A scaffold with regenerative activity was initially synthesized as a highly porous graft copolymer of type I collagen and chondroitin 6-sulfate (GAG). It was modified later by omitting the GAG. Regenerative activity depends on optimization of pore diameter and half-life for degradation; a minimal density of ligands for integrins α1β1 and α2β1 is also required on the collagen surface. When the structure has been appropriately optimized to yield a so-called dermis regeneration template (DRT), the scaffold induces regeneration of skin (guinea pig, swine, human) peripheral nerves (rat) and of the conjunctiva (rabbit). (Photo by Alexandra Kourgiantaki)
Several examples of regeneration in spontaneously healing wounds (no scaffold present) in various species support the contraction blockade theory. Such examples include the competition between contraction and regeneration for closure of skin wounds in the developing tadpole; regeneration of skin in the rabbit ear, related to tight binding of skin to the underlying cartilage which precludes closure by contraction; regeneration of the injured oral mucosa in mice and swine showing greatly reduced scar and contraction compared to healing in skin wounds; and regeneration of skin in the axolotl, associated with the reduced evidence of contraction.
The dermis regeneration template (DRT) is a temporarily insoluble scaffold with proven regenerative activity both experimentally and clinically (Fig. 10.1; Chaps. 5, 6). The vast majority of collagen scaffolds lack such activity. Three structural features are critically required for a contraction blockade by this scaffold. The optimal pore size of DRT is in the range 20–125 µm and the optimal degradation half-life is 2 ± 1 weeks (14 ± 7 days; Chap. 8). Binding of contractile cells via integrins α1β1 and α2β1 to the scaffold surface appears to be required for implementation of the contraction blockade. However, insufficient data are available to identify optimal values for ligand densities on the collagen surface , primarily GFOGER and GLOGEN, that populate naturally the collagen surface (Chap. 9). Using collagen libraries it has been shown that the regenerative activity of DRT practically disappears when these three key features of the scaffold structure deviate from optimal levels (Chap. 9).
An efficient contraction blockade is mounted when these three structural parameters of DRT are present at optimal levels. In the presence of the active collagen scaffold, the myofibroblast density is significantly reduced, the tight assembly of contractile cells normally observed in healing wounds is dispersed, and the axial orientation of cells becomes randomized. These profound changes in morphology of contractile cells appear to explain adequately the observed cancellation or near cancellation of the macroscopic force of wound contraction in the presence of the scaffold. Specific binding of contractile cells on the temporarily insoluble scaffold surface via integrin–ligand interactions as well as provision of a large enough surface through optimization of the pore diameter partly account for the observed contraction blockade. The scaffold half-life requires optimization as well in order to make contact with contractile cells over the period, 1–4 weeks, when contractile cells appear in the wound. Reduction in TGβ1 concentration is probably accounted for by the observed high-binding affinity of the cytokine to the collagen surface.
Impaired wounds close without contraction but also without regeneration. These examples include wounds in diabetics as well as pharmacologically impaired wounds; also included are wounds that were mechanically splinted. These examples of impaired healing show that a contraction blockade is not sufficient to induce regeneration .
Although required in the presence of the collagen scaffold, a contraction blockade does not provide by itself for synthesis of new stroma or of epithelial tissue. Stroma is synthesized by fibroblasts and differentiated fibroblasts in close contact with the scaffold which appears to provide steric guidance for the synthesis of collagen fibers. Somewhat speculatively, the new connective tissue is therefore a partial topological replica of the scaffold. In the absence of the mechanical field of a normally contracting wound, the cells are disoriented and synthesize connective tissue with collagen fibers that are largely oriented randomly in space, as in the physiological dermis , rather than being highly aligned, as in scar. Synthesis of stroma probably results from the interaction of two processes both of which provide spatial cues: The scaffold acts as a spatial template that breaks up cell–cell binding and instead binds cells on its surface, thereby assigning specific locations to the cells; and by cancelling the mechanical field of contraction it leads to randomization of axes of the cells that synthesize collagen fibers.
The two steps that comprise the overall process of induced regeneration, i.e., contraction blockade and stroma synthesis, have been shown capable of proceeding independently, as shown in an island graft experiment using DRT (Orgill and Yannas 1998).
Synthesis of the new epithelium is based on the spontaneous ability of epithelial cells to induce regeneration of parent tissue provided stroma is present. If a sufficient number of endogenous epithelial cells have survived the injury , they proliferate and mature into functional epithelial tissue using the surface of the new stroma as a spatial guide (Chap. 5; see also below in section on decellularized matrices). Other processes, not clearly understood at this time, lead to angiogenesis, neurogenesis, and completion of the remainder of regenerative processes for the organ.
10.4 Similarity Between Scarless Fetal Healing and Adult Healing in the Presence of DRT
Scarless healing of wounds in the early mammalian fetus has remained an enigma. The reviews of efforts to interpret data on fetal healing emphasize the difficulties involved and authors generally conclude that the precise mechanisms regulating scarless fetal healing remain unknown (Lin et al. 2010; Rolfe and Grobbelaar 2012; Leung et al. 2012; Lo et al. 2012; Ud-Din et al. 2014). The difficulty in reaching consensus is probably related to the challenging experimental conditions for studying healing in the fetus; results from the use of apparently similar protocols lead to conclusions that are occasionally contradictory. The available evidence on scarless mammalian fetal healing will be reviewed below in an effort to identify differences in the healing processes that occur spontaneously with fetal development.
An experimental variable that requires careful control in studies of fetal wound healing is the timing of the study relative to the onset of the ontogenetic transition from scarless healing to healing with scar (scarring; Fig. 10.2). This transition has been established in a small number of species and appears to be located at about two thirds of gestation period (Estes et al. 1994; Beanes et al. 2002; Soo et al. 2003). The incidence of scarless healing decreased dramatically after the ontogenetic transition from midgestational (“early”) to late-gestational (“late”) fetal healing (Beanes et al. 2002).
An ontogenetic transition in the mammalian fetus between scarless healing (regeneration) and healing with scar (repair) has been studied in a number of fetal models. It occurs at approximately two thirds of gestation time
An effective way of controlling for the developmental variable is to compare healing in fetuses with known gestational age that were wounded before and after the transition; or study the concentration kinetics of a cytokine or of another feature with wound healing time. Another approach is study of a model that spontaneously heals scarlessly by introducing an experimental variable that changes healing to a scarring outcome in animals of the same gestational age; or start with a scarring model and attempt to induce the reverse process. Relevant experimental evidence has been selected below on the basis of these criteria and will be summarized very briefly. In order to limit the discussion to anatomically well-defined skin wounds (Chap. 3), we will consider wounds that were produced by full-thickness excision. Results from incisional wound models in the adult (Shah et al. 1994; Liu et al. 2003) or the fetus (Cass et al. 1997b) require a qualitatively different interpretation from results with excisional wound models (Chap. 3) and will not be reviewed in the same context. Skin wounds smaller than 2 mm diameter in the lamb fetus, including incisional wounds, healed without scar while larger wounds healed with scar (Cass et al. 1997a, b). Control of wound size is, therefore, a critical experimental variable that requires control. There have been several reports implicating a number of molecules in scarless repair, including metalloproteinases (Dang et al. 2003), prostaglandin E2 (Parekh et al. 2009), decorin (Beanes et al. 2001; Järvinen and Ruoslahti 2013), basic fribroblast growth factor (Abe et al. 2012), fibromodulin (Zheng et al. 2014), as well as others. We will focus instead below on studies emphasizing the TGFβ1 presence in fetal wounds since there is considerable independent evidence that this cytokine is involved in scarring in a variety of animal models (see below) .
An early study of expression and clearance of TGFβ1, known to be required for myofibroblast (MFB) differentiation (Desmouliere et al. 2005), showed a rapid induction of TGFβ1, in scarless skin wounds in the fetal mouse, within 1 h postwounding; however, the cytokine was cleared rapidly from the wound site, leading to background levels by 18 h (Martin et al. 1993). In the fetal lamb, αSMA, the phenotype used to identify myofibroblasts (Hinz et al. 2012), cells credited with playing a major role during wound contraction (Gabbiani 1998; Daimon at al. 2013), was absent in wounds of early gestational age but was present to a progressively greater extent through late gestation (Estes et al. 1994). In another study, the effect of topical addition of TGFβ1 to wounds in the fetal rabbit was compared to the untreated control. In the rabbit model, the full-thickness skin wounds in the dorsum are known to initially expand following injury. Addition of TGFβ1 led to a much lower expansion (higher contraction) of wounds, increased density of fibroblasts, and presence of αSMA-staining cells in TGFβ1-treated but not in untreated wounds (Alaish et al. 1996). In the fetal lamb model, it was observed that αSMA was not expressed in 2-mm diameter wounds or incisional wounds that were observed to heal scarlessly (Cass et al. 1997a); however, with larger wounds that healed with scar, αSMA expression increased with wound size and myofibroblasts were observed in all wounds that expressed αSMA. There was lack of myofibroblasts in wounds that healed scarlessly while abundant myofibroblasts were observed in scarring wounds (Cass et al. 1997b). Exogenous addition of TGFβ1, by sustained release of discs implanted in the subcutaneous tissue next to the dorsal wound in the fetal rabbit, led to increased wound contraction (the untreated control in the rabbit model expanded, as expected), increased fibrosis, and increased procollagen expression (Lanning et al. 1999). Working with the same fetal rabbit model and using exogenous addition of TGFβ1 by sustained release next to the wound led to increased wound contraction compared to the untreated control as well as significant increase in staining for αSMA (Lanning et al. 2000).
The ontogenetic transition from scarless to scarring healing was defined more clearly in a study of fetal rats of gestational age 14.5 days (denoted E14 in the study), 16.5 days (E16), and 18.5 days (E18) (term = 21.5 days). E14 rats did not survive in the study. Wounds placed in E16 rats healed without scar, with regeneration of a normal dermis and epidermal appendages, while E18 rats healed with collagen scar formation and without hair follicles (Beanes et al. 2002). Using a similar rat fetal model, it was observed that, along the ontogenetic transition from E16 to E19 rats, TGFβ1 and TGFβ2 concentration levels became higher and cleared more slowly, while TGFβ3 levels were decreased and were more prolonged. Expression of TGFβ-receptors type I and II was also increased and their expression was prolonged in E19 rats (Soo et al. 2003). Focusing on TGFβ1 in this discussion, rather than its other two isoforms, it was observed that the transition from a scarless to a scarring model was accompanied by an increase in TGFβ1 concentration which also persisted longer in the wound after the transition (Soo et al. 2003).
In summary, these controlled studies with excisional skin wounds in various fetal models provided evidence that scarless healing differs from scarred healing in the following respects: In scarless wounds, TGFβ1 concentration was lower and cleared more rapidly, showed decreased density of fibroblasts, and wounds were free of αSMA-staining cells (myofibroblasts; Estes et al. 1994; Cass et al. 1997b; Soo et al. 2003). Addition of exogenous TGFβ1 led to increased wound contraction and increased density of αSMA-staining cells, apparently converting scarless wounds to scarring wounds (Alaish et al. 1996; Lanning et al. 1999, 2000).