Fibrosin, a novel fibrogenic cytokine, modulates expression of myofibroblasts

Granulation tissue fibroblasts, or myofibroblasts are characterized by the presence of alpha smooth muscle actin fibers (alpha SMA). These specialized cells are involved in wound contraction and in retractile phenomena observed during fibrotic disease. Myofibroblasts have also been shown to play a role in embryonic development. Growth factors such as Transforming growth factor beta TGFbeta and Nerve growth factor (NGF) can modulate the differentiation of myofibroblasts. In this report, we show that in vitro application of fibrosin, a novel fibrogenic cytokine, stimulates expression of alpha SMA-producing cells at least four-fold above that observed in control cultures. In addition, administration of fibrosin in a wound healing model in mice stimulates increased numbers of myofibroblasts 7 days after injury, when compared with untreated, or, control, wounded mice. These results suggest that fibrosin plays an important role in up regulating the appearance of myofibroblasts during wound healing, and possibly in fibrotic diseases. It may, therefore, be important in the process of scarring.     

瘢痕组织中α-平滑肌肌动蛋白的表达与细胞凋亡的关系

目的:观察α-平滑肌肌动蛋白在瘢痕组织中的表达,了解病理性瘢痕形成过程中凋亡所扮演的角色及其与肌成纤维细胞在真皮中变化的关系。方法:瘢痕标本来自烧伤后来我院进行整形手术的病人,同时取病人手术供皮区的正常皮肤作为对照。 8例瘢痕组织标本 (含 2例愈合较为平坦的瘢痕和 6例增生性瘢痕组织)被分成增殖期和成熟期两组。运用caspase-3mRNA及其蛋白的表达及TUNEL方法检测瘢痕及正常组织中的凋亡细胞,并以免疫组化法检测瘢痕及正常皮肤真皮内α-平滑肌肌动蛋白单克隆抗体的表达。结果:瘢痕组织中细胞凋亡的数目与正常组织明显不同。瘢痕内的TUNEL标记阳性细胞数多于正常组织;增殖期瘢痕内的细胞凋亡的数目多于成熟期。增殖期TUNEL标记阳性的细胞多于平坦瘢痕,而成熟期两者无显著差别,Caspase-3mRNA及其蛋白的表达与TUNEL标记结果具有一致性。随着瘢痕组织的成熟,α-平滑肌肌动蛋白单克隆抗体的表达逐渐降低,平坦的瘢痕组织中的表现尤为明显;增生性瘢痕中,增殖期与成熟期之间无显著差别。结论:正常伤口愈合过程中,肌成纤维细胞暂时性的表达,可引起伤口的收缩,随着真皮再塑形,含有α-平滑肌肌动蛋白的肌成纤维细胞因凋亡而消失,而病理性的愈合结局可能是它持续表达的结果。     

The myofibroblast in wound healing and fibrocontractive diseases

The demonstration that fibroblastic cells acquire contractile features during the healing of an open wound, thus modulating into myofibroblasts, has open a new perspective in the understanding of mechanisms leading to wound closure and fibrocontractive diseases. Myofibroblasts synthesize extracellular matrix components such as collagen types I and III and during normal wound healing disappear by apoptosis when epithelialization occurs. The transition from fibroblasts to myofibroblasts is influenced by mechanical stress, TGF-beta and cellular fibronectin (ED-A splice variant). These factors also play important roles in the development of fibrocontractive changes, such as those observed in liver cirrhosis, renal fibrosis, and stroma reaction to epithelial tumours.     

Scar‐Free Wound Healing and Regeneration Following Tail Loss in the Leopard Gecko,Eublepharis macularius

Many lizards are able to undergo scar‐free wound healing and regeneration following loss of the tail. In most instances, lizard tail loss is facilitated by autotomy, an evolved mechanism that permits the tail to be self‐detached at pre‐existing fracture planes. However, it has also been reported that the tail can regenerate following surgical amputation outside the fracture plane. In this study, we used the leopard gecko, Eublepharis macularius, to investigate and compare wound healing and regeneration following autotomy at a fracture plane and amputation outside the fracture plane. Both forms of tail loss undergo a nearly identical sequence of events leading to scar‐free wound healing and regeneration. Early wound healing is characterized by transient myofibroblasts and the formation of a highly proliferative wound epithelium immunoreactive for the wound keratin marker WE6. The new tail forms from what is commonly referred to as a blastema, a mass of proliferating mesenchymal‐like cells. Blastema cells express the protease matrix metalloproteinase‐9. Apoptosis (demonstrated by activated caspase 3 immunostaining) is largely restricted to isolated cells of the original and regenerating tail tissues, although cell death also occurs within dermal structures at the original‐regenerated tissue interface and among clusters of newly formed myocytes. Furthermore, the autotomized tail is unique in demonstrating apoptosis among cells adjacent to the fracture planes. Unlike mammals, transforming growth factor‐β3 is not involved in wound healing. We demonstrate that scar‐free wound healing and regeneration are intrinsic properties of the tail, unrelated to the location or mode of tail detachment. Anat Rec, 2012. © 2012 Wiley‐Periodicals, Inc.     

Wound healing in embryos: a review

 Skin wounds in young embryos heal rapidly, efficiently and perfectly without scar formation, an ability that is lost as development proceeds. The tissue movements of repair (re-epithelialisation and connective tissue contraction) are the same in embryos as adults, but the means by which these movements are achieved are very different. Whilst adult wound front epidermal cells crawl forwards over the exposed substratum to close a defect, a gap in the embryonic epidermis is closed by contraction of a rapidly assembled actin purse-string. In the adult wound situation connective tissue contraction is brought about by specialist contractile myofibroblasts, but in the embryo standard embryonic fibroblasts exert similar tractional forces to bring the wound margins together. We review what is known of the cellular sources and the levels of various growth factor signals that might activate wound closure movements during embryonic and adult repair and how this knowledge might help in designing therapeutic strategies to enhance adult healing. Finally we discuss how studies of the tissue movements of embryonic wound healing may guide our understanding of more natural tissue movements that occur during embryogenesis, such as gastrulation and neurulation. Accepted: 11 October 1996     

Morphologic examination of mesenchymal cells in healing wounds of normal and tight skin mice.

The healing process of an open wound as effected by wound contraction is complete by 3 weeks in the normal mouse. In contrast, its onset is delayed by 3 weeks and complete healing requires 6 weeks in the tight skin mouse (TSM), a mutant mouse strain with the autosomal dominant gene for tight skin. Possible mechanisms for this delay were evaluated. The frequency and distribution of myofibroblasts were studied during the 3-week delay in wound contraction by actin staining and electron microscopy. It was determined, by electron microscopy and phalloidin staining, that myofibroblasts were found in high density in noncontracting TSM wounds. Electron microscopy showed, however, that these myofibroblasts were surrounded by a pericellular matrix that separated their surface from adjacent collagen fibers. No pericellular matrix was found around cells in granulation tissue of normal mice. At 3 weeks, as TSM wounds began to contract, the number and intensity of cells stained by phalloidin in this tissue was less than that seen earlier. The pericellular matrix was fragmented at this time, and cell surface and collagen fiber associations were apparent. Finally, at 5 weeks, when wound contraction was well developed in the TSM, only a small area in the center of the healing wound beneath the epidermis contained phalloidin-positive myofibroblasts. Electron-microscopic examination of the residual granulation tissue at this time revealed the complete absence of the pericellular matrix. It is postulated that during the 3-week delay in wound closure, the presence of a localized pericellular matrix prevents the interaction between cells and collagen fibers necessary for the reorganization of collagen. It is also thought that the tightly adherent uninjured skin surrounding the healing wound may cause delayed wound closure. There was no evidence that the absence of myofibroblasts is responsible for delayed wound contraction.     

A comparative study of fibroblasts in healing freeze and burn injuries in rats.

In rats, the healing process of a full-thickness dermal freeze injury differs from that of a burn wound. Whereas burn wounds heal by wound contraction, the movement of surrounding normal skin over the defect, freeze wounds heal without wound contraction. That absence of contraction may be due to the freeze wound's lack of myofibroblasts, the cells reportedly associated with wound contraction. Myofibroblasts can be demonstrated histologically by staining the F-actin filaments of the stress fibers with NBD-phallacidin, a fluorescent reagent specific to F-actin filaments. Fibroblasts in normal dermis have no staining stress fibers. However, staining myofibroblasts are uniformly distributed in the granulation tissue of the healing burn and in the islands of granulation tissue between residual connective tissue fibers in the healing freeze wound. These residual dermal fibers were identified by their patterns of birefringence. Residual connective tissue matrix persists following cold trauma and acts like an internal splint. Burn trauma destroys cells and the connective tissue matrix, which is completely replaced with granulation tissue which undergoes wound contraction. Freeze trauma kills the cellular components of dermis, while some residual connective tissue fibers endure. This study shows that the connective tissue matrix can play an important role in the control of wound contraction.     

Cellular retinol-binding protein-1 is transiently expressed in granulation tissue fibroblasts and differentially expressed in fibroblasts cultured from different organs.

We have reported that cellular retinol-binding protein-1 (CRBP-1) is transiently expressed by arterial smooth muscle cells during experimental intimal repair (P. Neuville, A. Geinoz, G. Benzonana, M. Redard, F. Gabbiani, P. Ropraz, G. Gabbiani: Am J Pathol 1997, 150:509-521). We have examined here the expression of CRBP-1 during wound healing after a full-thickness rat skin wound. CRBP-1 was transiently expressed by a significant proportion of fibroblastic cells including myofibroblasts. Expression started 4 days after wounding, reached a maximum at 12 days, and persisted up to 30 days when a scar was formed. After wound closure, most CRBP-1-containing fibroblastic cells underwent apoptosis. We have further investigated CRBP-1 expression in rat fibroblasts cultured from different organs. CRBP-1 was abundant in lung and heart fibroblasts and was detected in decreasing amounts in muscle, tendon, subcutaneous tissue, and granulation tissue fibroblasts. Dermis fibroblasts contained no detectable levels of CRBP-1. All-trans retinoic acid and transforming growth factor-beta1 inhibited cell proliferation and increased CRBP-1 expression in fibroblastic populations except dermis fibroblasts. We demonstrate that during granulation tissue formation a subpopulation of fibroblastic cells express CRBP-1 de novo. We also demonstrate that CRBP-1 expression by fibroblasts is regulated in vitro by retinoic acid and transforming growth factor-beta1. Our results suggest that CRBP-1 and possibly retinoic acid play a role in the evolution of granulation tissue.     

An ex vivo model for functional studies of myofibroblasts

Migration, proliferation and invasive growth of myofibroblasts are key cellular events during formation of granulation tissue in situations of wound healing, arteriosclerosis and tumor growth. To study the invasive phenotype of myofibroblasts, we established an assay where arterial tissue from chicken embryos was embedded in fibrin gels and stimulated with growth factors. Addition of serum, PDGF-BB and FGF-2, but not VEGF-A, resulted in an outgrowth of cellular sprouts with a pattern that was similar to the organization of cells invading a provisional matrix in an in vivo model of wound healing using the chicken chorioallantoic membrane. Sprouting cells were defined as myofibroblasts based on being alpha-smooth muscle actin-positive but desmin-negative. There was no contribution of endothelial cells in outgrowing sprouts. The acquired myofibroblastic phenotype was stable since sprout-derived cells resumed sprouting in a growth factor-independent manner when re-embedded as spheroids in a fibrin matrix. Invasive growth and sprouting of vascular smooth muscle cells was not limited to chicken cells since a similar response was seen when spheroids composed of purified primary human aortic smooth muscle cells were embedded in fibrin. Finally, a technique for flat visualization of the three-dimensional sprouting and a quantification method is described. This ex vivo model allows quantitative analysis of invasive growth and differentiation of vascular smooth muscle cells and fibroblasts into myofibroblasts.     

Stromal changes in invasive breast carcinoma: an ultrastructural study

The stroma in infiltrating breast carcinomas, with particular reference to stromal spindle cells, has been studied by electron microscopy. A mixture of cells including resting fibroblasts, active fibroblasts, early myofibroblasts, and mature myofibroblasts has been identified. In loose stroma, myofibroblasts possessed prominent organelles and showed secretory products along the cell surface, whereas in dense stroma, there was relative prominence of cytoplasmic filaments as well as other features consistent with a contractile state of myofibroblasts. The degree of myofibroblastic proliferation was related to the growth pattern of the tumour. It is suggested that the infiltrating process of cancer cells is analogous to wound production and healing with continuous granulation tissue and scar formation resulting in the characteristic desmoplastic reaction seen in certain breast carcinomas.     

Dense fibrillar collagen matrices: a model to study myofibroblast behaviour during wound healing

Fibroblastic cells play an important part in wound healing. Human dermal fibroblasts seeded onto three-dimensional fibrillar collagen matrices migrate into the collagen network and differentiate into myofibroblasts. In order to evaluate the use of collagen matrices as model systems for studying myofibroblast phenotype during wound healing, myofibroblast behaviour migrating into dense or loose matrices was compared. The effect of collagen concentration on cell morphology, remodelling, proliferation and apoptosis of human myofibroblasts was evaluated. Myofibroblasts within dense collagen matrices (40 mg/ml) were spindle shaped, similar to cells observed during tissue repair. In contrast, cells within loose matrices (5mg/ml) were more rounded. Matrix hydrolysis activities (MT1-MMP and MMP2) did not differ between the two collagen concentrations. The myofibroblast proliferation rate was measured after 24h bromodeoxyuridine incorporation (BrdU). Cells in dense collagen matrices proliferated at a higher rate than cells in loose matrices at each culture time point tested. For example, 40% of cells in dense matrices were replicating compared to 10% of cells in loose matrices after 28 days in culture. Apoptotic cells were only detected in dense matrices from day 21 onwards when cells had already migrated into the collagen network. Taken together, these results show that a high collagen concentration has a stimulatory effect on myofibroblast proliferation and apoptosis, two important events in wound healing. Thus, dense matrices can be used to create controlled conditions to study myofibroblast phenotype.     

Heterogeneity of myofibroblast phenotypic features: an example of fibroblastic cell plasticity

Granulation tissue fibroblasts (myofibroblasts) develop several ultrastructural and biochemical features of smooth muscle (SM) cells, including the presence of microfilament bundles and the expression of -SM actin, the actin isoform present in SM cells and myoepithelial cells and particularly abundant in vascular SM cells. Myofibroblasts have been suggested to play a role in wound contraction and in retractile phenomena observed during fibrotic diseases. When contraction stops and the wound is fully epithelialized, myofibroblasts containing -SM actin disappear, probably as a result of apoptosis, and the scar classically becomes less cellular and composed of typical fibroblasts with well-developed rough endoplasmic reticulum but with no more microfilaments. In contrast, -SM actin expressing myofibroblasts persist in hypertrophic scars and in fibrotic lesions of many organs, including stroma reaction to epithelial tumours, where they are allegedly involved in retractile phenomena as well as in extracellular matrix accumulation. The mechanisms leading to the development of myofibroblastic features remain to be investigated. In vivo and in vitro investigations have shown that -interferon exerts an antifibrotic activity at least in part by decreasing -SM actin expression whereas heparin increases the proportion of -SM actin positive cells. Recently, we have observed that the subcutaneous administration of transforming growth factor-1 to rats results in the formation of a granulation tissue in which -SM actin expressing myofibroblasts are particularly abundant. Other cytokines and growth factors, such as platelet-derived growth factor, basic fibroblast growth factor and tumour necrosis factor-, despite their profibrotic activity, do not induce -SM actin in myofibroblasts. In conclusion, fibroblastic cells are relatively undifferentiated and can assume a particular phenotype according to the physiological needs and/or the microenvironmental stimuli. Further studies on fibroblast adaptation phenomena appear to be useful for the understanding of the mechanisms of development and regression of pathological processes such as wound healing and fibrocontractive diseases.     

Ultrastructural and Immunohistochemical Analysis of Intestinal Myofibroblasts During the Early Organogenesis of the Human Small Intestine

Intestinal myofibroblasts (IMFs), also known as pericryptal fibroblasts, are found at the basement membrane of the intestinal epithelium. They are characterized by well‐developed endoplasmic reticulum, cytoplasmic fibers, and fibrous extensions called fibronexi. IMFs have structural features in common both with fibroblasts and smooth cells. Vimentin, desmin, and α‐smooth‐muscle actin (α‐SM) are markers commonly used to discriminate between IMFs and smooth muscle cells. Immunohistochemical studies have shown that, when α‐SM and vimentin are positive in both IMFs and smooth muscle cells, desmin is negative in IMFs but positive in smooth muscle cells. In the adult intestine, IMFs play an important role in various functions, especially in tissue repair and scar formation during wound healing. In the embryonic intestine, however, wound healing does not occur, and to date, no studies have investigated the first appearance and subsequent evolution of IMFs. In this study, we have examined the human small intestine in embryos at 7, 9, and 11 weeks of development by ultrastructural and immunohistochemical analysis to shed light on the formation of IMFs during these early phases of organogenesis. At 7 weeks, the embryonic mesenchymal cells are similar to proto‐myofibroblasts and may be the precursors of the IMFs detected at 9 weeks and more abundantly at 11 weeks by immunohistochemistry. These IMFs seem to mediate information flow between the epithelium and the mesenchyme and thus contribute to the development of the small intestine. Anat Rec, 2011. © 2011 Wiley‐Liss, Inc.     

Relationship between the distribution of myofibroblasts,and stellar and circular scar formation due to the contraction of square and circular wound healing

Square skin wounds can heal to form a stellar scar with four protrusions at the four angles, whereas circular wounds can heal to form an ellipsoid scar. It is not clear why these differences occur and the aim of the present study was to clarify this phenomenon. Two square or circular full-thickness skin wounds were made on the dorsum of mice, and covered with hydrocolloid dressing. They were observed from day 0 to 15 after wounding, and used to prepare paraffin sections stained with anti-alpha-smooth muscle actin antibody to detect myofibroblasts. The square wound was transiently enlarged by edema and skin tension on day 3, at which time the angles became round, and thus the square form became more circular. Thereafter, the wound contracted rapidly and the circular form was maintained until day 11. On day 11 distinct angles appeared where the scar formation had progressed further, and there were fewer myofibroblasts than in any other section. A stellar scar with protrusions from the four angles was formed on day 15, when myofibroblasts almost disappeared in the protrusions. This indicates that due to the earlier disappearance of myofibroblasts and earlier scarring in the angles of the square wound, the scar angle cannot be pulled into the center of the wound but residual myofibroblasts on the side can pull the side into the center due to myofibroblastic contraction and consequently a stellar scar is formed. Thus, the earlier disappearance of myofibroblasts in the angles is very important for the formation of stellar scars.     

Evaluation of a collagen-glycosaminoglycan dermal substitute in the dog palate

Tissue shortage complicates surgery of cleft lip and palate. The healing of defects on the palate impairs growth of the dentoalveolar complex because of scar tissue formation. Implantation of a matrix into the wound might overcome this adverse effect. Integra with and without a silicone top layer was implanted into standardized full-thickness wounds (? 6 mm) in the palatal mucoperiosteum in beagle dogs. In some wounds, the silicone layer was removed after 14 days. Control wounds did not have an implant. At 2 and 4 weeks post-surgery, the wounds were assessed for epithelialization, inflammation (hematoxylin and eosin, leucocyte protein L1), number of myofibroblasts (alpha smooth muscle actin), and general histological characteristics. Wounds filled with Integra without the silicone layer showed fewer myofibroblasts and inflammatory cells than the sham wounds. Collagen fibers were more randomly orientated in these wounds than in the sham group. Wound closure was found to be retarded, and many inflammatory cells were present when Integra with silicone was implanted. The silicone layer was lost within 4 weeks in these wounds. We conclude that, in the moist oral environment, the silicone of Integra is not required. Re-epithelialization and tissue integration proceed more favorably without it. Further research in the dentoalveolar development with Integra will be conducted in a simulated cleft palate repair in the dog model.     

The dipeptide Lys-Pro restores the diminished wound healing following treatment with anti-T-helper cell monoclonal antibody

To determine the role of T-cell subsets in wound healing, we studied the granulation tissue proliferation after depletion of CD4 or CD8 positive cells. Granulation tissue proliferation in CD8-diminished AB mice was significantly higher than in the control group as a result of depleted suppressor cell activity. The CD4-depleted mice produced granulation tissue in less than 30% of the control group. To investigate the role of dipeptidylpeptidase IV on CD4 positive cells in wound healing we used Lys-[Z(NO2)]-Pro, a chemically modified dipeptide which may result by degradation of polypeptides by this peptidase. It was possible to restore the diminished capability of granulation tissue proliferation in CD4-depleted mice by a single treatment with the dipeptide Lys-[Z(NO2)]-Pro. Our results suggest that: (i) the CD4 positive T-helper subset regulates wound healing, and (ii) products of degradation by dipeptidylpeptidase IV may bypass T-helper cell function.     

Scar‐free cutaneous wound healing in the leopard gecko,Eublepharis macularius

Cutaneous wounds heal with two possible outcomes: scarification or near‐perfect integumentary restoration. Whereas scar formation has been intensively investigated, less is known about the tissue‐level events characterising wounds that spontaneously heal scar‐free, particularly in non‐foetal amniotes. Here, a spatiotemporal investigation of scar‐free cutaneous wound healing following full‐thickness excisional biopsies to the tail and body of leopard geckos (Eublepharis macularius) is provided. All injuries healed without scarring. Cutaneous repair involves the development of a cell‐rich aggregate within the wound bed, similar to scarring wounds. Unlike scar formation, scar‐free healing involves a more rapid closure of the wound epithelium, and a delay in blood vessel development and collagen deposition within the wound bed. It was found that, while granulation tissue of scarring wounds is hypervascular, scar‐free wound healing conspicuously does not involve a period of exuberant blood vessel formation. In addition, during scar‐free wound healing the newly formed blood vessels are typically perivascular cell‐supported. Immunohistochemistry revealed widespread expression of both the pro‐angiogenic factor vascular endothelial growth factor A and the anti‐angiogenic factor thrombospondin‐1 within the healing wound. It was found that scar‐free wound healing is an intrinsic property of leopard gecko integument, and involves a modulation of the cutaneous scar repair program. This proportional revascularisation is an important factor in scar‐free wound healing.     

Effects of angiotensin II receptor signaling during skin wound healing

The tissue angiotensin (Ang) system, which acts independently of the circulating renin Ang system, is supposed to play an important role in tissue repair in the heart and kidney. In the skin, the role of the system for wound healing has remained to be ascertained. Our study demonstrated that oral administration of selective AngII type-1 receptor (AT(1)) blocker suppressed keratinocyte re-epithelization and angiogenesis during skin wound healing in rats. Immunoprecipitation and Western blot analysis indicated the existence of AT(1) and AngII type-2 receptor (AT(2)) in cultured keratinocytes and myofibroblasts. In a bromodeoxyuridine incorporation study, induction of AT(1) signaling enhanced the incorporation into keratinocytes and myofibroblasts. Wound healing migration assays revealed that induction of AT(1) signaling accelerated keratinocyte re-epithelization and myofibroblasts recovering. In these experiments, induction of AT(2) signaling acted vice versa. Taken together, our study suggests that skin wound healing is regulated by balance of opposing signals between AT(1) and AT(2).