast behaviour and differentiation can be measured both on the population level and by characterising clones derived from single cells [47,48]. In bulk cultures, it is possible to measure overall proliferation rates, production and responsiveness to growth factors, and differences in cell morphology [13]. Clonal growth of DP cells in hanging drops or captured in hydrogels has shown that hair induction ability can be retained, both in mouse and human, at least at early passages [40,49–55]. Indeed, Sox2-positive cells retain their identity in culture, suggesting that they represent a distinct differentiated lineage [40]. By contrast, upper dermal fibroblasts isolated from E16.5 are competent to differentiate into adipocytes in the presence of adipogenic medium in hydrogel cultures. By P2, however, upper dermal fibroblasts
TICB-1095; No. of Pages 8
Review do not differentiate into adipocytes in culture because lineage restriction has occurred [37]. Mouse skin can be reconstituted from disaggregated epidermal and dermal cells by injecting them into a chamber implanted into a full thickness wound on the back of a syngeneic or immunodeficient mouse [31,37,47]. By sorting specific subpopulations of cells from Pdgfra–Histone-2Benhanced green fluorescent protein (EGFP) expressing dermis, it is possible to follow their fate when combined with an excess of unlabelled and unfractionated epidermal and dermal cells [44]. Transplantation studies have shown that fibroblast proliferation is necessary for new hair follicle formation, but cells that have been irradiated to inhibit proliferation can still contribute to new DP. Furthermore, although new hair follicles do not normally develop postnatally, adult dermal fibroblasts retain the ability to contribute to DP during hair follicle formation in transplantation assays [44] and in response to epidermal b-catenin activation [56]. In skin reconstitution assays, P2 upper dermal fibroblasts (CD26+Sca1–) only give rise to papillary fibroblasts, APM and DP, whereas lower dermal fibroblasts that express Sca1 are restricted to forming the reticular dermis and adipocytes. A multipotent population that is Dlk1+Sca1– can give rise to all the dermal mesenchymal cell types, but cells that express this combination of markers do not persist in adult skin [37]. Lin-CD34+CD29+Sca1+ adipocyte precursors stimulate local hair follicle cycling when injected intradermally [23]. When E14.5 skin was implanted onto kidney capsules it was found that adipocytes formed independently of hair follicle morphogenesis [57]. The availability of Cre mouse lines that target different dermal cell populations has facilitated the analysis of fibroblast function via genetic modification. The manipulation of gene expression via a variety of promoters that express noninducible forms of Cre-recombinase results in phenotypes of the skin, hair, wound repair, and malignancy [8,58–61]. It is more difficult to use transgenics that express tamoxifen-inducible CreER transgenes to induce skin phenotypes, in part because of variable efficiency of tamoxifen-induced activation, and in part because of the dynamic changes in expression of individual fibroblast marker genes [62]. Nonetheless, a thoroughly understood CreER line with specifically timed inductions with tamoxifen treatments can yield valuable information [8,37,63]. Indeed, Cre-mediated induction of the diptheria toxin suicide gene showed that reducing the number of fibroblasts during wounding can inhibit scar formation [64]. Other studies have shown that genetic ablation of Sox2 in the DP leads to a postnatal reduction in the number of awl/auchene hairs and a corresponding increase in zigzag hairs [65], as well as to a reduction in hair shaft length [58]. By contrast, DP expression of SOX18 is required for the formation of zigzag hair follicles [66]. Activation of b-catenin in the DP via Corin–Cre has been shown to regulate hair follicle morphogenesis and the hair follicle growth cycle, in addition to controlling hair pigmentation [59,60]. Although expression of Tbx18Cre and Prx1Cre is not confined to the developing DP, and also occurs in additional fibroblast populations, both Cre models have been used to manipulate genes that are specifically expressed in the dermal papilla
Trends in Cell Biology xxx xxxx, Vol. xxx, No. x
[63,67]. Such studies have shown that Tbx18 is not required for normal skin development and homeostasis. A key concept to emerge from studies of epidermal stem cell populations is that the properties they exhibit in undamaged epidermis may be different from those that they display in culture and following transplantation [11,68]. To some extent, this idea holds true for fibroblast subpopulations, although it is clear that different cell types can retain their identity, at least initially, in culture or following transplantation. Wnt-mediated epidermal regulation of dermal fibroblasts Fibroblasts can signal through various pathways to modulate their environment, and the environment itself can activate signalling pathways within fibroblasts to maintain skin homeostasis, stimulate wound healing, or promote tumorigenesis. Indeed, fibroblast derived activation of Yap, the key transcription factor of Hippo signalling, regulates the bulk stiffness of the ECM [69]. In addition, Notch signalling in fibroblasts regulates immune infiltration of connective tissue, as well as inflammation and ECM remodelling, which in turn regulates hair follicle morphogenesis and cancer progression, as shown by ablating immunoglobulin kappa J region recombination signal binding protein 1 (RBP-Jk) in fibroblasts in the skin [61,70]. In wound healing and scar formation, ECM deposition and contractility of myofibroblasts are mediated by transforming growth factor-beta (TGF-b) signalling [3]. Bmp/Noggin signals originating from keratinocytes, dermal papilla cells and adipocytes regulate hair follicle size and cycling [58,71,72]. In addition, reciprocal Wnt signalling between the epidermis and dermis has long been known to control hair follicle development and the hair growth cycle [51,53,55,73,74]. Epidermal Wnt ligands are required for hair follicle growth and adult wound-induced de novo hair formation [51,75]. In addition, recent studies have provided new insights into the profound effect of epidermal Wnt signalling on the adult dermis. A key effector of this pathway is bcatenin. When Wnt binds to cell surface receptors, b-catenin accumulates in the cytoplasm and translocates to the nucleus. There b-catenin activates transcription by binding to Lef/Tcf transcription factors. When b-catenin is activated via the keratin 14 promoter in all epidermal stem cell compartments, stem cell numbers increase and are accompanied by new hair follicle formation [56]. In addition to inducing new Sox2-negative DP [40], there is a general increase in fibroblast proliferation and extensive remodelling of the dermal ECM, such that the adult dermis acquires characteristics of the neonatal dermis [16]. The effects of epidermal b-catenin activation involve both local and long-range signalling, with an expansion of both the upper and lower dermal lineages [37]. An example of local signalling is deposition of the ECM protein nephronectin by epidermal stem cells in the hair follicle bulge. Fibroblasts expressing a8b1 integrin bind to nephronectin in the epidermal basement membrane. Specifically, as a Wnt target gene, when b-catenin is activated throughout the epidermal basal layer there is a corresponding upregulation of nephronectin and accumulation 5
TICB-1095; No. of Pages 8
Review of a8 integrin-positive cells in the dermis [17]. In addition, nephronectin promotes the differentiation of mesenchymal cells into APM, although the underlying mechanism remains to be elucidated. By contrast, the interaction between the epidermis and the hypodermis occurs over a longer range and is mediated, at least in part, by secreted factors such as bone morphogenetic proteins and insulin-like growth factors [23,71, 76]. The hair follicle growth cycle is synchronized with changes in the thickness of the adipocyte layer. Inhibition of epidermal b-catenin signalling reduces the size of the adipocyte layer, whereas its activation stimulates adipogenesis independent of hair follicle formation. Keratinocytederived adipogenic factors are induced by b-catenin activation and include ligands for the Bmp and Insulin pathways. Not only does the epidermis regulate the skin adipocyte layer, but adipocytes regulate the hair cycle. Intradermal pre-adipocyte precursors express high levels of PDGFA that stimulates hair follicle cycling by activating the dermal papilla [23]. In conclusion, activation of Wnt signalling in the epidermis leads to major effects in all dermal compartments and mediates those effects by a combination of short and long range signals. Wound repair, ageing, and cancer The functional heterogeneity of fibroblasts is not only important for skin homeostasis but also for wound healing. In addition it contributes to changes in the skin during ageing and disease. Lineage tracing studies have shown that the initial wave of dermal repair following wounding is mediated by lower lineage fibroblasts, which express myofibroblast markers such as a-SMA [37]. These cells secrete large amounts of ECM proteins such as collagen, and abnormal collagen deposition is a feature of scarring. By contrast, upper dermal fibroblasts are recruited during subsequent wound re-epithelialisation, during which they are required for hair follicle formation. The different responses between the upper and lower dermal fibroblast lineages towards wounding provide a powerful explanation for why scars are rich in fibrillar ECM and devoid of hair follicles. However, the activation of epidermal b-catenin can cause expansion of the upper dermal lineage prior to wound healing, ultimately resulting in the formation of hair follicles in the wound [37]. This finding is consistent with the observation that activation of Wnt in stem cells of the nail is required for regeneration of the nail or digit in neonatal mice [29,77]. It is of interest to discover the relative contributions of the different fibroblast lineages to changes in the dermis that occur during skin aging [78] and tumour formation in epithelia [79]. With age, the dermis becomes less elastic, and one potential reason, yet to be examined, is that it reflects a change in fibroblast types that populate aged skin versus young skin [69,78]. Indeed, in aged dermis, there is an expansion of the adipocyte layer, which might reflect a change in the relative abundance of reticular fibroblasts and pre-adipocytes. By contrast, increased ECM deposition and stiffness are common features of tumour stroma. One hallmark of cancer-associated fibroblasts (CAFs) is the upregulation of a-SMA, which could potentially reflect an expansion of the lower dermal fibroblast lineages in the skin [1]. 6
Trends in Cell Biology xxx xxxx, Vol. xxx, No. x
Concluding remarks In skin, the answer to the question ‘what is a fibroblast?’ is that a fibroblast represents a number of distinct differentiated mesenchymal cell types that have different origins, locations and functions. Key questions for the future are what pathways control selection of the different dermal lineages and whether our new knowledge of fibroblast heterogeneity can be of use clinically. Autologous fibroblast transplantation is already being evaluated to promote healing of acute wounds to stimulate healing of venous leg ulcers and to increase joint mobility that has been impaired by burn scars (www.clinicaltrials.gov). Allogeneic fibroblasts are being evaluated to repair scar contracture and as a therapy for epidermolysis bullosa (EB) [80]. Tumour stromal fibroblasts strongly influence tumour progression and, therefore, finding a way to target the tumour stroma could potentially decrease tumour growth [69,79]. The available evidence suggests that delivery of upper dermal fibroblasts could be beneficial in resolving scar formation and promoting scar-free wound healing, because they produce less fibrillar collagen than lower dermal fibroblasts. Conversely, lower dermal fibroblasts could have applications in breast reconstruction in mastectomy patients because of their ability to differentiate into adipocytes. There is also potential to explore the immunomodulatory properties of allogeneic fibroblast subpopulations, particularly given the large number of clinical trials that are underway in which bone marrow-derived mesenchymal stem/stromal cells (MSC) are being used in various indications [81]. References 1 Kalluri, R. and Zeisberg, M. (2006) Fibroblasts in cancer. Nat. Rev. Cancer 6, 392–401 2 Parsonage, G. et al. (2005) A stromal address code defined by fibroblasts. Trends Immunol. 26, 150–156 3 Tomasek, J.J. et al. (2002) Myofibroblasts and mechano-regulation of connective tissue remodelling. Nat. Rev. Mol. Cell Biol. 3, 349–363 4 Watt, F.M. and Fujiwara, H. (2011) Cell–extracellular matrix interactions in normal and diseased skin. Cold Spring Harb. Perspect. Biol. 1, 2011 Published online April 1, 2011, http://dx.doi.org/10.1101/ cshperspect.a005124 5 Driskell, R.R. et al. (2011) Hair follicle dermal papilla cells at a glance. J. Cell Sci. 124, 1179–1182 6 Fernandes, K.J. et al. (2004) A dermal niche for multipotent adult skinderived precursor cells. Nat. Cell Biol. 6, 1082–1093 7 Jinno, H. et al. (2010) Convergent genesis of an adult neural crest-like dermal stem cell from distinct developmental origins. Stem Cells 28, 2027–2040 8 Ohtola, J. et al. (2008) Beta-catenin has sequential roles in the survival and specification of ventral dermis. Development 135, 2321–2329 9 Rendl, M. et al. (2005) Molecular dissection of mesenchymal–epithelial interactions in the hair follicle. PLoS Biol. 3, e331 10 Wong, C.E. et al. (2006) Neural crest-derived cells with stem cell features can be traced back to multiple lineages in the adult skin. J. Cell Biol. 175, 1005–1015 11 Kretzschmar, K. and Watt, F.M. (2012) Lineage tracing. Cell 148, 33–45 12 Bianco, P. (2013) Reply to MSCs: science and trials. Nat. Med. 19, 813– 814 13 Sorrell, J.M. and Caplan, A.I. (2004) Fibroblast heterogeneity: more than skin deep. J. Cell Sci. 117, 667–675 14 Van Exan, R.J. and Hardy, M.H. (1984) The differentiation of the dermis in the laboratory mouse. Am. J. Anat. 169, 149–164 15 Bayreuther, K. et al. (1988) Human skin fibroblasts in vitro differentiate along a terminal cell lineage. Proc. Natl. Acad. Sci. U.S.A. 85, 5112–5116
TICB-1095; No. of Pages 8
Review 16 Collins, C.A. et al. (2011) Reprogramming adult dermis to a neonatal state through epidermal activation of beta-catenin. Development 138, 5189–5199 17 Fujiwara, H. et al. (2011) The basement membrane of hair follicle stem cells is a muscle cell niche. Cell 144, 577–589 18 Lee, J. and Tumbar, T. (2012) Hairy tale of signaling in hair follicle development and cycling. Semin. Cell Dev. Biol. 23, 906–916 19 Sennett, R. and Rendl, M. (2012) Mesenchymal–epithelial interactions during hair follicle morphogenesis and cycling. Semin. Cell Dev. Biol. 23, 917–927 20 Driskell, R.R. et al. (2009) Sox2-positive dermal papilla cells specify hair follicle type in mammalian epidermis. Development 136, 2815–2823 21 Biernaskie, J. et al. (2009) SKPs derive from hair follicle precursors and exhibit properties of adult dermal stem cells. Cell Stem Cell 5, 610–623 22 Toma, J.G. et al. (2001) Isolation of multipotent adult stem cells from the dermis of mammalian skin. Nat. Cell Biol. 3, 778–784 23 Festa, E. et al. (2011) Adipocyte lineage cells contribute to the skin stem cell niche to drive hair cycling. Cell 146, 761–771 24 Badiavas, E.V. et al. (2003) Participation of bone marrow derived cells in cutaneous wound healing. J. Cell. Physiol. 196, 245–250 25 Ishii, M. et al. (2005) Molecular markers distinguish bone marrow mesenchymal stem cells from fibroblasts. Biochem. Biophys. Res. Commun. 332, 297–303 26 Kataoka, K. et al. (2003) Participation of adult mouse bone marrow cells in reconstitution of skin. Am. J. Pathol. 163, 1227–1231 27 Barisic-Dujmovic, T. et al. (2010) Fibroblasts/myofibroblasts that participate in cutaneous wound healing are not derived from circulating progenitor cells. J. Cell. Physiol. 222, 703–712 28 Higashiyama, R. et al. (2011) Differential contribution of dermal resident and bone marrow-derived cells to collagen production during wound healing and fibrogenesis in mice. J. Invest. Dermatol. 131, 529–536 29 Rinkevich, Y. et al. (2011) Germ-layer and lineage-restricted stem/ progenitors regenerate the mouse digit tip. Nature 476, 409–413 30 Hall, A.P. (2006) Review of the pericyte during angiogenesis and its role in cancer and diabetic retinopathy. Toxicol. Pathol. 34, 763–775 31 Paquet-Fifield, S. et al. (2009) A role for pericytes as microenvironmental regulators of human skin tissue regeneration. J. Clin. Invest. 119, 2795–2806 32 Dellavalle, A. et al. (2011) Pericytes resident in postnatal skeletal muscle differentiate into muscle fibres and generate satellite cells. Nat. Commun. 2, 499 33 Rinn, J.L. et al. (2006) Anatomic demarcation by positional variation in fibroblast gene expression programs. PLoS Genet. 2, e119 34 Rinn, J.L. et al. (2008) A dermal HOX transcriptional program regulates site-specific epidermal fate. Genes Dev. 22, 303–307 35 Dhouailly, D. (1973) Dermo–epidermal interactions between birds and mammals: differentiation of cutaneous appendages. J. Embryol. Exp. Morphol. 30, 587–603 36 Gomez, C. et al. (2013) The interfollicular epidermis of adult mouse tail comprises two distinct cell lineages that are differentially regulated by Wnt, Edaradd, and Lrig1. Stem Cell Rep. 1, 19–27 37 Driskell, R.R. et al. (2013) Distinct fibroblast lineages determine dermal architecture in skin development and repair. Nature 504, 277–281 38 Janson, D.G. et al. (2012) Different gene expression patterns in human papillary and reticular fibroblasts. J. Invest. Dermatol. 132, 2565–2572 39 Harper, R.A. and Grove, G. (1979) Human skin fibroblasts derived from papillary and reticular dermis: differences in growth potential in vitro. Science 204, 526–527 40 Driskell, R.R. et al. (2012) Clonal growth of dermal papilla cells in hydrogels reveals intrinsic differences between Sox2-positive and negative cells in vitro and in vivo. J. Invest. Dermatol. 132, 1084–1093 41 Enshell-Seijffers, D. et al. (2008) The serine protease Corin is a novel modifier of the Agouti pathway. Development 135, 217–225 42 Collins, C.A. and Watt, F.M. (2008) Dynamic regulation of retinoic acidbinding proteins in developing, adult and neoplastic skin reveals roles for beta-catenin and Notch signalling. Dev. Biol. 324, 55–67 43 Johnston, A.P. et al. (2013) Sox2-mediated regulation of adult neural crest precursors and skin repair. Stem Cell Rep. 1, 38–45 44 Collins, C.A. et al. (2012) Polyclonal origin and hair induction ability of dermal papillae in neonatal and adult mouse back skin. Dev. Biol. 366, 290–297
Trends in Cell Biology xxx xxxx, Vol. xxx, No. x
45 Schmidt, B.A. and Horsley, V. (2013) Intradermal adipocytes mediate fibroblast recruitment during skin wound healing. Development 140, 1517–1527 46 Ambler, C.A. and Watt, F.M. (2010) Adult epidermal Notch activity induces dermal accumulation of T cells and neural crest derivatives through upregulation of jagged 1. Development 137, 3569–3579 47 Jensen, K.B. et al. (2010) Assaying proliferation and differentiation capacity of stem cells using disaggregated adult mouse epidermis. Nat. Protoc. 5, 898–911 48 Mulder, K.W. et al. (2012) Diverse epigenetic strategies interact to control epidermal differentiation. Nat. Cell Biol. 14, 753–763 49 Higgins, C.A. et al. (2013) Microenvironmental reprogramming by three-dimensional culture enables dermal papilla cells to induce de novo human hair-follicle growth. Proc. Natl. Acad. Sci. U.S.A. 110, 19679–19688 50 Horne, K.A. et al. (1986) Whisker growth induced by implantation of cultured vibrissa dermal papilla cells in the adult rat. J. Embryol. Exp. Morphol. 97, 111–124 51 Kishimoto, J. et al. (2000) Wnt signaling maintains the hair-inducing activity of the dermal papilla. Genes Dev. 14, 1181–1185 52 Lichti, U. et al. (1993) In vivo regulation of murine hair growth: insights from grafting defined cell populations onto nude mice. J. Invest. Dermatol. 101, 124S–129S 53 Ohyama, M. et al. (2010) The mesenchymal component of hair follicle neogenesis: background, methods and molecular characterization. Exp. Dermatol. 19, 89–99 54 Rendl, M. et al. (2008) BMP signaling in dermal papilla cells is required for their hair follicle-inductive properties. Genes Dev. 22, 543–557 55 Yang, C.C. and Cotsarelis, G. (2010) Review of hair follicle dermal cells. J. Dermatol. Sci. 57, 2–11 56 Silva-Vargas, V. et al. (2005) Beta-catenin and Hedgehog signal strength can specify number and location of hair follicles in adult epidermis without recruitment of bulge stem cells. Dev. Cell 9, 121–131 57 Wojciechowicz, K. et al. (2013) Development of the mouse dermal adipose layer occurs independently of subcutaneous adipose tissue and is marked by restricted early expression of FABP4. PLoS ONE 8, e59811 58 Clavel, C. et al. (2012) Sox2 in the dermal papilla niche controls hair growth by fine-tuning BMP signaling in differentiating hair shaft progenitors. Dev. Cell 23, 981–994 59 Enshell-Seijffers, D. et al. (2010) Beta-catenin activity in the dermal papilla regulates morphogenesis and regeneration of hair. Dev. Cell 18, 633–642 60 Enshell-Seijffers, D. et al. (2010) Beta-catenin activity in the dermal papilla of the hair follicle regulates pigment-type switching. Proc. Natl. Acad. Sci. U.S.A. 107, 21564–21569 61 Hu, B. et al. (2010) Control of hair follicle cell fate by underlying mesenchyme through a CSL–Wnt5a–FoxN1 regulatory axis. Genes Dev. 24, 1519–1532 62 Jeffery, E. et al. (2014) Characterization of Cre recombinase models for the study of adipose tissue. Adipocyte 3, 206–211 63 Grisanti, L. et al. (2013) Tbx18 targets dermal condensates for labeling, isolation, and gene ablation during embryonic hair follicle formation. J. Invest. Dermatol. 133, 344–353 64 Dulauroy, S. et al. (2012) Lineage tracing and genetic ablation of ADAM12(+) perivascular cells identify a major source of profibrotic cells during acute tissue injury. Nat. Med. 18, 1262–1270 65 Lesko, M.H. et al. (2013) Sox2 modulates the function of two distinct cell lineages in mouse skin. Dev. Biol. 382, 15–26 66 James, K. et al. (2003) Sox18 mutations in the ragged mouse alleles ragged-like and opossum. Genesis 36, 1–6 67 Woo, W.M. et al. (2012) Shh maintains dermal papilla identity and hair morphogenesis via a Noggin–Shh regulatory loop. Genes Dev. 26, 1235–1246 68 Watt, F.M. and Jensen, K.B. (2009) Epidermal stem cell diversity and quiescence. EMBO Mol. Med. 1, 260–267 69 Calvo, F. et al. (2013) Mechanotransduction and YAP-dependent matrix remodelling is required for the generation and maintenance of cancer-associated fibroblasts. Nat. Cell Biol. 15, 637–646 70 Hu, B. et al. (2012) Multifocal epithelial tumors and field cancerization from loss of mesenchymal CSL signaling. Cell 149, 1207–1220 71 Plikus, M.V. et al. (2011) Self-organizing and stochastic behaviors during the regeneration of hair stem cells. Science 332, 586–589 7
TICB-1095; No. of Pages 8
Review 72 Sharov, A.A. et al. (2006) Bone morphogenetic protein signaling regulates the size of hair follicles and modulates the expression of cell cycle-associated genes. Proc. Natl. Acad. Sci. U.S.A. 103, 18166– 18171 73 Millar, S.E. (2002) Molecular mechanisms regulating hair follicle development. J. Invest. Dermatol. 118, 216–225 74 Schneider, M.R. et al. (2009) The hair follicle as a dynamic miniorgan. Curr. Biol. 19, R132–R142 75 Myung, P.S. et al. (2013) Epithelial Wnt ligand secretion is required for adult hair follicle growth and regeneration. J. Invest. Dermatol. 133, 31–41 76 Donati, G. et al. (2014) Epidermal Wnt/beta-catenin signaling regulates adipocyte differentiation via secretion of adipogenic factors. Proc. Natl. Acad. Sci. U.S.A. 111, E1501–E1509 77 Takeo, M. et al. (2013) Wnt activation in nail epithelium couples nail growth to digit regeneration. Nature 499, 228–232
8
Trends in Cell Biology xxx xxxx, Vol. xxx, No. x
78 Giangreco, A. et al. (2008) Epidermal stem cells are retained in vivo throughout skin aging. Aging Cell 7, 250–259 79 Hanahan, D. and Weinberg, R.A. (2011) Hallmarks of cancer: the next generation. Cell 144, 646–674 80 Petrof, G. et al. (2013) Fibroblast cell therapy enhances initial healing in recessive dystrophic epidermolysis bullosa wounds: results of a randomized, vehicle-controlled trial. Br. J. Dermatol. 169, 1025–1033 81 Jones, E. and McGonagle, D. (2008) Human bone marrow mesenchymal stem cells in vivo. Rheumatology 47, 126–131 82 Pittenger, M.F. et al. (1999) Multilineage potential of adult human mesenchymal stem cells. Science 284, 143–147 83 Handjiski, B.K. et al. (1994) Alkaline phosphatase activity and localization during the murine hair cycle. Br. J. Dermatol. 131, 303–310 84 Egawa, G. et al. (2009) Transient expression of ephrin b2 in perinatal skin is required for maintenance of keratinocyte homeostasis. J. Invest. Dermatol. 129, 2386–2395