The revascularization process of human skin grafts
The origin of the blood vessels in a skin graft after transplantation has been subjected to debate for long time. It is not entirely clear whether the vascularization of adult skin grafts is achieved by penetration of new vessels from the host into the transplant (neovascularization), by anastomosis of host blood vessels with preexisting graft vessels, or by combination of these processes.
In the present article, we summarize data that have allowed to better understand the revascularization processes of normal human skin(#Demarchez et al., 1987) or of two types of in vitro reconstructed skin(#Demarchez et al., 1992) which are composed of human keratinocytes seeded either onto a acellular desepidermised dermis or on a lattice made of human fibroblasts embedded in a type I collagen gel, a living dermal equivalent.
In most studies, the ability to draw definite conclusions about the revascularization process of grafts was limited because of the lack of an analytical method with which components originating from the host or from the donor could be easily distinguished by specific markers. In the data reported here, we have performed double labeling immunofluorescence microscopy with cross-reacting or species-specific antibodies directed against endothelial components or type IV collagen, a major constituent of vascular basement membrane.
- 1. Origin of endothelial cells of human skin transplanted onto (...)
- 2. The process of revascularization of normal human skin grafted (...)
- 3. The revascularization of in vitro reconstructed skin
- 3.1 The skin reconstructed with human keratinocytes seeded on a (...)
- 3.2 The skin reconstructed with human keratinocytes seeded on a (...)
- 4. The revascularization of the dermis during wound healing of (...)
- 5. Conclusion
1. Origin of endothelial cells of human skin transplanted onto the nude mouse
In skin, the blood vessel wall is made of endothelial cells lying on a basement membrane. In a first step, we have determine whether the endothelial cells of the graft blood vessels were of mouse or human origin. For this purpose, we hava used a non-species-specific rabbit antibody directed against a factor VIII associated protein, which specifically labels endothelial cells. To determine the species origin (mouse or human), we have used mouse anticlonal antibodies directed against human cells. Thus, the first antibody allowed to localize the endothelial cells, and the second antibody allowed to determine wheher they were human. To specifically reveal these two antibodies, we have used an anti-rabbit antibody antibody coupled to rhodamin and an anti mouse antibody antibody coupled to fluorescein. Thus, the endothelial cells of both species appeared in red and the human endothelial cells appeared in red and green.
In normal human skin before grafting, we could verify that all the endothelial cells were stained with the anti- factor VIII associated protein and with antibodies directed against HLA-DR antigen, a MHC class II protein or against human vimentin, an intermediate filament of the endothelial cell cytoskeleton. One week after grafting onto the nude mouse, The HLA-DR antigen is only detected on a part of the graft endothelial cells. At two weeks after grafting, no endothelial cells were HLA-DR positive while few of them were positive for human vimentin.
These experiments seem to show that human endothelial cells of the graft progressively disappear and that one of the first degenerative signs is the loss of the HLA-DR antigen at their surface. Concomitantly, the mouse endothelial cells penetrate in the graft. This last observation is confirmed by progressive deosit of mouse type IV collagen inside the graft blood vessels.
Since the mouse endothelial cells appeared to migrate into the graft it appeared interesting to determine the process they use to make it. In vivo, endothelial cells are known to be surrounded by and to rest on a basement membrane composed of macromolecules including type IV collagen. When this work was performed it was admitted that the endothelial cells synthesize the type IV collagen of the basement membrane .
In the blood vessels of the grafted human skin, the human type IV collagen is detected at all stages after grafting. It seems therefore that while the human endothelial cells progressively disappear, the pre-existing basement membranes remain in place. We therefore decide to make double labeling with antibodies directed against factor VIII associated protein and directed against human type IV collagen.
Our working hypothesis was the following : if the graft endothelial cells remained constantly associated to th human type IV collagen, this would demonstrate that the mouse endothelial cells use the basement membrane of the pre-existing blood vessels of the graft to migrate into the grafted dermis. On the other hand, if this co-distribution endothelial cells- human type collagen was only partial or absent, this would indicate that the mouse endothelial cells use an other way to penetrate into the transplant.
The results were in favor of the first hypothesis, namely, that the graft endothelial cells are constantly co-distributed with human type IV collagen.
- In the blood vessels of the grafted skin, the endothelial cells positive for the factor VIII-associated protein are constantly associated with human type IV collagen.
To confirm this result, we made double labeling with species specific antibodies directed against human or mouse type IV collagen. In the graft blood vessels, the type IV collagens of the two species appear constantly co-distributed. This co-localisation is not observed at the dermal-epidermal junction of the grafted skin.
- In the blood vessels of the grafted skin, the human and mouse type IV collagens are constantly co-distributed, with the exception of more central areas in which mouse endothelial cells have not yet penetrated. At the dermal-epidermal junction, the human type IV collagen is only detected in the human skin and the mouse type IV collagen only in the mouse skin.
This last observation seemed to indicate that the mouse endothelial cells build their own basement membrane above the pre-existing basement membrane of the original blood vessels.
In summary, from these results, we could conclude that the revascularization of human skin transplanted onto the nude mouse probably proceeds according to the following mechanism:
1) there is anastomosis between the graft blood vessels and those of the host;
2) the human endothelial cells lose their specific markers and probably degenerate while the mouse endothelial cells penetrate into the grafted dermis by migrating on the basement membrane of the preexisting blood vessels of the grafted skin;
3) the mouse endothelial cells produce their own basement membrane above the pre-existing ones.
However, we could draw this conclusion from the previous data only if the endothelial cells were the only one to produce the type IV collagen of the blood vessel basement membrane. As we are going to see, this is not the case as it is observed by studying the vascularization of in vitro reconstructed skin grafts.
Two types of in vitro reconstructed skinhave been grafted onto the nude mouse (#Demarchez et al., 1992) . They are composed of human keratinocytes seeded either onto an acellular desepidermised dermis or on a lattice made of human fibroblasts embedded in a type I collagen gel, a living dermal equivalent.
In in vitro reconstructed skin composed of human keratinocytes seeded onto an acellular desepidermised dermis, all the dermal cells have been previously killed by successive freezings and thawings. The dermis before grafting is constituted of the extracellular matrix; in particular, human endothelial cells are absent and the blood vessels basement membranes are preserved.
By using the previously described antibodies, we could observe that, at the early stages, the mouse endothelial cells which were migrating into the grafted dermis, were constantly associated with the human type IV collagen of the preexisting basement membrane of the blood vessels. At later stages, this human type IV collagen was only detected at the derma-epidermal junction of the grafted reconstructed skin but was not detected in the basement membrane of the blood vessels while mouse type IV collagen was detected in both areas.
- One month after transplantation onto the nude mouse, the blood vessels of the grafted skin are positive only for the mouse type IV collagen. On the opposite, the dermal epidermal junction is positive for both human and mouse type IV collagen.
These observations are in agreement with those made with the grafts of normal human skin; after anastomosis of blood vessels of the two species, the mouse endothelial cells penetrate into the graft by migrating on the original basement membrane of the graft blood vessels. However, on the opposite of what we have observed with the normal human skin grafts in which mouse and human type IV were constantly co-distributed in the blood vessels, the human type IV collagen eventually disappears to be replaced by mouse type IV collagen.
3.2 The skin reconstructed with human keratinocytes seeded on a lattice made of human fibroblasts embedded into type I collagen
The dermis of the skin reconstructed with a living dermal equivalent contains human fibroblasts embedded in a bovine type I collagen gel but does not include blood vessels and therefore contains no human endothelial cells.
After grafting, the transplant is progressively revascularized by penetration of blood vessels from the host (=inosculation). These blood vessels contain mouse endothelial cells labeled with the antibody directed against the factor VIII associated protein and their basement membrane are, as expected, positive for mouse type IV collagen. However, they were also labeled by the antibody directed against human type IV collagen. Since the only human cells present in the human dermal equivalent were human fibroblasts, this last result indicates that graft human fibroblasts interact with the mouse endothelial cells that penetrate into the graft and produce human type IV collagen in the basement membrane of the newly formed blood vessels.
The fibroblastic origin of this human type IV collagen is in agreement with the observation made with dermal equivalents in which the human fibroblasts have been killed by successive freezings and thawings: in this case, in the absence of human fibroblasts, no human type IV collagen were observed in the blood vessels of the dead dermal equivalent after grafting.
Wound healing of human skin transplanted onto the nude mouse is the subject of another article on this Web site. In the context of the present article, some observations are of interest. When a wound is made through the entire thickness of a human skin that has been transplanted for two months onto the nude mouse, the human skin will entirely regenerate. The reconstruction of the dermal part occurs in two steps. First, a granulation tissue of mouse origin is produced in the wound under the crust; second, after reepithelialization, it is replaced by a human neodermis produced by human fibroblasts that have migrated from the surrounding noninjured human dermis.
No human type IV collagen was associated with the blood vessels in the granulation tissue while mouse type IV collagen is detected. However, mouse and human type IV collagens were codistributed in the vascular basement membrane of the human neodermis.
- In the blood vessels of the neodermis, the human or mouse type IV collagens are co-localized while at the dermal-epidermal junction, only human type IV collagen is observed.
In this last situation, it is probable that the human fibroblasts that have migrated from the noninjured dermis to build the neodermis, by interacting with mouse endothelial cells are induced to produce human type IV collagen. This result is in agreement with the observation made with the skin reconstructed with a living dermal equivalent.
From all these data, it can be concluded that
-* the dermal fibroblasts participate to the construction of the blood vessel basement membrane; the local production of the basement membrane components such as type IV collagen by the fibroblasts would occur following to their interaction with endothelial cells of the blood vessels; in the other areas of the dermis, the fibroblasts mainly produce the dermal extracellular matrix mainly composed of type I or III collagens. We will see in another article that the fibroblasts also produce basement membrane components at the dermal-epidermal junction.
-* the revascularization of human skin grafts begins by anastomosis of host blood vessels with those of the graft if they are present.
-* in the model of human skin grafted onto the nude mouse, the human skin is vascularized by mixed blood vessels made of mouse endothelial cells resting on a basement membrane which is partly synthesize by these cells and for another part, is produced by surrounding human fibroblasts
Demarchez M, Hartmann DJ, Regnier M, Asselineau D. The role of fibroblasts in dermal vascularization and remodeling of reconstructed human skin aftertransplantation onto the nude mouse. Transplantation. 1992 Aug;54(2):317-26.