Regenerative Healing in Fetal Skin: A Review of the Literature
Exploration of fetal wound healing began in the 1950s with examination of animal models. Initial studies by Hess1,2 indicated that fetal wounds heal rapidly and show signs of regeneration, in some cases making it difficult to identify the wound site. In subsequent studies,3-7 a lack of inflammation and healing without a scar4 were described in fetal wounds. Similar differences in healing were noted in the human fetus8 and were confirmed when the first human fetal surgeries were performed.9 As experimental studies continued, it became evident that fetal wounds heal differently depending on the gestational age of the fetus (see Figure 1).10-17 In the first and second trimesters of development, fetal skin undergoes rapid healing with little or no inflammation and no scarring. Scarless healing in early fetal skin is a form of regeneration, with renewal of skin appendages such as hair follicles and sebaceous glands in addition to the restoration of a normal dermal matrix and no scar. Near the third trimester, a transition period occurs. At this point, the skin begins to lose its ability to regenerate and instead undergoes fibrotic healing similar to that seen in postnatal skin. After this transition to fibrotic fetal healing, cutaneous wounds heal more slowly, exhibiting inflammation and significant scarring. After the transition period, wounded fetal skin will display typical signs of scar tissue, with the loss of subepidermal appendages and an overproduction of densely packed, disorganized collagen. Pathology reports and basic research studies8,17 have shown that scarless healing occurs in fetal skin until around 22 to 24 weeks of gestation in the human fetus.
The point of transition from scarless to fibrotic healing also has been determined for many animals; they have been used in experimental models of fetal wound healing. The author’s laboratory prefers the mouse model, owing to the abundance of available reagents and the availability of genetically modified mice. In mice, the transition occurs at about embryonic day 16 (E16), so wounds made at E16 or earlier heal without a scar (regenerative or scarless repair) and wounds made later than E16 heal with a scar (fibrotic repair).18-21 Scarless and fibrotic fetal wounds then can be compared and used to identify pathways involved in regenerative healing. It is important to note that scarless fetal repair is somewhat dependent on the size and severity of the wound14 and, with the exception of bone,22 fetal skin appears to be unique in its capacity to heal by regeneration.23
The development of animal models of fetal wound healing sparked great interest in the study of fetal repair for several reasons. The use of invasive diagnostic procedures such as amniocentesis, chorionic villus sampling, and fetal blood or tissue sampling all have the potential to cause injury to the fetus,24-26 making it important to know how the fetus will respond should damage occur. In addition, with the use of fetal surgery in cases of congenital cystic adenomatoid malformation of the lung, sacrococcygeal teratoma, myelomeningocele, congenital diaphragmatic hernia, and cleft lip and palate,27 knowledge of how the fetus will heal following surgery is essential. This information is even more critical when considering the likelihood that with the development of less invasive fetal surgical techniques, the number of these surgical procedures performed will continue to rise. Finally, studying fetal repair could provide mechanistic insights into regenerative healing, offering key information about how to accelerate healing in postnatal wounds as well as how to improve healing from a cosmetic standpoint. This review discusses the differences between scarless and fibrotic repair and how they might be exploited in the future to make adult wounds heal more like those in the fetus.
Environmental versus Inherent Factors
Many of the earliest studies on fetal wound healing considered whether the fetal environment or characteristics inherent in fetal skin were responsible for scarless repair. The uterine environment in which fetal wounds heal is unique, with amniotic fluid surrounding the healing wounds. Originally, this warm sterile amniotic fluid, rich in growth factors and extracellular matrix components, was considered important for the scarless fetal healing process. Although it has been suggested that the sterile nature of amniotic fluid and anti-inflammatory factors that it contains may help facilitate non-inflammatory, scarless healing,28,29 the amniotic fluid environment is not required for this process. Studies utilizing the developing opossum12 indicate that amniotic fluid is not essential for scarless healing. In this marsupial model, offspring develop in a pouch instead of a uterine environment, but the developmental process in the pouch resembles the in utero development of a mammalian fetus. For the developing opossum, young animals up to pouch day 9 heal without scarring even though the wounds do not heal in an amniotic fluid environment.
Transplantation studies also were used to investigate the importance of amniotic fluid in scarless tissue repair. Studies in sheep have shown that wounds made in adult skin or late gestation fetal skin transplanted onto fetal lambs before wounding heal with a scar30; therefore, skin beyond the transition to fibrotic healing continues to heal with a scar even if repair takes place in a fetal environment. In addition, early human fetal skin transplanted subcutaneously in nude mice heals without a scar after wounding, demonstrating that scarless healing in fetal skin is independent of amniotic fluid or perfusion by fetal serum.17 Together, these studies led to the conclusion that intrinsic characteristics exclusively found in early fetal skin mediate scarless repair.
One of the first distinguishing characteristics of scarless fetal healing identified was a lack of inflammation.3-7 A diminished inflammatory response in scarless fetal wounds has been demonstrated repeatedly in many different models of fetal wound healing.4,5,8,12,18,20,21,31-36 The presence of inflammation during repair is believed to contribute to the transition from scarless to fibrotic healing in fetal skin because a significant inflammatory response to injury does not manifest until the third trimester when the skin begins to heal with a scar.12,18,21 Additional experimental studies support the relevance of a lack of inflammation in scarless healing. Inducing inflammation with killed or live bacteria,28 chemical agents,37 or various mediators of inflammation including cytokines,38 growth factors,18,33,39-43 and prostaglandins21 in early fetal wounds results in the formation of a scar, when normally these wounds would have healed without a scar. Therefore, despite early speculation that the absence of inflammation in fetal wounds resulted from the inability to mount a proper immune response, inflammation can be induced in fetal wounds; when it is, fibrotic healing ensues.
Once the idea that minimal inflammation defines scarless fetal repair was established, studies focused on characterizing the specific inflammatory cell types that were missing and determining the mechanisms of reduced inflammation. Virtually all of the cells involved in acute inflammation respond differently to injury in early fetal skin, including platelets, neutrophils, macrophages, and lymphocytes.
Platelets. Platelet aggregation and subsequent cytokine production by these cells are initial events in adult wound healing and help trigger an acute inflammatory reaction. In a series of studies, Olutoye et al44 demonstrated that fetal and adult platelets behave differently. Although mid-trimester fetal platelets are structurally similar by electron microscopy, have a similar number of α-granules, and produce similar levels of collagen receptors (α2 subunit of the α2β1 integrin) compared to adult platelets,44,45 fetal platelets show less aggregation in response to collagen, a matrix component that stimulates platelet aggregation during adult wound repair. In addition, fetal platelets release lower levels of cytokines, which likely contributes to the reduced levels of inflammation in early fetal wounds.45
Immune cells. Lower numbers of infiltrating inflammatory cells, including fewer neutrophils,4,12,33,46,47 macrophages and activated macrophages,35,36 T cells,33 and B cells36 all have been described in scarless fetal wounds compared to fibrotic wounds. Although it is not known whether one cell type is more important than another for scar formation, studies48-50 have confirmed a role for neutrophils, macrophages, and T cells in scar formation in adult skin.
Inflammatory mediators. More current in vivo fetal wound healing studies have examined the role of specific inflammatory mediators such as cytokines and prostaglandins in fetal wound healing. Liechty et al38,51 showed that several members of the interleukin family of inflammatory mediators play a role in scarless fetal repair. They demonstrated that the pro-inflammatory mediators IL-6 and IL-8 are produced at lower levels by fetal fibroblasts compared to adult fibroblasts and that fetal skin wounds express lower levels of IL-6 and IL-8 than adult wounds. In addition, injecting recombinant IL-6 into fetal wounds caused fibrotic healing.38 The same authors also showed that fetal skin deficient in the anti-inflammatory IL-10 heals with a scar compared to skin with normal levels of IL-10, which heals scarlessly.52
Early experiments by Morykwas et al53 examined the effects of prostaglandins on inflammation in fetal rabbit wounds. These studies showed that implanted slow-release pellets containing PGF2α or PGE2 induced inflammation; however, potential effects on scar formation were not examined. Studies from the author’s lab showed that scarless fetal wounds produce lower levels of cyclooxygenase-2 (COX-2), one of the enzymes that produces PGE2, and contain less PGE2 than fibrotic fetal wounds.21 In addition, injecting exogenous PGE2 into early fetal wounds delays healing and induces the formation of scar tissue.21 Adult wound healing studies have shown that treatment with PGE2 or a PGE2 analog can induce collagen synthesis and fibrosis54,55 and that inhibition of COX-2 and reduction of PGE2-mediated inflammation with celecoxib can reduce scarring,49 further supporting the idea that prostaglandins can influence scar formation.
Reactive oxygen species. Reactive oxygen species, produced during an active inflammatory response, can influence a variety of healing parameters.56,57 Studies conducted in the author’s laboratory have shown that hydrogen peroxide, a potential fibrogenic byproduct of inflammation, causes scars to form when added to early fetal wounds in which scarless healing would normally occur. Hydrogen peroxide appears to induce scar formation in otherwise scarless wounds without altering the level of neutrophil infiltration but may involve increases in TGF-β1.58
Leukocyte-endothelial cell interactions. Although minimal inflammation is important for healing to proceed scarlessly in fetal skin, it still is not completely clear why early fetal skin fails to elicit an acute inflammatory response. However, a recent study by Olutoye et al59 suggests that altered interactions between fetal endothelial cells and inflammatory leukocytes may be one reason these cells do not migrate into early fetal wounds. Studies showed that neutrophils adhere less to and have a higher rolling velocity over fetal endothelial cells compared to adult endothelial cells after endothelial cell stimulation with TNF-α; they also display a reduced ability to transmigrate through fetal endothelial cell monolayers compared to monolayers of adult endothelial cells. This may be due to the fact that fetal endothelial cells express lower mRNA levels of the integrin P-selectin compared to adult endothelial cells in response to TNF-α stimulation.59 These results suggest that in addition to reduced levels of pro-inflammatory mediators, a diminished capacity of fetal endothelial cells to facilitate emigration of leukocytes from the bloodstream into the wound bed could contribute to the dampened inflammatory response seen in scarless fetal wounds.
Reepithelialization and Keratinocytes
Because early fetal wounds heal without a scar, much of the work in the field of fetal wound healing has focused on fibroblasts and regeneration of the dermis. Even though faster reepithelialization (generally within 24 hours) occurs in scarless fetal wounds,4,10,12,20,32,60,61 relatively little attention has been given to characterizing differences in keratinocytes or the reepithelialization process. However, several studies have suggested that different mechanisms are involved in fetal reepithelialization and that alterations in fetal keratinocytes may influence epidermal healing.
Martin et al, who conducted most of the studies on fetal wound reepithelialization, identified fundamental differences in the mechanics of reepithelialization in embryonic wounds. Although adult wounds have been shown to reepithelialize through extension of lamellipodia followed by epidermal cells at the wound edge crawling over the wound bed,62,63 embryonic wounds exhibit no signs of lamellipodia or filopodial extensions. Instead, epidermal cells at the edge of wounds in both chick and mouse embryos assemble an actin cable that contracts like a purse string to close the wound.61,64-66 Proof of the importance of actin purse string formation was provided by studies showing that when the assembly of the actin cable is disrupted by treatment with cytochalasin D, wound reepithelialization is blocked.65 Myosin II, another component of the actin cable, mediates the contractile forces necessary to close the actin purse string. Localization of cadherin molecules to adherens junctions, which potentially acts to anchor the actin cable to neighboring epidermal cells at the wound edge, is also important.66 These studies also demonstrated the requirement for the small GTP-binding protein Rho, but not Rac, in proper assembly of the actin cable and reepithelialization of fetal wounds.66
An in vitro model67 of fetal wound healing confirmed the presence of actin cables in wounded E17 fetal rat skin but not E19 rat skin. In this model, E17 wounds (a time that correlates to scarless healing in vivo) reepithelialize but E19 wounds do not, confirming the role of the actin cytoskeleton in reepithelialization. Paxillin and gelsolin are differentially regulated proteins also associated with the actin cytoskeleton. Paxillin showed co-localization with actin in E17 wounds but not E19 wounds; whereas, gelsolin was associated with actin in E19 wounds but not E17 wounds.67
Rapid alteration in epidermal integrin expression also may allow early fetal skin to quickly reepithelialize. Cass et al68 showed that a variety of integrins (molecules responsible for facilitating migration by allowing cells to adhere to the extracellular matrix) are quickly upregulated by keratinocytes after wounding in fetal skin. This rapid upregulation of integrins in fetal keratinocytes occurs much sooner than what has been reported in adult keratinocytes and may be important for quick reepithelialization in fetal wounds.
Transcription factor induction and signal transduction also is different in fetal wound keratinocytes. Martin and Nobes69 reported early induction of the nuclear transcription factor c-fos in epidermal cells in response to wounding in the fetus, which may regulate downstream signals involved in rapid healing. Using an in vitro fetal wound healing model, Gangnuss et al70 showed that expression of AP-1 transcription factors persisted in E19 wounds compared to E17 wounds. They also demonstrated that the keratinocyte differentiation markers keratin 10 and loricrin were upregulated during healing in E19 skin but not in E17 skin, suggesting that keratinocyte maturation and differentiation state may influence the ability of keratinocytes to rapidly restore epidermal integrity after injury.
Based on the collective information gained by these studies, it is evident that the process of wound reepithelialization and the response of keratinocytes to wounding are much different in early fetal skin. Additional studies to further characterize the reepithelialization process by fetal keratinocytes and to find new ways to force adult keratinocytes to more quickly reestablish the epidermal barrier following injury are needed.
Whether a fetal skin wound will heal with or without a scar depends on the developmental state of the skin. However, relatively little is known about how the developmental status of the skin and the molecular programming during skin development contribute to scarless healing. Clearly, molecular pathways involved in skin and skin appendage development are altered at times coincident with the transition from scarless to fibrotic healing. The transition period occurs as epidermal stratification and skin appendage development (eg, hair follicles, sebaceous glands) is taking place.71
A few studies have explored the potential involvement of developmental genes in scarless healing. The bone morphogenetic proteins (BMPs) belong to one family of developmental growth factors involved in skin and hair follicle development.72 Stelnicki et al73 demonstrated BMP-2 expression in the epidermis and developing hair follicles of human fetal skin. When exogenous BMP-2 was added to fetal lamb wounds, epidermal growth was augmented and the number of hair follicles and other skin appendages increased. However, fibroblast number and fibrosis also increased. Stelnicki et al74 also found differential regulation of two homeobox genes, PRX-2 and HOXB13, in a model of scarless fetal repair. Expression of PRX-2 was induced during scarless fetal wound healing in contrast to adult wounds that displayed no increase in expression. In addition, expression of HOXB13 was reduced during fetal wound healing compared to unwounded skin or adult wounds. Subsequent studies from the same group showed that deletion of the PRX-2 gene altered in vitro wound healing parameters of fetal but not adult skin fibroblasts,75 which may have implications for the regulation of scarless healing.
The Wnt signaling pathway also is known to be important for skin and hair follicle development.76 Recently, Colwell et al77 showed that Wnt-4 is expressed at higher levels in uninjured fetal skin than in postnatal skin. However, Wnt-4 expression increased during the healing of both fetal and postnatal wounds, suggesting that Wnt-4 is not likely to be an important mediator of scar tissue formation.
A study in adult mice by Fathke et al78 suggests that another Wnt family member may be involved in skin regeneration. Applying retroviral vectors containing Wnt-5a to adult skin wounds resulted in rudimentary hair follicles and sebaceous gland formation in the healed wound.
Together, these studies lend credence to the idea that altering signaling pathways important for skin and skin appendage development may someday be used to encourage regenerative healing in adult wounds. However, additional studies to better define the role of BMPs, homeobox genes, and Wnt signaling in regenerative repair and to examine other signals known to be important for skin and skin appendage development such as the hedgehog and delta/notch signaling pathways are necessary.76 Determining whether other processes that occur during skin development — ie, epidermal stratification and keratinocyte commitment to terminal differentiation — are involved in regenerative healing also is important. A review by Koster and Roop71 concluded that alterations in the expression of certain differentiation markers, such as keratin family members and p63 isoforms, also appear to coincide with the transition from scarless to fibrotic healing in developing fetal skin and may influence the healing phenotype.
Angiogenesis, the process of new blood vessel growth, is a key element of the proliferative phase of healing in adult wounds. Until recently, no studies quantitatively comparing angiogenesis in early and late fetal wounds were available. Several early studies suggested that robust angiogenesis is not a prominent feature in scarless fetal wounds. Whitby and Ferguson20 reported a noticeable lack of neovascularization in fetal mouse wounds compared to adult wounds with immunostaining for collagen IV and laminin, both of which stain endothelial basement membranes. In addition, in a rat model of fetal wound healing, Ihara et al31 observed angiogenesis in E19 fetal wounds but not in E16 wounds. Several pro-angiogenic factors have been found to be absent or present at lower levels in scarless versus scar-forming wounds, including basic fibroblast growth factor (bFGF), TGF-β1, PDGF, and PGE2.19,21,41,79-81 It also has been shown that the addition of substances capable of inducing angiogenesis to early fetal wounds (eg, bacteria,28 TGF-β1,18,33,40-43,82 PDGF,39 PGE2,21 and hyaluronidase83) causes scar formation.
More recently, Colwell et al84 reported that blood vessel counts and VEGF mRNA levels measured by real-time reverse transcriptase-polymerase chain reaction (RT-PCR) were higher in E16 scarless excisional rat wounds compared to fibrotic wounds made at E18. However, ongoing studies by the author’s lab suggest that this may not apply to every fetal system.85 Although the role of angiogenesis and VEGF in scarless fetal repair is not entirely clear, data from adult studies support a role for angiogenesis in scar formation. The author’s results have shown that neutralizing VEGF in adult wounds limits scar tissue production85; others have shown that inhibiting angiogenesis or VEGF signaling can reduce scar formation.86-89 Additional studies are needed to evaluate the use of anti-angiogenesis therapy to reduce scarring for appropriate types of wounds and patient populations.
Extracellular Matrix (ECM)
The most obvious difference between fetal and adult skin healing is the lack of dermal scar tissue formation. Instead of producing a scar with a significantly altered ECM, fetal skin has the unique ability to lay down new ECM with a composition and arrangement similar to normal unwounded skin, complete with the regeneration of hair follicles and other appendages. Fetal wound healing studies examining several ECM components, including glycosaminoglycans, glycoproteins, proteoglycans, collagen, and matrix-degrading enzymes, suggest that ECM differences may contribute to regenerative healing (see Table 1).
Glycosaminoglycans. One of the first differences identified in the ECM of fetal wounds was the presence of high levels of glycosaminoglycans (GAGs).34,90 Glycosaminoglycans are polysaccharides comprised of repeating acidic and basic disaccharides found at high levels in connective tissues.91 Important GAGs in the skin include chondroitin sulfate, dermatan sulfate, heparin sulfate, heparin, and hyaluronic acid (HA). Whitby and Ferguson20 studied the expression of two GAGs, chondroitin sulfate and heparin sulfate, in fetal lip wounds. They found that significant amounts of chondroitin sulfate are produced in fetal but not adult wounds, suggesting chondroitin sulfate may help facilitate regenerative healing. However, no differences were observed in heparin sulfate expression between fetal and adult wounds.
Many initial fetal wound healing studies focused on one particular GAG, hyaluronic acid (HA), which appears to be present at the highest levels in fetal wounds.34,90 Hyaluronic acid can influence the structure, assembly, and hydration of the ECM and appears to support cell growth and migration, especially during development.92,93 Amniotic fluid has high levels of HA, especially earlier in gestation94,95; high levels of HA-stimulating activity (HASA), a factor that promotes HA deposition, also are present in amniotic fluid.96 Higher levels of HA have been described in wounds or PVA sponge implants in fetal rabbits.34,90 In sheep, wound fluid from early gestation fetal wounds contains higher levels of HA and HASA than late gestation fetal or adult wound fluid.95 Higher HA and HASA levels are maintained for a longer period of time in fetal versus adult wound fluid.95,97,98 Fetal wounds also have less of the HA-degrading enzyme hyaluronidase than adult wounds.99 Together, the high levels of HA and HASA, combined with low levels of hyaluronidase, are thought to be responsible for the persistent, high levels of HA in fetal wounds.
The relevance of high HA levels to scarless healing has been highlighted by several studies. When forced degradation of HA occurs in fetal skin through the addition of hyaluronidase in PVA sponge implants, a fibrotic response occurs; this is not the case when control sponges are implanted.83 Similarly, in adult mice, PVA sponge implants containing hyaluronidase caused a fibrotic response; whereas, sponges containing HA resulted in normal ECM deposition.100 An in vitro model101 of wounded mouse limbs where E14 limbs underwent scarless repair and E18 limbs healed with scar was used to study the effects of exogenous HA on the repair process. When E18 wounds, which normally heal with a scar, were cultured in the presence of HA, scarless healing resulted.101 Together, these data suggest that continual high levels of HA are important for regenerative healing.
Glycoproteins. Integral components of the ECM, glycoproteins bind integrins, collagen, and proteoglycans, holding tissues together and functioning in cell adherence.102 The glycoproteins laminin, fibronectin, and tenascin have been studied during fetal wound repair. In murine wounds, laminin is present in the endothelial basement membrane and also in the epithelial basement membrane but only after reepithelialization is complete.20 This is consistent for both fetal and adult wounds. Fibronectin, part of the provisional wound matrix, displays similar staining patterns and timing of expression in fetal and adult wounds in mice and sheep,20,60 but may be expressed a few hours earlier in fetal rabbit wounds.103 Tenascin-C appears to be the most differentially expressed glycoprotein in scarless repair. Although tenascin is expressed in all wounds, it is deposited much sooner in early fetal wounds compared to either late gestation fetal or adult wounds.20,60 In mice, tenascin is expressed as early as 1 hour post-wounding in fetal wounds compared to 24 hours post-wounding in adult wounds.20 Tenascin is deposited before reepithelialization – the fact that it is deposited more quickly in fetal wounds suggests it may mediate the quick reepithelialization seen in early fetal skin.
Proteoglycans. Proteoglycans are made up of a core protein linked to one or more GAGs.102 The expression of several members of the small leucine-rich proteoglycan (SLRP) family has been studied in fetal wound repair. This family of proteoglycans, including decorin and fibromodulin, regulate collagen fibrillogenesis, growth factor activity, and cellular proliferation.104 By RT-PCR, reduced decorin expression was found in early scarless fetal wounds but did not change in scar-forming fetal wounds.105 In adult wounds, decorin expression increases during healing.106 These data suggest that reduced mRNA levels of decorin may correlate with scarless healing. In other models, reports conflict about the role of decorin in scar formation and fibrosis.107-116 One potential explanation for the conflicting data is that mRNA and protein levels of decorin do not always correlate.117 Additional studies to clarify the role of decorin in scar formation and fibrosis are needed.
Protein levels of fibromodulin, another SLRP family member, are significantly increased in scarless fetal wounds compared to scar-forming fetal wounds.106 This correlates with data demonstrating the ability of small proteoglycans to reduce collagen fibrillogenesis and of fibromodulin to bind TGF-b with high affinity.118,119 Further studies to more clearly define the role of proteoglycans in scarless repair and to determine whether their influence on collagen fibrillogenesis or TGF-b activity play a role in this process should be performed.
Collagen. Collagen production obviously differs between scarless and scar-forming fetal wounds. In the first and second trimesters of development, fetal skin is capable of healing wounds with newly formed collagen in a fine reticular or basket-weave pattern identical to normal skin.12,13,15,17,20,30 This type of collagen network allows for regeneration rather than scar formation and is distinctly different from the thick, disorganized, parallel bundles of collagen that make up scar tissue. Many aspects of collagen synthesis have been studied in scarless fetal healing, including the amount, types, and speed of collagen production, tensile strength, and level of cross-linking.
In normal fetal skin, the amount of total collagen increases during development.120-122 However, reports conflict about the timing and amount of collagen production in scarless fetal wounds compared to scar-forming wounds. Some studies have reported less collagen production in fetal wounds compared to adult wounds46,123 and in early fetal wounds compared to late fetal wounds in a PVA sponge model.121 One study showed that relative collagen synthesis (normalized to non-collagen protein synthesis) is similar in fetal wounds compared to unwounded skin but increases significantly in adult wounds and several studies have reported more rapid synthesis of collagen in early fetal wounds.20,47,120,124
Similar discrepancies have been noted in the types of collagen produced. Some investigators have reported lower collagen I:III ratios in fetal wounds.123 Others have reported an increased ratio of collagen I:III in early gestation fetal sponge implants120 or no difference in the ratio121,125 when compared to scar-forming fetal or adult wounds. It appears that in both scarless and scar-forming wounds, type III collagen is primarily deposited early in the repair process and later switches to production of collagen I.121,125 Although reports conflict as to whether the ratio of collagen I:III differs in scarless and fibrotic wounds, uninjured fetal skin clearly contains higher levels of type III collagen compared to type I. 120-123,125,126
Aside from characterizing the collagen content of scarless fetal wounds, studies also have determined the amount of collagen cross-linking and the tensile strength of these wounds. Collagen cross-linking has been reported to be lower in earlier stages of fetal development in sponge implants, indicating that less mature collagen is manufactured.120 It has been suggested that mature, more cross-linked collagen may correlate to increased rigidity and enhanced scar formation. Although less collagen cross-linking has been described in early fetal skin, this does not appear to correlate with a reduction in strength – the tensile strength of fetal wounds is similar to adult wounds when normalized to age-matched, unwounded skin.46
Inconsistencies in the reported amounts and types of collagen produced during fetal wound healing likely stem from differences in the animal models used, whether collagen assessments were made in wounds or in sponge implants, and what methods were used to examine the collagen. Regardless of the absolute amount, timing, or types of collagen produced during the healing of scarless fetal wounds, a limited amount of organized collagen is made, unlike the overproduction of disorganized collagen in scar tissue. It is evident that the collagen is deposited and arranged in a fundamentally different way during scarless repair, allowing for the regeneration of normal skin with the re-growth of subepidermal appendages and a reticular network of collagen.
Matrix metalloproteinases (MMPs). During normal adult wound repair, extensive ECM remodeling leads to collagen reorganizing and formation of a mature scar. The remodeling process consists of both the production and the degradation of collagen and other ECM components; a relative imbalance in this process by either an abnormally high production of collagen or inadequate degradation could result in excessive scarring. According to a review of the subject,127 MMPs and TIMPs (tissue inhibitors of the proteolytic activity of MMPs) are key to maintaining the appropriate balance between matrix production and degradation. Although little is known about the role of MMPs and TIMPs in scarless fetal healing, a few studies have suggested that alterations in the control of matrix degradation could be a factor in scarless repair. Bullard et al82 used immunohistochemistry to examine MMP levels in a model of subcutaneous transplantation of human fetal skin in immunodeficient mice and reported higher levels of MMP-1 (interstitial collagenase), MMP-2 (gelatinase A), and MMP-3 (stromelysin-1) in mid-gestation human fetal skin compared to adult skin. In addition, the authors showed that adding TGF-b reduced MMP levels in fetal skin. Although MMP levels were examined only in unwounded fetal and adult skin, the authors suggested that high levels of MMPs in fetal skin could contribute to scarless repair. Their results also suggest that one way TGF-b may augment scar formation is by reducing matrix degradation through the MMPs.
More recently, mRNA levels of a panel of MMPs and TIMPs were determined during scarless and fibrotic fetal wound healing in rat skin. Dang et al128 showed that scarless fetal wounds express MMP-1, MMP-9 (gelatinase B), and MMP-14 (membrane type-1 MMP) more quickly or at higher levels than fibrotic fetal wounds. In addition, while MMP-2, TIMP-1, and TIMP-3 expression is not altered during scarless healing, MMP-2 levels significantly decrease and TIMP-1 and TIMP-3 levels increase in fibrotic wounds. Together, these data indicate that high levels of MMPs exist in fetal skin and that a higher MMP:TIMP ratio is present in scarless versus scar-forming fetal wounds. The high levels of matrix-degrading enzymes, supporting matrix degradation rather than matrix production, may play an important role in the ability of fetal skin to repair wounds without the production of scar tissue.
Fibrogenic Growth Factors
Several growth factors with pro-fibrotic properties have been studied in fetal wound healing. Many of the published fetal wound healing experiments have focused on TGF-b family members because this family of proteins has been shown to have a defined role in fibrosis.129 Differences in TGF-b expression, with lower levels and rapid clearance of TGF-b1 and -b2 during scarless fetal repair compared to fibrotic healing, have been demonstrated repeatedly.19,41,43,79-81,130 In particular, reduced expression of TGF-b1 in early fetal wounds has been confirmed in incisional and excisional wounds in murine, rat, and human skin. The consistency of these findings in several diverse models suggests that minimal TGF-b1 expression in response to injury is a conserved response in early fetal skin.
Numerous studies demonstrate that adding TGF-b causes a fibrotic healing response in early fetal skin that would normally heal scarlessly or without fibrosis, further solidifying a role for TGF-b in scar formation.18,33,40-43,82 The TGF-b3 isoform, which has anti-fibrotic effects in adult wounds,131 is reportedly higher in non-scarring fetal wounds.81 In addition, TGF-b receptors TGF-bRI and TGF-bRII are present at lower levels in scarless fetal wounds than in fibrotic wounds.79,81
Aside from TGF-b, platelet-derived growth factor (PDGF) is also considered a fibrogenic cytokine. Whitby and Ferguson19 performed immunohistochemistry on scarless and fibrotic wounds and found that while PDGF is present in both types of wounds, it is produced for a shorter length of time in scarless fetal wounds. Additionally, in fetal rabbits, adding PDGF induces fibrosis,39 suggesting that it may play a role in scar formation.
Fibroblasts and Myofibroblasts
The fibroblast is the primary cell type responsible for determining whether scarless or fibrotic healing will occur; therefore, regenerative fetal healing must ultimately depend on the ability of fetal fibroblasts to produce and arrange new collagen and other ECM components in similar quantities, ratios, and arrangements to unwounded skin. These characteristics appear to be unique to early gestation fetal fibroblasts because skin fibroblasts past the transition period lose the ability to make normal ECM in response to wounding. The critical role of the fetal fibroblast in scarless healing was highlighted by Lorenz et al.16 Human fetal skin retains the ability to heal scarlessly when transplanted subcutaneously in nude mice but heals with a scar when transplanted as a cutaneous graft. Utilizing antibodies specific for either mouse or human collagen types I and III, the authors demonstrated that the healed dermis in scarless, subcutaneous grafts was made up of human collagen. The collagen contained within the newly formed dermis of the subcutaneous graft wounds was assessed histologically and found to be arranged in a reticular pattern similar to unwounded skin. The collagen is these scarless wounds was identified as human collagen and as such must have been produced by the human fetal fibroblasts. Conversely, the new matrix in the cutaneous grafts that healed with a scar consisted of mouse collagen, suggesting that the murine adult fibroblasts, not the human fetal fibroblasts, were involved in the production of scar tissue.16 These studies underscore the unique ability of fetal fibroblasts to facilitate scarless healing in the fetus.
Aside from normal fibroblasts, myofibroblasts, specialized contractile fibroblasts, also can contribute to wound repair. These cells express a-smooth muscle actin (a-SMA) and are characterized using transmission electron microscopy by a well-developed rough endoplasmic reticulum, nuclei with irregular borders, secretory vesicles denoting active collagen synthesis, and organized microfilament bundles.132,133 In fetal wounds, myofibroblast numbers appear to correlate with fibrotic healing. Studies in sheep133 have indicated that myofibroblasts are absent in early scarless fetal wounds but are present during healing at later stages when prominent scarring occurs. In addition, incisional or small excisional wounds (2 mm) heal without a scar and do not contain myofibroblasts.134 In contrast, larger excisional wounds that heal with a scar contain strong a-SMA staining, with the number of myofibroblasts correlating with the size of the wound and amount of scarring.134 A lack of myofibroblasts also has been reported in wounded mouse embryos.65 In addition, adding TGF-b1 to early fetal rabbit wounds induces fibrosis40 and increases the number of a-smooth muscle actin-positive myofibroblasts in the wounds,135 further supporting the association of myofibroblasts with scar formation in fetal wounds.
In the past 20 or 30 years, scientists have discovered the remarkable ability of the fetus to heal cutaneous wounds by regeneration through a process that involves no inflammation, rapid reepithelialization, and no scarring. Studying this process has the potential to provide clues about how to make postnatal wounds heal more quickly and with less scarring. Several observations made in fetal wounds have been applied to adult wound healing with some success, at least in animal models.
Many studies have emphasized the importance of low levels of TGF-b1 to scarless fetal repair so it is not surprising that TGF-b1 inhibition is now a major anti-scarring strategy. Ferguson’s group initially showed the benefits of treatment with neutralizing TGF-b1 and TGF-b2 antibodies on minimizing scarring136,137 and since then several others have shown that blocking TGF-b1 activity or reducing TGF-b1 levels can reduce scarring in adult animals.138-140 Currently, clinical trials of drugs designed by the biotech company Renovo are underway in the UK to determine whether inhibiting TGF-b1 activation can improve scarring.
Another prominent feature of fetal skin is its ability to heal wounds without the induction of an acute inflammatory response. In the past few years, several groups have shown that limiting inflammation in adult wounds can enhance repair, either by increasing the rate of reepithelialization or by reducing scarring.48-50,141 Modulating angiogenesis, which is often linked to inflammation and may be dampened in scarless wounds, could turn out to be another way to reduce scar formation in adult wounds. Although blocking angiogenesis is frequently associated with delayed healing, in vivo evidence exists indicating that reducing angiogenesis has no effect on overall healing rates or can even enhance healing.89,142-147 In addition, several studies have shown that reducing angiogenesis or interrupting the VEGF signaling pathway can minimize scarring.85-89 More studies to determine whether reducing scar formation through anti-angiogenic therapy is a valid option and to more clearly define the effects this may have on reepithelialization and healing rates are needed. Whether limiting angiogenesis is an option will likely depend on the type of wound and whether the rate of healing or the amount of scarring is the primary concern.
The newest experimental wound healing therapies involve utilizing the regenerative capacity of fetal cells directly. A recent study by Hohlfeld et al148 emphasized the potential impact of fetal skin cells in healing postnatal wounds. These Swiss researchers generated skin constructs using skin cells from a 14-week aborted fetus and used the constructs as grafts for pediatric burn patients. The authors reported success using the fetal skin constructs, with relatively quick healing and minimal scarring. Although these studies are not without controversy and the mechanisms underlying the ability of these fetal grafts to facilitate healing are not known, they show that using fetal cells as a direct form of therapy may be a viable wound care option in the future.
Although quite a bit is known about the process of scarless fetal repair, many aspects of this unique type of healing still are not understood. Of particular concern is the fact that to date, no information is available about whether epithelial or mesenchymal stem cells are involved in regenerative fetal healing. However, with the development of new tools to study and identify stem cells and the increasing use of high-throughput analytical techniques such as gene array and proteomics, it is likely that in the coming years much more information will be uncovered about the mechanisms that allow fetal skin to heal perfectly. While the exact mechanisms behind scarless fetal healing are unknown, studies are beginning to show the feasibility of exploiting current knowledge about this process to benefit postnatal wound repair. Based on these studies, it is clear that learning more about regenerative healing in the fetus could have a tremendous impact on the future of wound care.