The Contractile Phenotype of Dermal Fetal Fibroblasts in Scarless Wound Healing

Abstract

Purpose of Review

Injured skin in the mammalian fetus can heal regeneratively due to the ability of fetal fibroblasts to effectively reorganize the extracellular matrix (ECM). This process occurs without fetal fibroblasts differentiating into highly contractile myofibroblasts which cause scarring and fibrosis in adult wounds. Here, we provide a brief review of fetal wound healing and the evidence supporting a unique contractile phenotype in fetal fibroblasts. Furthermore, we discuss the biomechanical role of the ECM in driving myofibroblast differentiation in wound healing and the implications for new clinical modalities based on the biophysical properties of fetal fibroblasts.

Recent Findings

We and others have found that fetal fibroblasts are refractory to the environmental stimuli necessary for myofibroblast differentiation in adult wound healing including mechanical stress.

Summary

Understanding the biomechanical mechanisms that regulate the contractile phenotype of fetal fibroblasts may unlock new avenues for anti-scarring therapies that target myofibroblast differentiation of adult fibroblasts.

INTRODUCTION

Injuries to tissues resulting from traumas, surgeries, and diseases are a leading global health concern. In the United States alone, treatments for dermal wounds cost tens of billions of dollars in health care expenditures [1]. Wound healing and repair are important considerations when treating tissue injuries. Many therapies have been developed to regulate the wound healing process; however, injured tissues still heal imperfectly resulting in scar formation or fibrosis which can have significant consequences. For example, dermal scarring can range from cosmetic abnormalities to major body deformations and impaired physical function [2]. Therefore, there exists a pressing need for new clinical strategies to regulate tissue repair and drive healing towards a more regenerative outcome.

Nature provides examples of regeneration in the mammalian fetus (Fig. 1). Fetal tissues can heal scarlessly depending on wound size and gestational age [3]. This regenerative response was first discovered in the dermis [4] but has also been observed in the upper airway mucosa [5] and tendon [6]. While differences in the extracellular matrix (ECM) and inflammation of fetal wounds contribute to scarless healing [7], the intrinsic properties of fetal fibroblasts are also thought to play a crucial role in this regeneration [8]. In particular, this review will focus on the biochemical and biophysical properties of dermal fetal fibroblasts that indicate a unique contractile phenotype capable of appropriately contracting and remodeling the ECM leading to regenerative repair. Understanding the mechanisms that regulate these phenotypic characteristics of fetal fibroblasts may lead to new therapeutic targets with the potential for reducing scarring and fibrosis.

Fig. 1.

After traumatic injury to tissues such as the skin, adult wounds heal with imperfect repair and scar formation resulting in a loss of natural structure and function. In contrast, fetal dermal wounds heal in a regenerative manner and are virtually indistinguishable from non-injured tissue. In either case, the quality of healing is dictated by fibroblasts that are responsible for invading the wound bed, depositing new collagenous tissue, and contracting and remodeling this tissue. Since fibroblast activity in the wound bed determines healing outcomes, regenerative or scarless wound healing is highly dependent on the intrinsic properties of fetal fibroblasts.

THE ROLE OF FIBROBLASTS IN WOUND HEALING

Fibroblasts are responsible for the production and organization of new tissue in the wound bed. In post-natal or “adult’ wounds, fibroblasts differentiate into myofibroblasts which synthesize and deposit new collagenous ECM that is initially composed of type III collagen but is later dominated by type I collagen [9]. Myofibroblasts are characterized by mature focal adhesions and stress fibers containing α-smooth muscle actin (α-SMA) that generate large contractile forces via actomyosin contractility [10]. These forces are used to excessively contract and remodel the newly deposited ECM; this mechanical response is perpetuated in a positive feedback loop as mechanical stress continues to build in the ECM due to sustained cellular forces ultimately resulting in scarring and fibrosis [11]. α-SMA expression and collagen synthesis are regulated by the cytokine transforming growth factor-β1 (TGF-β1) which is the major promoter of myofibroblast differentiation [12]. In addition, the splice variant of fibronectin containing the extra domain A is upregulated in wounds and necessary for the conversion of fibroblasts to myofibroblasts [13].

In contrast, scarless healing occurs in wounded skin of the mammalian fetus early in gestation and within the first two trimesters in humans [7]. Fetal dermal wounds heal much faster and regain their natural structure and function including skin appendages and sebaceous glands [14]. α-SMA-containing fibroblasts are not present in fetal wounds or are only transiently expressed [15], and these wounds close with minimal contraction [16]. Fetal wounds are characterized by a diminished inflammatory response resulting in decreased TGF-β1 expression and altered collagen production [17]. Fetal skin is composed of a higher ratio of type III to type I collagen which has also been reported in some fetal wounds as well [9, 18]. The newly deposited collagenous ECM in fetal wounds is organized in a fine reticular pattern similar to unwounded skin [19] resulting in the full restoration of tensile strength which does not occur with adult scars [20, 21]. In addition, higher levels of glycosaminoglycans such as hyaluronic acid are also produced in fetal wounds and have been found to be important in regenerative healing [7]. Despite these environmental differences, potential therapeutic strategies targeting cytokines such as TGF-β1 and ECM synthesis in adult wounds have yet to restrain fibrotic healing [22, 23].

THE FETAL FIBROBLAST

Pioneering transplantation studies performed at Stanford University in the early 1990’s have established the unique ability of fetal fibroblasts to facilitate scarless healing. Human fetal skin transplanted subcutaneously in adult nude mice and later wounded healed without scarring despite exposure to adult serum and inflammatory cells [8, 14]. The collagen in the regenerated skin was identified as human indicating that it was produced by the fetal fibroblasts. In contrast, the transplantation of adult sheep skin into the backs of fetal lambs and later wounded healed with scars despite exposure to amniotic fluid rich in growth factors and ECM components originally thought to be important for scarless repair [24, 25, 16]. These results suggested that scarless healing is an intrinsic property of fetal skin that is orchestrated by fetal fibroblasts. Therefore, the fetal fibroblast was concluded to be the key effector of scarless repair [8]. Intriguingly, a clinical study in which collagen constructs seeded with human fetal skin cells were transplanted into the burns of pediatric patients found that these wounds healed without retraction and with minimal scarring [26] further supporting a unique role for these cells in regenerative repair.

These findings are supported by in vitro studies showing differences in the molecular and cellular characteristics of fetal and adult fibroblasts [27–30]. For example, significant differences have been found in the expression of growth factors and ECM components by fetal fibroblasts as well as in TGF-β1 signaling [31–33]. While exogenous TGF-β1 increases whole tissue α-SMA expression and scarring in fetal wounds [34], α-SMA expression in fetal fibroblasts in vitro does not change on stress-free collagen gels or plastic despite similar TGF-β receptor and Smad mRNA levels when compared to adult fibroblasts [33, 35]. Fetal fibroblasts have also been shown to express lower levels of α1 and α3 integrins which are thought to affect their contractile capacity [36]. In fact, contractility of collagen gels has served as the primary in vitro surrogate for understanding differences in wound contraction that occurs in vivo [16, 25, 37, 33]. Consistent with other reports [16, 25], our work also found that fetal fibroblasts contract free floating collagen gels to a greater extent than their adult counterparts [38]. However, these gels lack tension and thus compact due to the migration of cells into the collagen matrix which is thought to mimic the early stages of wound healing in which fibroblasts migrate into the wound bed [39, 40]. These results parallel other studies we performed on two-dimensional surfaces and in three-dimensional collagen plugs in which fetal fibroblasts consistently migrated at a faster rate [41, 42]. In addition, we also found differential effects on migration and contraction due to a defective prostaglandin E2 (PGE2) pathway in fetal fibroblasts despite similar expression levels of the PGE2 receptors when compared to adult fibroblasts [41, 38, 43]. Therefore, fetal fibroblasts exhibit distinct phenotypic characteristics that may contribute to their ability to enable scarless wound healing such as increased migration and altered responses to soluble mediators.

ECM BIOMECHANICS AND CELLULAR CONTRACTILITY

The mechanical state of the ECM is known to play a fundamental role in driving pathological conditions in a number of conditions and diseases by regulating cellular phenotypes [44]. The tension and stiffness that develop in granulation tissue promote myofibroblast differentiation and subsequent fibrotic ECM reorganization of adult dermal wounds [45]. TGF-β1 is expressed early in the healing process but does not induce myofibroblast differentiation until a mechanical threshold is reached in the ECM [10]. In fact, latent TGF-β1 bound to the ECM must first be activated through conformational changes of the latency associated peptide caused by integrin-mediated cellular forces [46]. Similar findings in vitro have shown that ECM rigidity is a significant factor in promoting TGF-β1-induced myofibroblast differentiation [47]. Interestingly, fetal skin and wounds are more compliant than their adult counterparts [48, 20] which is thought to be due to less cross-linking of a more immature ECM composed of higher levels of type III collagen [21]. This notion is supported by an in vivo study using a type III collagen-deficient mouse model that found increases in myofibroblasts, contraction, and scarring during dermal wound repair [49]. While β1 integrins are responsible for binding fibrillar collagens and transducing cellular forces [50], types I and III collagen have different integrin-binding motifs that have been shown to affect fibroblast adhesion [51, 52]. Since the strength of cellular adhesion is critical for force generation [53], these ECM components may play an important role in determining myofibroblast differentiation. Furthermore, other integrins and ECM components may contribute to this process such as ανβ3 and fibronectin [53, 54]. Therefore, the mechanical nature of the fetal wound environment may not be conducive to myofibroblast differentiation.

To gain insight into whether mechanical stress can promote cellular contractility in fetal fibroblasts, we began to look at the biomechanical role of the ECM in adult versus fetal wound healing in vitro [38, 43]. In particular, we developed our own version of an attached type I collagen gel model as a 3-D dermal equivalent (Fig. 2a) [43]. These models have been shown to approximate granulation tissue due to biological and mechanical similarities [55], and fibroblasts will differentiate into myofibroblasts once stress develops in the gels [56]. Interestingly, fetal fibroblasts contracted the anchored collagen gels to a greater extent than adult fibroblasts at an early time point when tension was not yet present [43] similar to free floating collagen gels (Fig. 2b) [38]. However, once the gels were under noticeable tension, we found no difference between adult and fetal fibroblast contraction (Fig. 2b) [43]. At longer time points, adult fibroblasts had significantly surpassed fetal fibroblasts in their ability to contract the collagen gels (Figs. 2b–c) [43]. At the transition point in which mechanical tension was present, adult fibroblasts exhibited more prominent actin stress fibers that were rich in α-SMA (Fig. 2d) [43]. Despite differences in contraction, fetal fibroblasts were able to reorganize the anchored collagen gels to the same extent as adult fibroblasts suggesting that cellular forces exerted by untransformed fibroblasts are sufficient for ECM remodeling [43, 57].

Fig. 2.

Anchored collagen gel contraction by human dermal adult and fetal fibroblasts. (A) Anchored collagen gel contraction over time by adult fibroblasts where contraction is calculated based on the change in area of the collagen gel between the stainless steel anchors (top and bottom). As the gels contract, they become taut between the anchors within 4 hours. (B) While fetal fibroblast contraction is initially greater, adult fibroblast contraction increases at a faster rate once mechanical tension develops in the collagen gel (4 hours) and becomes greater than fetal fibroblast contraction even at (C) later time points. (D) This difference in contractile ability coincides with the expression of prominent stress fibers containing α-SMA (yellow in overlay) throughout the cytoplasm of adult but not fetal fibroblasts at 4 hours suggesting a myofibroblast phenotype. Reprinted from Parekh A, Sandulache VC, Singh T, Cetin S, Sacks MS, Dohar JE, and Hebda PA, Prostaglandin E2 differentially regulates contraction and structural reorganization of anchored collagen gels by human adult and fetal dermal fibroblasts, Wound Repair and Regeneration, 2009, 17(1), 88–98 [43] with permission from John Wiley & Sons, Inc

THERAPEUTIC PERSPECTIVE

Initial therapeutic strategies targeting scarring and fibrosis have primarily focused on inflammation and collagen deposition [58]; however, there are currently no accepted anti-fibrotic therapies [59]. While these approaches have typically been designed to indirectly affect myofibroblasts, their lack of translational success suggests that new clinical avenues must be sought for directly targeting myofibroblasts to combat fibrotic healing. An alternative approach is to interfere with the contractile forces generated by myofibroblasts that lead to excessive ECM contraction and remodeling [10, 23]. Myofibroblast differentiation and force generation are ultimately a result of fibroblasts sensing increased mechanical stress in granulation tissue [10, 45]. From this perspective, potential methods include blocking specific integrins to interfere with the transduction of mechanical signals from the ECM to the cell thus reducing cellular force generation, contraction and remodeling, and/or activation of latent TGF-β1 [23]. Similarly, targeting molecules in integrin-based signaling pathways that respond to mechanical stress such as focal adhesion kinase may also impede myofibroblast differentiation and/or activity as well as other downstream profibrotic pathways [60–62]. In addition, altering the ECM mechanical environment can modulate myofibroblast contractile activity such as by interfering with cross-linking [63] or by offloading the mechanical stress in granulation tissue [64].

Careful consideration must be given to any potential therapies regarding the degree to which fibroblast activity is impeded since some activity is necessary for proper wound healing to occur. Coming from a fetal wound healing perspective, we have advocated for reducing adult fibroblast contractile forces to a “fetal level” in an attempt to promote regenerative healing [41, 38, 43]. Evidence from our work suggests that the unique contractile phenotype of fetal fibroblasts could serve as a biomechanical benchmark for fibroblast activity. While we and others have found some signaling defects in fetal fibroblasts which have been well documented [27–30], intrinsic biochemical variations in fetal fibroblasts could be further explored and potentially exploited in adult fibroblasts to compromise their fibrotic behavior. In particular, fetal fibroblasts appear to exhibit distinct responses to mechanical factors, thus we feel strongly that this phenotype warrants further exploration in an attempt to uncover new therapeutic targets that likely exist in mechanical signaling pathways. In addition, molecular methods aimed at reprogramming adult fibroblasts could be combined with bioengineered scaffolds to provide a mechanical environment more favorable for scarless healing [65, 29]. Alternatively, cell therapies using fetal fibroblasts could also be utilized to take advantage of their innate ability for regenerative repair [66–69] given their early but promising clinical results [26]. While the use of fetal cells and tissues is highly controversial and now significantly limited [70], human dermal fetal fibroblasts may hold the key for regeneration and warrant consideration for future work in this understudied area of wound healing.

CONCLUSIONS

While the properties of the fetal wound environment contribute to scarless healing, fetal fibroblasts exhibit distinct biochemical and biophysical characteristics that are necessary for regenerative repair. In particular, fetal fibroblasts exhibit a unique contractile phenotype that facilitates arrangement of the ECM in a manner that recapitulates its natural structure and function in the dermis. This cellular behavior is conserved in vitro, in vivo, and ex vivo indicating that fetal fibroblasts have intrinsic differences in their molecular makeup when compared to their adult counterparts. While more studies are necessary, fetal fibroblasts may not be able to differentiate into myofibroblasts due to altered biomechanical responses that reduce their contractility. Understanding the molecular mechanisms that regulate the contractile phenotype of fetal fibroblasts may reveal new therapeutic targets in mechanical signaling pathways to minimize scarring and fibrosis caused by adult fibroblast activity.