**5.3. Importance of blood vessels for tissue homoeostasis**

remodeling, and after an additional 2 weeks of culture, the reseeding process results in the generation of a stroma with the same thickness as the one obtained by classical stacking of three sheets produced without reseeding. This dense stroma supported the development and maturation of the epithelium [88]. The reseeding technique offers a remarkable alternative to the classical self-assembly protocol because it is faster and it reduces costs associated with

**Figure 3. Graphical illustration of the improvements to the self-assembly technique (SS) when reseeding of cells (RS) is applied**. Reseeding allows the reduction of costs, handling time, and incubator space requirements throughout

Over the years, it has been shown that the origin of mesenchymal cells has a direct impact on the quality of bioengineered tissues. Carrier *et al*. [109] showed that reconstructed human cornea substitutes had great macroscopic and histological differences; especially in the corneal epithelium thickness and differentiation whether dermal fibroblasts or keratocytes (corneal fibroblasts) were used to produce the ECM. Constructs made with dermal fibroblasts were less transparent and lacked ultraviolet absorption characteristics compared to corneal tissues which were produced using autologous keratocytes [109]. The importance of this cross talk was further evaluated by mass spectrometry analyses performed on human stromal and epithelial layers of corneal substitute. Not only is the origin of mesenchymal cell important, but a fully differentiated and stratified epithelium is required for appropriate ECM synthe‐ sis and organization. This reciprocal regulation between the ECM and epithelial components is initiated as epithelial cells adhere to the stroma, they emit a continuous signal that medi‐

the steps of cell culture.

**5.2. Organ-specific stroma**

ates ECM remodeling accordingly [63].

extensive culture medium consumption as well as material [88] (**Figure 3**).

352 Composition and Function of the Extracellular Matrix in the Human Body

Blood vessel formation can occur through two distinct mechanisms: angiogenesis and vasculogenesis. Vasculogenesis involves the recruitment of progenitors of ECs from the bone marrow, which leads to the formation of a vascular plexus *de novo*, whereas angiogenesis occurs when pre-existing vessels form new branches or sprouts [112, 113]. Finally, inoscula‐ tion of a tissue is the anastomosis of the pre-established capillary-like network within the graft or engineered tissue with the host's vasculature [114].

ECs are in a stable quiescent state, however, they can become activated upon angiogenic stimuli in engineered tissues [115]. For instance, in response to conditions, such as tissue ischemia or chronic hypoxia, new collateral vessels can grow. Endothelium proliferation is stimulated by growth factors such as vascular endothelial growth factor (VEGF) that induce sprouting of new blood vessels. Additionally, proangiogenic signals increase MMP activity that prompts ECs to break apart their basement membrane allowing sprouting [116].

### *5.3.1. Vascularization strategies in tissue engineering*

ECs are a promising angiogenic cell source for therapeutic vasculogenesis because they have the potential to proliferate and rearrange themselves into functional capillary-like networks. Human umbilical vein endothelial cells (HUVEC) are an important source of ECs widely used in vasculogenesis [117]. Advantages of this cell type use are the noninvasive cell source, the profusion of medical wastes, and the impressive source of ECs in umbilical cords and placental tissues. Although the therapeutic use of HUVECs is limited because of their allogeneic nature, they remain a valuable EC source for basic and applied research needs [117].

To overcome this issue, the potential of ASCs for their differentiation into ECs was explored. Isolated adipose stromal vascular fraction (SVF) from white human adipose tissue is rich in adult stem cell populations, including EC progenitors. Freshly harvested SVF containing mixed white adipose stromal cells and white adipose ECs was cultured in 3D collagen hydrogels. Within the first week, the culture showed a formation of capillary network with continuous lumen [118], and after 3 weeks it gave rise to a functional 3D vascularized skin substitute that responded well to implantation in mice. This experiment demonstrates the synergy of vascular and stromal cells in blood network formation de novo. Hence, white ASCs Demonstrate promising results with minimal cell handling. Finally, human microvascular endothelial cells (HMVEC), which originate from small superficial capillaries, also represent a promising avenue fortissue endothelialization. These cells have been incorporated in in vitro

models using the self-assembly method and formed vascular networks with lumen [115]. HMVECs are an easily accessible source because they can be derived from a skin biopsy or any other tissue. The use of HMVECs could be particularly suitable for therapeutic applica‐ tion because it is best adapted for organ specific reconstructs.

#### *5.3.1.1. Vascularization of self-assembled tissues*

The co-culture of dermal fibroblasts and keratinocytes with HUVEC on a chitosan/collagen sponge showed the establishment of a capillary-like network similar to the microvasculature found in vivo [119]. Prevascularization of tissues prior to implantation has yield impressive improvements in regenerative medicine. In 2005, human endothelialized reconstructed skin models revealed an important reduction in the delay of functional vascularization after implantation in mice. Early signs of vascularization were observed in the endothelialized human skin grafts within 4 days following tissue implantation, as opposed to 14 days in the nonendothelialized reconstructed skin. Mouse blood vessels were only detected after 14 days in both models demonstrating that neovascularization is a latter process. The uniform distribution of ECs across the reconstruct ensures adequate perfusion of the entire graft. The colocalization of human and host mouse ECs inside a human capillary within the graft suggests the formation of chimeric microvessels and confirms inosculation between both microvascu‐ lar networks [114] (later confirmed in Gibot *et al*. [120] generated self-assembled tissue).

The progress of endothelialized tissue–engineered dermal substitutes lead to the introduc‐ tion of a new in vitro model of capillary-like network formation in self-assembled skin substitutes without the use of an exogenous scaffold. In this approach, stromal sheets, formed by culturing dermal fibroblast during 4 weeks, were seeded with ECs. To generate the 3D skin, two endothelialized stromal sheets were stacked and allowed to fuse [121] (**Figure 1B**). Although a capillary network was observed, the fact that ECs were seeded in a single plane orientation, on top of the stromal sheets, resulted in a vascularized skin model with mainly a 2D vascular network ratherthan a 3D network. In orderto provide the reconstructed skin with the optimal 3D capillary network, ECs were co-seeded with fibroblasts (**Figure 1D**). Incorpo‐ ration of ECs in the reconstructed model using the reseeding technique produced a capillarylike network with increased tissue elasticity and mechanical strength [88]. Moreover, because fibroblasts were seeded at high density, ECM was readily generated and allowed the dermal stroma to be rapidly embedded with ECs [88].This vascularized stroma had pericyte-like cells that expressed the neuron-glial 2 (NG2) marker, which characterizes the surrounding of capillary-like structures.

#### **6. Conclusions**

The self-assembly approach is used to generate several tissues for fundamental and clinical research applications. Over the years, adjustments to the stroma elaboration protocols and especially the ECM generation were proposed to improve the quality of the bioengineered substitutes. As one of the main objectives is to reduce the production time and costs, mechan‐ ical, biological, and chemical modifications were also introduced. Organ-specific ECM was associated with a better epithelial differentiation and an overall tissue architecture that closely mimics native tissues. To improve clinical applications, endothelialized tissues were generat‐ ed and grafted with better survival and functions compared to nonvascularized substitutes.
