Mechanically challenged tissue-engineered organs, such as blood vessels, traditionally relied on synthetic or modified biological materials for structural support. In this report, we present a novel approach to tissue-engineered blood vessel (TEBV) production that is based exclusively on the use of cultured human cells, i.e., without any synthetic or exogenous biomaterials. Human vascular smooth muscle cells (SMC) cultured with ascorbic acid produced a cohesive cellular sheet. This sheet was placed around a tubular support to produce the media of the vessel. A similar sheet of human fibroblasts was wrapped around the media to provide the adventitia. After maturation, the tubular support was removed and endothelial cells were seeded in the lumen. This TEBV featured a well-defined, three-layered organization and numerous extracellular matrix proteins, including elastin. In this environment, SMC reexpressed desmin, a differentiation marker known to be lost under standard culture conditions. The endothelium expressed von Willebrand factor, incorporated acetylated LDL, produced PGI2, and strongly inhibited platelet adhesion in vitro. The complete vessel had a burst strength over 2000 mmHg. This is the first completely biological TEBV to display a burst strength comparable to that of human vessels. Short-term grafting experiment in a canine model demonstrated good handling and suturability characteristics. Taken together, these results suggest that this novel technique can produce completely biological vessels fulfilling the fundamental requirements for grafting: high burst strength, positive surgical handling, and a functional endothelium.

For patients with extensive burns, wound coverage with an autologous in vitro reconstructed skin made of both dermis and epidermis should be the best alternative to split-thickness graft. Unfortunately, various obstacles have delayed the widespread use of composite skin substitutes. Insufficient vascularization has been proposed as the most likely reason for their unreliable survival. Our purpose was to develop a vascular-like network inside tissue-engineered skin in order to improve graft vascularization. To reach this aim, we fabricated a collagen biopolymer in which three human cell types keratinocytes, dermal fibroblasts, and umbilical vein endothelial cells were cocultured. We demonstrated that the endothelialized skin equivalent (ESE) promoted spontaneous formation of capillary-like structures in a highly differentiated extracellular matrix. Immunohistochemical analysis and transmission electron microscopy of the ESE showed characteristics associated with the microvasculature in vivo (von Willebrand factor, Weibel-Palade bodies, basement membrane material, and intercellular junctions). We have developed the first endothelialized human tissue-engineered skin in which a network of capillary-like tubes is formed. The transplantation of this ESE on human should accelerate graft revascularization by inosculation of its preexisting capillary-like network with the patient's own blood vessels, as it is observed with autografts. In addition, the ESE turns out to be a promising in vitro angiogenesis model.

The major limitation for the application of an autologous in vitro tissue-engineered reconstructed skin (RS) for the treatment of burnt patients is the delayed vascularization of its relatively thick dermal avascular component, which may lead to graft necrosis. We have developed a human endothelialized reconstructed skin (ERS), combining keratinocytes, fibroblasts and endothelial cells (EC) in a collagen sponge. This skin substitute then spontaneously forms a network of capillary-like structures (CLS) in vitro. After transplantation to nude mice, we demonstrated that CLS containing mouse blood were observed underneath the epidermis in the ERS in less than 4 days, a delay comparable to our human skin control. In comparison, a 14-day period was necessary to achieve a similar result with the non-endothelialized RS. Furthermore, no mouse blood vessels were ever observed close to the epidermis before 14 days in the ERS and the RS. We thus concluded that the early vascularization observed in the ERS was most probably the result of inosculation of the CLS network with the host's capillaries, rather than neovascularization, which is a slower process. These results open exciting possibilities for the clinical application of many other tissue-engineered organs requiring a rapid vascularization.