**1. Introduction**

The skin is the largest organ of the body, accounting for about 15% of the total adult body weight. It is made up of three layers: the epidermis, dermis, and the hypodermis (Figure 1, [1]). Bioengineered skin substitutes, in the form of either cellularized engineered skin grafts or acellular dermal regeneration templates (DRT), have been developed to address two main issues still affecting the repair of extended deep wounds [2–10]: firstly, limiting the amount of healthy skin removed from the patient needed for closure; and secondly, acting as promoter for the restoration of the physiologic conditions of the skin avoiding the formation of severe scars. Skin acts not only as a barrier between the organism and the environment preventing invasion of pathogens and fending off chemical and physical assaults [11], it also plays a crucial role in the regulation of body temperature, moisture, and trafficking of water and solutes [12]. In addition, the sensory system of the skin allows the sensing of pain, temperature, light touch, discriminative touch, vibration and pressure. Finally, other important adnexal structures, such as sweat glands and hair follicles, contribute to the functionality of the healthy skin. In addition to the epidermis, severe damage due to burns, chronic ulcers and reconstructive surgeries, induce the destruction of the dermis. Unlike the epidermis, the dermis is characterized by an impaired healing process in which the final assembly of the extracellular matrix (ECM) is far from physiologic conditions. This mismatch compromises the reestablishment of the aforementioned regulatory functions of the whole organ [13–15]. The components of the ECM of the dermis (collagen elastin, hyaluronic acid, fibronectin, perlacan, water and other molecules), possess specific three-dimensional arrangements of sequences orchestrating the cross-talk among the different cell populations comprising the skin. Ultimately, such cross-talk affects the attachment, migration, differentiation and morphogenetic phenomena. For instance, the ECM promotes 'appropriate' communications between keratinocytes and the fibroblasts, and it is responsible for the formation and maintenance of the adnexal structures such as hair follicles, sweat glands and innervations [16–18]. Furthermore, when such adnexal structures become compromised, the self-regeneration of the epidermis cannot occur, and the wound becomes hard to heal. The repair of a deep wound can be divided into four subsequent phases [19,20]: (i) the coagulation and homeostasis phase (immediately after injury); (ii) the inflammatory phase (shortly after injury to tissue), during which swelling takes place; (iii) the proliferation period, where new tissues and blood vessels are formed; and (iv) the maturation phase, in which remodeling of new tissues takes place. How the maturation phase takes place determines the difference between repair and regeneration. The former is a "mere" closure process where fibroblasts bridge the wound gap by organizing their ECM differently from the healthy status. The latter restores the organization of the ECM that will appear indistinguishable from the healthy status [21]. The impaired ECM organization featuring the repair process depresses the regulatory and repository role of the extracellular space that ultimately forms an extended scar characterized by the loss of biological functionalities, inducing the insurgence of severe aesthetics and mobility-associated concerns. A classical approach to skin grafting and repairing is depicted in Figure 2. After debridement of the wound bed, a DRT is applied. Fibroblasts and endothelial cells from the recipient take at least one month to invade the DRT. After this time, it is possible to apply a split thickness skin graft (STSG): an epidermis with a layer of dermis removed from healthy sites of the patient. Both vascularization and fibroblast-secreted ECM molecules affect the take of the STSG that serves to trigger the regeneration of the epidermis due to the lack of adnexa and basal lamina [22–26]. As shown in Figure 2, with a period of two years, the remodeling of the neodermis occurs. To date, even though progress in biomaterials science and tissue engineering has led to the realization of different classes of skin substitutes (either cellularized or not), their healing potential is still limited in triggering a repair process instead of regeneration. In addition to economics, safety and regulatory (in case of allogenic or xenogeneic materials) concerns, in this work the currently available skin substitutes will be reviewed in the light of the composition of the dermis compartment and how this can affect the regeneration process. DRT can be fabricated starting from connective tissues of either allogenic or xenogeneic origin after removal of the cellular component [27,28]. The decellularization processes remove the associated risk of transmission of pathogens, preserving the composition of the ECM. On the other hand, the functionality of molecules of the ECM resulting from the decellularization processes compromise the correct signal presentation to the cells [16–18]. Pre-cellularization with endothelial cells, fibroblasts and keratinocytes seems to improve the biological performance of reconstructed three-dimensional matrices of both natural or synthetic origins [2,27,29–31], by speeding up the vascularization, the synthesis of neodermis, the take of the STSG and the closure of the wounds [32–34]. Nevertheless, the reconstructed three-dimensional matrices used to accommodate living cells prior to the implant are composed of exogenous biomatrices that possess composition, stiffness and three-dimensional arrangements that are quite different from the native dermis. For this reason, some doubts on their effectiveness in triggering a regeneration process have been raised [17]. Finally, a tissue engineering strategy that use patients' own cells to build up in vitro a human-like vascularized ECM featured by the absence of any exogenous material, is presented as an alternative to guide the wound toward a physiological regeneration process [35–37].

#### **2. Tissue Engineering Strategies for Skin Regeneration**

Tissue engineering aims at developing strategies to allow tissue and organ regeneration [32] by two approaches: (i) In in vitro tissue engineering the patient's human skin is re-built in a laboratory using either endogenous or allogenic cell lines (keratinocytes and fibroblasts); after a period of cultivation in three-dimensional matrices [24] and bioreactors [10,38], the engineered skin is then implanted [32]; (ii) in in vivo tissue engineering a three-dimensional matrix is introduced in the wound bed; such matrices are bio-functionalized in order to attract both cells and growth factors supporting skin regeneration [38]. The use of de novo fabricated skin becomes necessary when skin self-regeneration is hindered by adverse conditions [10]; in particular, when severe burns (second-, third-, and fourth-degree burns), chronic ulcers, surgery or trauma lead to the destruction of the dermis and underlying tissues (fat, muscle or bone) are exposed [39,40]. Bioengineered skin substitutes can be classified according to the following categories.


**Figure 1.** Main components of the human skin. (Image source: brgfx/Freepik).

**Figure 2.** The main steps of a two-step procedure to treat deep and partial wounds with the application of a DRT. (**A**) Healthy skin and wound bed after debridement. (**B**) Application of a DRT possessing an artificial silicone epidermis and covered with gauze. (**C**) Removal of the silicone epidermis and application of the STSG. (**D**) Long-term appearance of the repaired dermis. (**E**) Cellular end extracellular dynamics occurring during the wound healing process after the application of a DRT. W = week; M = month. DRT, dermal regeneration templates; STSG, split thickness skin graft; ECM, extracellular matrix.

#### *2.1. Dermal Regeneration Templates (DRT): Materials and Fabrication Techniques*

Porous and fibrous materials. Regardless of the tissue engineering strategy used (in vitro vs. in vivo approaches), a 3D scaffold supporting cell growth is required [42]. Scaffolds are biomaterials acting as temporary porous structures mimicking the 3D architecture of human tissue. In the case of skin tissue engineering, the tissue that one would like to mimic is the dermis. The scaffolds mimicking the dermis can be made by either natural or synthetic polymers (or a combination of both) and, regardless of their origin, they must have different characteristics: non-immunogenic; biocompatible; be able to resist the activity of proteolytic enzymes; stiff and flexible in order to withstand surgical procedures; be able to control the wound contracture; and to possess a degradation rate synchronous with the neo-dermis ingrowth and assembling. Furthermore, the scaffold should support epidermis attachment, maintenance and stratification, and it should promote the blood vessels' influx when implanted [23,34,42]. Three-dimensional porous structures can be obtained by different fabrication techniques allowing the production of different 3D architectures. Starting from melt polymers or polymer solutions, the use of porogen agents, or phase separation techniques, allows the production of sponge-like structures with interconnected pores and porosity ranging from 50 to 500 μm [44]. When porogens are used, the final mechanical properties of the scaffold can be modulated by varying the polymer concentration, the volume fraction and the dimension of the porogens. Phase separation technique exploits the thermodynamic instability of either polymeric solutions or blends [42]. The instability can be induced physically (i.e., temperature) or chemically (i.e., introduction of a solvent/non-solvent agent). The thermodynamic instability induces a segregation process with the formation of a dispersed and a continuum phase. The dispersed phase forms globular structures and after its removal porosity is created. The parameters affecting the final properties of the scaffold (mechanical properties, porosity, pore diameter and pore interconnection) are the initial composition of the polymer solution and, in the case of thermally induced phase separation, the cooling/heating rate is used to induce the thermodynamic instability. Woven and non-woven assembly of nano-fibers using electrospinning allows the production of porous structures categorized as fibrillar scaffolds [30,45,46]. Nanofibrous materials can be also produced by means of different techniques such as self-assembly, phase separation, fiber bonding and electrospinning. This kind of structure is able to mimic better the fibrous nature of the natural extracellular matrix. In in vitro tissue engineering, such preformed structures (either porous or fibrous) are colonized by the cells after the preparation, since their fabrication techniques represent severe conditions for cell viability. This could represent a limitation because a homogenous cell seeding through the thickness of the scaffolds is difficult to achieve and sophisticated bioreactors need to be used [38,47–49]. Hydrogels represent another class of fibrous scaffolds. These are highly hydrated 3D structures obtained by physical, ionic or covalent cross-linking of different polymers of both natural or synthetic origins [49–52]. Three-dimensional hydrogels can be obtained by the assembly and crosslinking of a liquid monomeric phase. This represents an advantage because the monomeric solution can be mixed with the cell suspension. At the end of the gelling process, cells remain entrapped in the 3D structures obtaining a homogenous cellularization of the final 3D scaffold. Although the final properties can be modulated by different parameters (monomer concentration, temperature, pH, UV radiation, ionic strength) the presence of cells poses the same constraints on the control of the final mechanical properties. For instance, UV radiation, pH, temperature and other methods used to induce cross-link of polymeric networks may affect cell viability. Hydrogel are, in general, considered soft materials. Hydrogels composed of gelatin–chitosan hydrogels [53], fibronectin–hyaluronan, or dextran-based hydrogels, in combination with nano-fibrous poly lactic-co-glycolic acid (PLGA) scaffolds have been extensively used for wound healing and regeneration applications [54–56]. In particular, dextran-based hydrogels demonstrated efficient vascularization. Composite hydrogels formed by glycosaminoglycan (GAG)–collagen showed good wound healing in rabbit models [56]. Self-assembling peptide-based hydrogel scaffolds have been reported to reduce burn wound healing time and skin cell proliferation [56].

Protein-based naturals biomaterials. Biomaterials of protein nature can be realized using (but not limited to) collagen, gelatin, silk and fibrinogen. Collagen is the most abundant structural protein of the human dermis secreted by fibroblasts, and is responsible for tensile stiffness. Collagen in skin tissue engineering is used as both acellular scaffold or cell-populated scaffold [56]. Collagen for tissue engineering application is extracted from animals: bovine, ovine and avian are the mostly exploited sources. Examples of acellular/porous scaffolds are the commercially available dermis substitutes Alloderm® or Integra®, while examples of cell-populated collagen hydrogels are represented by Apligraft® and Transcyte®. In addition, collagen-based biomaterials have been processed in the form of membranes, sponges, composite sponges and electrospun biomaterials with nanometric features [56]. Gelatin is a protein obtained by collagen denaturation possessing higher advantages in terms of cell adhesion and inducing a reduced immunogenic response. In the treatment of wounds and burns, gelatin has been used as electrospun nanofibers [57], membranes [58] and gelatin sponges loaded with growth factors [59]. Silk fibroin in the form of sponges, nanofibers and porous films have shown decreased inflammatory response, and promising results in wound healing and skin regeneration has been reported [60].

Polysaccharide-based biomaterials. Polysaccharide-based biomaterials are mainly used in the form of hydrogels. Those mostly used in skin regeneration and wound healing are dextran [54], cellulose [26], chitosan [61], alginate [62] and hyaluronic acid (HA) [63–65]. Among these, HA has been extensively used in skin regeneration leading to the commercialization of different skin substitutes, such as Hyaff®, Laserskin® and Hyalograft®.

Synthetic and composite biomaterials. Synthetic biomaterials comprise [30] the class of aliphatic polyesters, such as polylactic acid (PLA), polyglycolic (PGA) and polycaprolactone (PCL). They possess controllable mechanical stiffness and high process flexibility and are biocompatible and nontoxic. Moreover, PLA, PGA, PCL, and their blends and copolymers are FDA approved. An example of a commercially available skin substitute made using such materials is Dermagraft®.

Decellularized matrices. To date, no biomaterials yet exist that are able to mimic the composition of the native extracellular matrix as a whole [16–18]. Decellularized matrices of allogenic or xenogeneic origin should bridge such a gap. This is very important in skin regeneration because the lack of a functional extracellular matrix is the main cause affecting the impaired wound healing process. Currently available decellularized biomaterials comprise decellularized mesothelium, intestine, amniotic membrane, dermis and skin flaps [9]. Allogenic dermis can be obtained by treating fresh dermis from a cadaver with Dispase–Triton X-100 or NaCl–sodium dodecyl sulfate (SDS). In this way, collagen bundles and basement membranes retain their structure. The abstained biomaterial is further lyophilized. An FDA-approved decellularized dermis obtained with such a technique is Alloderm® [9]. Its use in combination with split-thickness autologous skin grafts allows complete cellularization of skin defects after 12 weeks post application when applied to full-thickness or partial-thickness burn wounds, thereby reducing subsequent scarring [65]. Other acellular dermal matrices are of porcine origin (e.g., Permacol®) and the decellularization process is similar to that of Alloderm®. Different techniques use a foaming process in order to destroy the cellular component, as well as any immunogenic agent. On the other hand, foaming compromises extracellular matrix structure and functions. In general, this kind of decellularized dermis is similar to porous scaffolds and is used for hemostatic applications. Other kinds of decellularized matrices are derived from the mesothelium, including peritoneum, pleura and pericardium [27,66,67]. Decellularized peritoneum is obtained using detergent agents and the processes are designed to maximize the preservation of the extracellular matrix architecture and composition. Because growth factors have a limited shelf life, such biomaterials are often combined with fibroblast growth factor (FGF) and epidermal growth factor (EGF) to promote wound repair. Decellularized mesothelium of porcine origin is used, for instance, in breast reconstruction. Bovine sources are another font of decellularized mesothelium showing a faster healing process than that observed for other decellularized dermis substitutes such as Alloderm® [27]. The intestine is another source to obtain decellularized extracellular matrices. Interestingly, many cytokines and growth factors (FGF and TGF-β families) are retained after the decellularization process. Different applications have been reported comprising the reconstruction of cornea, urethra, vagina, and lung. In the case of dermis reconstruction, decellularized intestine and, in particular, the decellularized small intestine submucosa, is a very promising scaffold due to its capability in promoting angiogenesis. OaSIS® and SurgySIS® (Cook Surgical, Bloomington, IN, USA) are two decellularized matrices from small intestine submucosa [9,68]. The human amniotic membrane is rich in basement membrane and avascular stromal matrix. Decellularization can be obtained using ethylenediaminetetraacetic acid (EDTA) and aprotinin, SDS, DNAase and RNAase. Nevertheless, such a process induces the reduction of anti-inflammatory and anti-scarring components by reducing their superiority compared to other matrices. Clinical applications can be found in the reconstruction of the ocular surface, while different studies have been performed on skin reconstruction in nude mice models [27].
