Differential Capability of Clinically Employed Dermal Regeneration Scaffolds to Support Vascularization for Tissue Bioengineering

The loss of skin integrity has always represented a major challenge for clinicians dealing with dermal defects, such as ulcers (diabetic, vascular and chronic), postoncologic resections (i.e., radical vulvectomy) or dermatologic disorders. The introduction in recent decades of acellular dermal matrices (ADMs) supporting the repair and restoration of skin functionality represented a significant step toward achieving clean wound repair before performing skin grafts. Hard-to-heal ulcers generally depend on local ischemia and nonadequate vascularization. In this context, one possible innovative approach could be the prevascularization of matrices with vessel-forming cells (inosculation). This paper presents a comparative analysis of the most widely used dermal templates, i.e., Integra® Bilayer Matrix Wound Dressing, PELNAC®, PriMatrix® Dermal Repair Scaffold, Endoform® Natural Dermal Template, and Myriad Matrix®, testing their ability to be colonized by human adult dermal microvascular endothelial cells (ADMECs) and to induce and support angiogenesis in vitro and in vivo. By in vitro studies, we demonstrated that Integra® and PELNAC® possess superior pro-adhesive and pro-angiogenetic properties. Animal models allowed us to demonstrate the ability of preseeded ADMECs on Integra® to promote the engraftment, integration and vascularization of ADMs at the site of application.


Introduction
The loss of skin integrity has always represented a major challenge for clinicians, especially when dealing with fragile patients. Burns, ulcers (diabetic, vascular and chronic), postoncologic resections (i.e., radical vulvectomy with inguinofemoral lymphadenectomy followed by chemoradiotherapy) and dermatologic disorders (i.e., epidermolysis bullosa) are among the most common causes of skin integrity defects [1][2][3]; these conditions represent major disabilities for patients, and have the potential to be fatal in severe cases.
The treatment of skin lesions can be either conservative or surgical, depending on the size and type of wound, as well as on patient characteristics and comorbidities. Generally, when dealing with vulnerable patients, the decision to perform surgery rather than applying conservative treatment is not always straightforward, since patient overall health status has a preponderant influence on the healing process. In fact, several local Table 1. Summary of the characteristics of the dermal substitutes analyzed in the current study. Dermal substitute structures were designed using the Blender 3D software (Blender Foundation, Stichting Blender Foundation, Buiklotermeerplein, Amsterdam, the Netherlands).

Commercial Scaffolds Components
Integra ® Bilayer Matrix Wound Dressing The dermal layer is composed of bovine type I tendon collagen and chondroitin-6-sulfate; the epidermal layer is made of silicon.
PELNAC ® A bilayer scaffold made of porcine tendon atelocollagen and a silicon layer.

PriMatrix ® Dermal Repair Scaffold
Acellular dermal tissue matrix from fetal bovine dermis, particularly rich in type II collagen.
Endoform ® Natural Dermal Template Derived from ovine forestomach tissue that is minimally processed to separate tissue layers and decellularize the tissue extracellular matrix (ECM). It maintains components essential to tissue repair (e.g., collagen I, III, IV, fibronectin, laminin, elastin, hyaluronic acid, heparin sulfate, GAGs, growth factors and chemokines).

Myriad Matrix ®
A collagen matrix with an intact ECM. Derived from ovine forestomach, it retains the innate biological structure and function of the native ECM associated macromolecules including elastin, fibronectin, glycosaminoglycans and laminin.
Biomedicines 2021, 9, x FOR PEER REVIEW 3 of 19 This report presents a comparative analysis of the most widely used dermal templates, i.e., Integra ® Bilayer Matrix Wound Dressing, PELNAC ® , PriMatrix ® Dermal Repair Scaffold, Endoform ® Natural Dermal Template, and Myriad Matrix ® (Table 1), testing their ability to be colonized by human adult dermal microvascular ECs (ADMECs) and to induce and support angiogenesis, both in vitro and in vivo, using a mouse model of full-thickness skin wounds. Table 1. Summary of the characteristics of the dermal substitutes analyzed in the current study. Dermal substitute structures were designed using the Blender 3D software (Blender Foundation, Stichting Blender Foundation, Buiklotermeerplein, Amsterdam, the Netherlands).

Commercial Scaffolds Components
Integra ® Bilayer Matrix Wound Dressing The dermal layer is composed of bovine type I tendon collagen and chondroitin-6-sulfate; the epidermal layer is made of silicon.
PELNAC ® A bilayer scaffold made of porcine tendon atelocollagen and a silicon layer.

PriMatrix ® Dermal Repair Scaffold
Acellular dermal tissue matrix from fetal bovine dermis, particularly rich in type II collagen.

Endoform ® Natural Dermal Template
Derived from ovine forestomach tissue that is minimally processed to separate tissue layers and decellularize the tissue extracellular matrix (ECM). It maintains components essential to tissue repair (e.g., collagen I, III, IV, fibronectin, laminin, elastin, hyaluronic acid, heparin sulfate, GAGs, growth factors and chemokines).

Myriad Matrix ®
A collagen matrix with an intact ECM. Derived from ovine forestomach, it retains the innate biological structure and function of the native ECM associated macromolecules including elastin, fibronectin, glycosaminoglycans and laminin.

Commercial Scaffolds Components
Integra ® Bilayer Matrix Wound Dressing The dermal layer is composed of bovine type I tendon collagen and chondroitin-6-sulfate; the epidermal layer is made of silicon.
PELNAC ® A bilayer scaffold made of porcine tendon atelocollagen and a silicon layer.

PriMatrix ® Dermal Repair Scaffold
Acellular dermal tissue matrix from fetal bovine dermis, particularly rich in type II collagen.

Endoform ® Natural Dermal Template
Derived from ovine forestomach tissue that is minimally processed to separate tissue layers and decellularize the tissue extracellular matrix (ECM). It maintains components essential to tissue repair (e.g., collagen I, III, IV, fibronectin, laminin, elastin, hyaluronic acid, heparin sulfate, GAGs, growth factors and chemokines).

Myriad Matrix ®
A collagen matrix with an intact ECM. Derived from ovine forestomach, it retains the innate biological structure and function of the native ECM associated macromolecules including elastin, fibronectin, glycosaminoglycans and laminin.
This report presents a comparative analysis of the most widely used dermal templates, i.e., Integra ® Bilayer Matrix Wound Dressing, PELNAC ® , PriMatrix ® Dermal Repair Scaffold, Endoform ® Natural Dermal Template, and Myriad Matrix ® (Table 1), testing their ability to be colonized by human adult dermal microvascular ECs (ADMECs) and to induce and support angiogenesis, both in vitro and in vivo, using a mouse model of full-thickness skin wounds.

Commercial Scaffolds Components
Integra ® Bilayer Matrix Wound Dressing The dermal layer is composed of bovine type I tendon collagen and chondroitin-6-sulfate; the epidermal layer is made of silicon.
PELNAC ® A bilayer scaffold made of porcine tendon atelocollagen and a silicon layer.
PriMatrix ® Dermal Repair Scaffold Acellular dermal tissue matrix from fetal bovine dermis, particularly rich in type II collagen.

Endoform ® Natural Dermal Template
Derived from ovine forestomach tissue that is minimally processed to separate tissue layers and decellularize the tissue extracellular matrix (ECM). It maintains components essential to tissue repair (e.g., collagen I, III, IV, fibronectin, laminin, elastin, hyaluronic acid, heparin sulfate, GAGs, growth factors and chemokines).

Myriad Matrix ®
A collagen matrix with an intact ECM. Derived from ovine forestomach, it retains the innate biological structure and function of the native ECM associated macromolecules including elastin, fibronectin, glycosaminoglycans and laminin.
This report presents a comparative analysis of the most widely used dermal templates, i.e., Integra ® Bilayer Matrix Wound Dressing, PELNAC ® , PriMatrix ® Dermal Repair Scaffold, Endoform ® Natural Dermal Template, and Myriad Matrix ® (Table 1), testing their ability to be colonized by human adult dermal microvascular ECs (ADMECs) and to induce and support angiogenesis, both in vitro and in vivo, using a mouse model of full-thickness skin wounds.

Commercial Scaffolds Components
Integra ® Bilayer Matrix Wound Dressing The dermal layer is composed of bovine type I tendon collagen and chondroitin-6-sulfate; the epidermal layer is made of silicon.
PELNAC ® A bilayer scaffold made of porcine tendon atelocollagen and a silicon layer.
PriMatrix ® Dermal Repair Scaffold Acellular dermal tissue matrix from fetal bovine dermis, particularly rich in type II collagen.

Endoform ® Natural Dermal Template
Derived from ovine forestomach tissue that is minimally processed to separate tissue layers and decellularize the tissue extracellular matrix (ECM). It maintains components essential to tissue repair (e.g., collagen I, III, IV, fibronectin, laminin, elastin, hyaluronic acid, heparin sulfate, GAGs, growth factors and chemokines).

Myriad Matrix ®
A collagen matrix with an intact ECM. Derived from ovine forestomach, it retains the innate biological structure and function of the native ECM associated macromolecules including elastin, fibronectin, glycosaminoglycans and laminin.

Cell Isolation and Culture
Human umbilical vein ECs (HUVECs) were isolated following the protocol described by Jeffe et al. [26]. Pregnant women were enrolled at the Institute for Maternal and Child Health, IRCCS Burlo Garofolo, Trieste, Italy. The study was reviewed and approved by the Regional Ethical Committee of FVG (CEUR), Udine, Italy (Prot. 0010144/P/GEN/ARCS 2019). Informed consent for participation in the study was obtained from all women. Human ADMECs were harvested from skin biopsies of patients undergoing reductive plastic surgery, following the protocol described by Kraling et al. [27]. Patients were enrolled at the Cattinara Hospital of Trieste (Plastic Surgery Unit); informed consent for participation in the study was obtained from all women. Cells were seeded onto a fibronectin-gelatin coated flask and maintained in culture with human endothelial serum free medium (HESFM, Gibco, Carlsbad, CA, USA) supplemented with 20 ng/mL basic fibroblast growth factor, 10 ng/mL epidermal growth factor (Immunological Sciences), 1% penicillin-streptomycin (Sigma Aldrich), 10% v/v fetal bovine serum (FBS, Life Technologies), 10% v/v human serum (Sigma Aldrich) and 1:100 hydrocortisone (Sigma Aldrich). ECs were positively selected with Dynabeads ® CD31-conjugated magnetic beads (Invitrogen, Thermo Fisher Scientific-11155D). ADMECs were maintained at 37 • C in a humidified atmosphere in a 5% v/v CO 2 incubator.

Immunofluorescence
Cells were seeded onto round glass coverslips of 10 mm diameter previously coated with fibronectin-gelatin (2 µg/cm 2 ). Confluent cells were fixed with 3% paraformaldehyde (PFA) for 15 min in the dark. In order to perform quenching, blocking and permeabilization, cells were then incubated with a solution of 1% bovine serum albumin (BSA), 0.1 % Triton X-100 and 50 mM glycine in Dulbecco's Phosphate Saline Buffer (dPBS, Sigma Aldrich) for 30 min at room temperature (RT). Primary antibodies diluted in dPBS-2% BSA and 0.7 mM CaCl 2 and 0.7 mM MgCl 2 + were added for 1 h at RT; secondary antibodies were incubated for 30 min at RT. Nuclei were stained with DAPI. Glass slides were mounted with a fluorescence mounting medium (Dako), and images were acquired with a Leica DM3000 microscope (Leica, Wetzlar, Germany) using a Leica DFC320 digital camera (Leica).

RNA Isolation, cDNA Synthesis and Real-Time Quantitative PCR
ADMECs were seeded onto different matrices previously cut into small pieces with a biopsy punch under sterile conditions. Cells were cultured for 36 h at 37 • C in a humidified atmosphere in a 5% v/v CO 2 incubator. RNA was extracted from cells with a Total RNA purification kit (Norgen Biotek Corp., Thorold, ON, Canada) following the manufacturer's protocol, and retrotranscribed to cDNA with the SuperMix kit (Bioline, Meridian Life Science, Memphis, TN, USA) as previously described [28]. The expression level of the genes of interest was evaluated by comparative quantification based on the reaction efficacy and normalized to the expression of 18S, GAPDH, and TBP as housekeeping genes [29]. The expression level of the following genes was investigated: VEGFA, PGF, ANGPT1, KDR, FLT1, TEK, IL6, IL8/CXCL8, TNF, MCP1/CCL2, MMP2, and MMP9. Primer sequences are reported in Table 2: The reaction was performed using the Rotor-Gene 6000 (Corbett Research, Mortlake, Australia), using SYBR™ Green PCR Master Mix (Applied Biosystems, Milan, Italy), following a program of 45 cycles of denaturation (60 s at 95 • C), annealing (30 s at 60 • C, the melting temperature of the primers) and amplification (60 s at 72 • C).

MTT Assay
Matrices were cut into small pieces (8 mm in diameter) using a biopsy punch under sterile conditions. ADMECs were seeded onto the matrices at a concentration of 2.5 × 10 5 cells/matrix. After 5, 15 or 30 min of adhesion at 37 • C in a 5% CO 2 v/v incubator, nonadherent cells were washed away with dPBS supplemented with 0.7 mM CaCl 2 and 0.7 mM MgCl 2 . To analyze cell viability, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) was added and incubated at 37 • C for 2-4 h. The purple formazan crystals were solubilized by adding dimethyl sulfoxide (DMSO) under shaking for 20 min. Absorbance was read at 570 nm using a plate reading spectrophotometer.

Time Course Adhesion Assay
Matrices were cut into small pieces (8 mm in diameter) using a biopsy punch under sterile conditions. Cells were stained with 10 µg/mL of fluorescent dye FAST DiI (molecular probes, Invitrogen) diluted in dPBS. After 15 min of incubation at 37 • C in a 5% v/v CO 2 incubator, 2.5 × 10 5 cells were resuspended in HESFM supplemented with 10% FBS and 1:100 hydrocortisone and seeded onto the matrices. After 5, 15 or 30 min of adhesion at 37 • C in a 5% v/v CO 2 incubator, nonadherent cells were washed away with dPBS supplemented with 0.7 mM CaCl 2 and 0.7 mM MgCl 2 . Cells were lysed and the plate was immediately read with Infinite200 (TECAN Italia, Milan, Italy).

Proliferation Assay
ADMECs were cultured for 4 days in an incubator at 37 • C, 5% v/v CO 2 . After 4 days, matrices were labelled with 10 µg/mL of FAST DiI diluted in dPBS and lysing buffer was added. The fluorescence was read with Infinite200 (TECAN Italia, Milan, Italy).

Staining for Ki-67 and Vimentin
Matrixes were cut into small pieces (8 mm in diameter) with a biopsy punch under sterile conditions. ADMECs were seeded onto the matrices and cultured for 4 days in an incubator at 37 • C, 5% v/v CO 2 . Matrices were fixed with 3% PFA for 20 min at RT under shaking in the darkness. Matrices were incubated for 40 min at RT under shaking with primary antibodies anti-Ki-67 and antivimentin diluted in saponin A, composed of 60 mg Saponin (Farmitalia Carlo Erba, Milan, Italy) in 10 mL dPBS; and for 30 min at RT in darkness under shaking with LI-COR Biosciences-specific secondary antibodies. Images were acquired using Odissey CLx (LI-COR Biosciences, Lincoln, NE, USA) and analyzed with LI-COR Image Studio Acquisition software (LI-COR Biosciences, Lincoln, NE, USA).
To chemically mimic hypoxia and diabetes, cells were cultured for 4 days in the presence of 100 µM deferoxamine (Sigma Aldrich) or 20 µM advanced glycation endproduct-BSA (AGE-BSA), kindly donated by Dr. Cristina Bellarosa (Italian Liver Foundation). The concentration of deferoxamine was selected based on a previous work by Sánchez-Elsner et al. [30]. The concentration of AGE-BSA was determined by previous studies by Gallo et al. [31]. The seeding and staining protocol are described above.

Growth Factors and Chemokines Detection
ADMECs were seeded onto different matrices previously cut into small pieces with a biopsy punch under sterile conditions, or grown to confluence in 24-well plates (BD Falcon) in serum-free medium and stimulated for 24 h with interferon (IFN)-γ (100 U/mL), tumor necrosis factor (TNF)-α (100 ng/mL) or interleukin (IL)-1β (10 ng/mL) (all purchased from Peprotech, Milan, Italy). Cells seeded onto matrices were cultured for 36 h at 37 • C in a humidified atmosphere in a 5% v/v CO 2 incubator; the supernatant was then collected. The levels of placental growth factor (PlGF), VEGF-A, TNF-α, and IL-8 were determined with a commercial ELISA kit following the manufacturer's protocol (Human PlGF ELISA kit, Human ANGPT-1 ELISA kit and Human VEGF-A ELISA kit, Sigma Aldrich; Human TNFα ELISA KIT, Diaclone SAS; IL-8 Human ELISA Kit, Invitrogen Thermo Fisher Scientific).

Subcutaneously Application of ADMECs and In Vivo Experiments
First, 2.5 × 10 5 ECs were resuspended in HESFM and seeded for 30 min at 37 • C, 5% v/v CO 2 , Integra ® Dermal Regeneration Template without a silicone layer (Integra ® Life Science Corporation) or PELNAC ® Dermal Substitute (Eurosurgical), that had previously been cut into small pieces (6 mm × 5 mm) using a sterile scalpel. After cell seeding, the scaffolds were directly implanted on the quadriceps of immunodeficient NOD scid gamma (NSG) mice [14]. Surgical implantation of the scaffolds was performed on adult NSG (8 weeks old) mice under general anesthesia by ketamine/xylazine (100 mg/kg and 10 mg/kg body weight, i.p.). An incision of approximately 1 cm was made at the level of the knee towards the medial thigh on the left hind limb of the mice. The skin was separated from the quadricep muscle and the scaffolds were applied with the cells facing down. The incision was closed with 5-0 suture thread. Institutional guidelines in compliance with national and international laws and policies were followed for animal care and treatment. All experimental procedures were approved by the ICGEB Animal Welfare Board, with the requirements of the EU Directive 2010/63/EU, and by the Italian Ministry of Health.
The mice were sacrificed by cervical dislocation after anesthesia with 5% isoflurane. Integra ® , PELNAC ® and the surrounding tissue on the quadricep were collected, transversely cut in half and fixed overnight using 2% v/v PFA (paraformaldehyde) in dPBS 1X solution (Santa Cruz). The following day, PFA solution was removed and replaced with 20% sucrose cryoprotector solution (Sigma-Aldrich, St. Louis, MO, USA). The samples were cut into 8-µm sections using cryostat. The sections were thawed and dried at room temperature for 5 min. After a wash in 1X dPBS, samples were permeabilized using 0.5% v/v Triton X-100 in 1X dPBS for 15 min. The blocking step was performed using 5% w/v in 1X dPBS solution for 1 h. The sections were incubated overnight at 4 • C using a humid chamber with the primary antibody, diluted at a ratio of 1:200 v/v with 1% v/v BSA, 0.1% v/v Tween-20 in 1X dPBS solution. The next day, the slides were washed once in 0.02% v/v Tween 20 in 1X dPBS solution for 5 min and twice in 1X dPBS for 5 min each. The incubation with secondary antibody was performed for 2 h at RT in a humid chamber diluting the secondary antibody 1:500 v/v with 1% v/v BSA, 0,1% v/v Tween-20 in 1X dPBS solution. The tissues were washed once in 0.02% v/v Tween 20 in 1X dPBS solution for 5 min and twice in 1X dPBS for 5 min each. Nuclei were counterstained with Hoechst 33342 (#H3570 Invitrogen) diluted 1:5000 in 1X dPBS for 7 min. Subsequently, slides were washed three times with 1X dPBS and mounted using Mowiol mounting medium (Sigma-Aldrich).

Phenotypic Characterization of ADMECs and Evaluation of their Responsiveness to Pro-Inflammatory Cytokines
ADMECs were isolated from skin biopsies of adult patients undergoing reductive plastic surgery. We carried out an extensive characterization by immunofluorescence (IF) ( Figure 1A) in order to determine the purity of the isolated cells. ADMECs were positively stained for classical vascular endothelium markers such as PECAM/CD31, vWF, VEGFR3 and Vimentin, but were completely negative for Podoplanin, CK8/18 and CD45 (data not shown). We evaluated the behavior of ADMECs in response to classical inflammatory stimuli, as compared to HUVECs ( Figure 1B-F). As reported in Figure 1B, ADMECs manifested a low responsiveness to IFN-γ and TNF-α stimulation in terms of IL-8/CXCL8, MCP1/CCL2 and IL-6 secretion, but expressed higher levels of MIP1α/CCL3, RANTES/CCL5, TNF-α and IL-1β stimulation.

Evaluation of ADMEC Adhesive Properties and Colonization Capability of Different Dermal Substitutes
ADMECs were seeded onto five different dermal substitutes (Integra ® , PELNAC ® , PriMatrix ® , Endoform ® , or Myriad ® ). After 5, 15 or 30 min of incubation, the scaffolds were extensively washed to remove unattached cells, and adherent ADMECs were labelled with the viability marker MTT in a 24-well plate (Figure 2A). Successively, MTT crystals were solubilized with DMSO and absorbance was read at 570 nm. As shown in Figure  2B,C, Integra ® and PELNAC ® demonstrated superior pro-adhesive properties compared to PriMatrix ® , Endoform ® and Myriad ® . We observed that the cells were able to colonize the entire thickness/depth of these scaffolds (Videos S1 and S2), as demonstrated by IF analysis with antihuman CD31. The higher pro-adhesive properties of Integra ® and PELNAC ® were further confirmed by labelling ADMECs with FAST DiI, a viable fluorescent dye, prior to seeding. After 5 or 15 min of adhesion, labelled cells were lysed and the fluorescence was quantified. In this case, Myriad ® showed the greatest ability to stimulate cell adhesion, but also a high variability among the different ADMEC populations. The results confirmed data obtained by the MTT technique, highlighting the slightly better behavior of Integra ® in terms of EC colonization capability.

Assessment of the Proliferative Drive Induced by the Different Dermal Substitutes
In order to evaluate the proliferative boost induced by the different dermal substitutes, ADMECs were grown onto the scaffolds for 36 h and then labelled with FAST DiI. As shown in Figure 3A, the presence of living ADMECs was significantly higher on Integra ® compared to the other scaffolds, except Myriad ® , which demonstrated highly variable behavior. To determine the percentage of proliferating cells present in the different dermal substitutes, the attached cells were stained for both vimentin in order to determine the total number of cells, and for Ki-67 to calculate the number of proliferating cells. Both stainings were quantified using a Biosciences Infrared Odyssey imaging system (LI-COR Biosciences, Lincoln, NE, USA), a fluorescence scanner which is able to detect multiple layers on the z-axis. As reported in Figure 3B, ADMECs grown on Integra ® and PELNAC ® exceeded a proliferation percentage of 40%, whereas those on PriMatrix ® , Endoform ® and Myriad ® showed a lower proliferation index.

Evaluation of ADMEC Adhesive Properties and Colonization Capability of Different Dermal Substitutes
ADMECs were seeded onto five different dermal substitutes (Integra ® , PELNAC ® , PriMatrix ® , Endoform ® , or Myriad ® ). After 5, 15 or 30 min of incubation, the scaffolds were extensively washed to remove unattached cells, and adherent ADMECs were labelled with the viability marker MTT in a 24-well plate (Figure 2A). Successively, MTT crystals were solubilized with DMSO and absorbance was read at 570 nm. As shown in Figure  2B,C, Integra ® and PELNAC ® demonstrated superior pro-adhesive properties compared to PriMatrix ® , Endoform ® and Myriad ® . We observed that the cells were able to colonize the entire thickness/depth of these scaffolds (Video S1,S2), as demonstrated by IF analysis with antihuman CD31. The higher pro-adhesive properties of Integra ® and PELNAC ® were further confirmed by labelling ADMECs with FAST DiI, a viable fluorescent dye, With the aim of mimicking a chronic wound microenvironment, we cultured ADMECs under hypoxic conditions (i.e., by incubating with deferoxamine) or diabetogenic conditions (i.e., by adding AGE-BSA to the culture media). The results shown in Figure 3C,D indicate that there were no significant differences among the proliferative rates of cells seeded onto the different scaffolds, except for Endoform ® , which manifested a statistically lower proliferative rate in both conditions. Even if not significantly, Integra ® and Myriad ® demonstrated the best performance under hypoxic and diabetic conditions. fluorescence was quantified. In this case, Myriad ® showed the greatest ability to stimulate cell adhesion, but also a high variability among the different ADMEC populations. The results confirmed data obtained by the MTT technique, highlighting the slightly better behavior of Integra ® in terms of EC colonization capability.   (E) A representative image of labelled scaffolds was acquired by using LI-COR Odyssey imaging system, and data were processed using Image Studio system 5.0 software (LI-COR Biosciences, Lincoln, NE, USA). The formula used for the calculation is reported in the Supplementary Material section. Data from four independent experiments conducted in triplicate are presented as mean ± SE. * p < 0.01 vs. Integra; n.s. = not significant.

Proangiogenic and Remodeling Factors Produced by ADMECs on Different Dermal Substitutes
Having demonstrated the poor pro-adhesive and proproliferative properties of PriMatrix ® , Endoform ® , and considering the great variability of the results obtained with Myriad ® , subsequent studies were conducted only on Integra ® and PELNAC ® . To assess the mRNA expression of angiogenic-promoting factors (VEGFA, PGF, ANGPT1), receptors (KDR, FLT1, TEK) and tissue remodeling factors (MMP2 and MMP9) ex-pressed by ADMECs cultured on Integra ® and PELNAC ® , we performed RT-qPCR after seeding the cells onto the scaffolds for 36 h.
The interaction of ADMECs with Integra ® induced the activation of almost all angiogenetic factor genes compared to 2D cell culture (except for TEK and MMP2). PELNAC ® was able to activate numerous angiogenic genes and, in some cases, in an even stronger manner than Integra ® (AGPT1, KDR, FLT1, MMP9) ( Figure 4A-H).

Dermal Substitutes Modulate the Pro-Inflammatory Behavior of ADMEC
Since the formation of granulation tissue and leukocyte recruitment is fundamental for the restitutio ad integrum of wounds [14], we wanted to investigate the ability of the different scaffolds to induce ADMEC expression of the chemokine and cytokine genes involved in the inflammatory process. The results shown in Figure 5A-D indicate that Moreover, the upregulation of soluble angiogenetic factors (VEGF-A, PlGF and ANGPT-1) was also confirmed in terms of protein secretion ( Figure 4I-K).

Dermal Substitutes Modulate the Pro-Inflammatory Behavior of ADMEC
Since the formation of granulation tissue and leukocyte recruitment is fundamental for the restitutio ad integrum of wounds [14], we wanted to investigate the ability of the different scaffolds to induce ADMEC expression of the chemokine and cytokine genes involved in the inflammatory process. The results shown in Figure 5A-D indicate that Integra ® was able to upregulate the expression of the healing-promoting cytokines as follows: a two-fold increase was observed in the expression of TNF, a 2.5-fold increase in the upregulation of IL8/CXCL8 was observed; a seven-fold increase of MCP1/CCL2 was observed; and a two-fold of IL6 was observed, as compared to cells seeded without scaffolds. On the other hand, PELNAC ® yielded only a two-fold upregulation of IL8. follows: a two-fold increase was observed in the expression of TNF, a 2.5-fold increase in the upregulation of IL8/CXCL8 was observed; a seven-fold increase of MCP1/CCL2 was observed; and a two-fold of IL6 was observed, as compared to cells seeded without scaffolds. On the other hand, PELNAC ® yielded only a two-fold upregulation of IL8. The cytokine upregulation observed for gene expression was not confirmed at the protein level. We confirmed a slight, nonsignificant upregulation of IL-8/CXCL8 ( Figure  S1A), but we detected a significant downregulation of TNF-α ( Figure S1B).

In Vivo Studies
To determine the capacity of ADMECs to support the vascularization of clinically employed dermal regeneration scaffolds in vivo, we used host immunodeficient NSG mice to assess the engraftment and persistence of these cells following subcutaneous implantation. ADMECs were seeded either on Integra ® or PELNAC ® for 30 min and subcutaneously implanted on the quadriceps of NSG recipient mice. Histological analysis at 10 days revealed the presence of CD31-stained human ECs that colonized Integra ® ( Figure  6A). In some cases, we observed the formation of human derived vascular structures inside the scaffold that were approaching mouse ECs, stained for mouse CD31. In addition, the host ECs colonized Integra ® , demonstrating the successful engraftment of the scaffold beside the application and its vascularization. In contrast, very few CD31 + cells, of either human or mouse origin, were present inside PELNAC ® , and they did not form any vascular structures ( Figure 6B). Overall, these data indicate that Integra ® is a more suitable scaffold for tissue engineering approaches, as it promotes a high level of EC engraftment, integration and vascularization. The cytokine upregulation observed for gene expression was not confirmed at the protein level. We confirmed a slight, nonsignificant upregulation of IL-8/CXCL8 ( Figure S1A), but we detected a significant downregulation of TNF-α ( Figure S1B).

In Vivo Studies
To determine the capacity of ADMECs to support the vascularization of clinically employed dermal regeneration scaffolds in vivo, we used host immunodeficient NSG mice to assess the engraftment and persistence of these cells following subcutaneous implantation. ADMECs were seeded either on Integra ® or PELNAC ® for 30 min and subcutaneously implanted on the quadriceps of NSG recipient mice. Histological analysis at 10 days revealed the presence of CD31-stained human ECs that colonized Integra ® ( Figure 6A). In some cases, we observed the formation of human derived vascular structures inside the scaffold that were approaching mouse ECs, stained for mouse CD31. In addition, the host ECs colonized Integra ® , demonstrating the successful engraftment of the scaffold beside the application and its vascularization. In contrast, very few CD31 + cells, of either human or mouse origin, were present inside PELNAC ® , and they did not form any vascular structures ( Figure 6B). Overall, these data indicate that Integra ® is a more suitable scaffold for tissue engineering approaches, as it promotes a high level of EC engraftment, integration and vascularization.
Biomedicines 2021, 9, x FOR PEER REVIEW 15 of 19 Figure 6. ADMECs engrafted and formed vascular structures on Integra ® but not on PELNAC ® . Representative images of the Integra ® (A) and PELNAC ® (B) scaffolds, populated by ADMECs at day 10 after subcutaneous cell implantation. Human and mouse ECs were stained using species-specific anti-CD31 antibodies. Both Integra ® and PELNAC ® are visible in green due to scaffold autofluorescence. Scale bar 100 μm.

Discussion
Tissue-engineered skin substitutes are essential for tissue repair and regeneration when dealing with wounds which are large in size and penetrate deep below the dermis [33]. Although skin was the first engineered organ that went from laboratory research to patient care, there are still several limitations on the use of dermal substitutes, such as reduced vascularization, poor mechanical integrity, failure to integrate, scarring and immune rejection [34]. In the present study, we report on the differential ability of clinically employed dermal regeneration scaffolds to support ADMECs for the revascularization process in a chronic wound microenvironment.
We selected five different dermal substitutes which are currently being used in clinics (see Table 1), i.e., Integra ® Bilayer Matrix Wound Dressing, PELNAC ® , PriMatrix ® Dermal Repair Scaffold, Endoform ® Natural Dermal Template, and Myriad Matrix ® , and compared their capacity to be colonized by adult dermal microvascular ECs, inducing angiogenesis. Integra ® Bilayer Matrix Wound Dressing is a double layer membrane consisting of a porous coprecipitate of type I bovine cross-linked tendon collagen, rich in shark chondroitin-6-sulfate and a removable silicon layer. PELNAC ® is a dermal regenerating matrix composed of atelocollagen derived from pig tendon and reinforced with a silicone layer. PriMatrix ® Dermal Repair Scaffold is derived from fetal bovine dermis and is particularly rich in type III collagen that is active in developing and healing tissues. Endoform ® Natural Dermal Template is a recently developed matrix based on AORA ECM ® technology. It is derived from decellularized ovine forestomach, harvested exclusively from New Zealand pasture-raised animals. This bioscaffold contains more than 150 ECM proteins which are important for the wound healing process, and maintains residual vascular channels to support the establishment of new vasculature. Myriad Matrix ® is an engineered ECM that contains the natural porous structure of AROA ECM ® with interstitial perforations that allow cell infiltration.
Microvascular ECs are active participants in, and regulators of, inflammatory processes, secreting inflammatory mediators, modulating adhesion and migration of leukocytes through the expression of adhesion molecules and chemokines [28]. As demonstrated by Agostinis et al. [28], ECs are not all the same, and they respond very differently Human and mouse ECs were stained using species-specific anti-CD31 antibodies. Both Integra ® and PELNAC ® are visible in green due to scaffold autofluorescence. Scale bar 100 µm.

Discussion
Tissue-engineered skin substitutes are essential for tissue repair and regeneration when dealing with wounds which are large in size and penetrate deep below the dermis [33]. Although skin was the first engineered organ that went from laboratory research to patient care, there are still several limitations on the use of dermal substitutes, such as reduced vascularization, poor mechanical integrity, failure to integrate, scarring and immune rejection [34]. In the present study, we report on the differential ability of clinically employed dermal regeneration scaffolds to support ADMECs for the revascularization process in a chronic wound microenvironment.
We selected five different dermal substitutes which are currently being used in clinics (see Table 1), i.e., Integra ® Bilayer Matrix Wound Dressing, PELNAC ® , PriMatrix ® Dermal Repair Scaffold, Endoform ® Natural Dermal Template, and Myriad Matrix ® , and compared their capacity to be colonized by adult dermal microvascular ECs, inducing angiogenesis. Integra ® Bilayer Matrix Wound Dressing is a double layer membrane consisting of a porous coprecipitate of type I bovine cross-linked tendon collagen, rich in shark chondroitin-6sulfate and a removable silicon layer. PELNAC ® is a dermal regenerating matrix composed of atelocollagen derived from pig tendon and reinforced with a silicone layer. PriMatrix ® Dermal Repair Scaffold is derived from fetal bovine dermis and is particularly rich in type III collagen that is active in developing and healing tissues. Endoform ® Natural Dermal Template is a recently developed matrix based on AORA ECM ® technology. It is derived from decellularized ovine forestomach, harvested exclusively from New Zealand pasture-raised animals. This bioscaffold contains more than 150 ECM proteins which are important for the wound healing process, and maintains residual vascular channels to support the establishment of new vasculature. Myriad Matrix ® is an engineered ECM that contains the natural porous structure of AROA ECM ® with interstitial perforations that allow cell infiltration.
Microvascular ECs are active participants in, and regulators of, inflammatory processes, secreting inflammatory mediators, modulating adhesion and migration of leukocytes through the expression of adhesion molecules and chemokines [28]. As demonstrated by Agostinis et al. [28], ECs are not all the same, and they respond very differently depending on the tissue of origin. In order to demonstrate the differential behavior of ECs of different provenance in response to inflammatory conditions, we compared ADMECs isolated from adult dermal biopsies with HUVECs isolated from human umbilical veins, which have been extensively characterized in terms of angiogenesis and inflammation. ADMECs were found to be less responsive to IFN-γ compared to HUVECs, although they expressed a high level of cytokines and chemokines in response to TNF-α and IL-1β. Their low responsiveness to pro-inflammatory cytokines, such as IFN-γ, resulted in a lower production of IL-8/CXCL8, MCP1/CCL2 and IL-6. In contrast, stimulation with TNF-α and IL-1β increased the production of MIP1α/CCL3 and RANTES/CCL5, two chemokines involved in lymphocyte and neutrophil recruitment and secreted at very high levels by ADMECs. Based on these differences, the use of ADMECs rather than HUVECs appeared to be more promising for our in vitro model. Moreover, we decided to use ADMECs for an easier translational feasibility, since the future application of autologous ADMECs, isolated from patient punch biopsies, would be desirable to improve wound healing in clinical practice.
The principal aim of this study was to enhance revascularization to successfully promote wound healing [35]. Since the first step of matrix colonization is cell adhesion, we performed adhesion assays to assess the different adhesive properties of the five scaffolds. All of the analyzed matrices, with the exception of the Endoform ® , reached a plateau value after just 15 min. Integra ® and PELNAC ® showed excellent pro-adhesive properties after 5 min, making them the most promising scaffolds. Myriad ® also showed some interesting adhesive features despite its variability among replicates and cell populations. Another property under evaluation was the abilities of the scaffolds to induce EC proliferation. To this end, we used two different approaches: (i) quantification of the total number of living cells present in the matrix after 36 h, and (ii) the percentage of proliferating cells (positive for Ki-67) normalized for the total number of cells (positive for vimentin). The results confirmed those obtained for the adhesion experiments, indicating that Integra ® , PELNAC ® , and Myriad ® were the most promising dermal substitutes in terms of local proliferation of vascular cells.
Chronic wounds are defined as a discontinuity in the skin barrier lasting longer than 42 days. Risk factors for the delay of physiological wound healing are chronic disease, diabetes, age, vascular insufficiency, nutritional deficit and some local factors [36]. With the aim of mimicking a chronic wound microenvironment, we cultured ADMECs in the scaffolds under hypoxic conditions by using deferoxamine. Furthermore, we wanted to reproduce diabetogenic conditions by adding AGE-BSA to the culture media, as advanced glycation end products and their receptors (RAGE) have been shown to be involved in the pathogenesis of diabetes [37]. The results suggested that only cells seeded onto Myriad ® were able to proliferate under the hypoxia condition. In general, hypoxia signaling is essential during tissue and organ development, but this condition is also important in specific organs in healthy adults. In fact, since the hypoxia-inducible factor(HIF)-α is not hydroxylated by prolyl hydroxylase under hypoxic conditions, it can activate the HIFmediated transcriptional program, which also includes angiogenesis. Hypoxia triggers the transcription of angiogenic genes including VEGFA, PGF, PDGFB and ANGPT1 and 2. Furthermore, it posttranscriptionally controls pro-angiogenic chemokines and receptors, inducing the migration of progenitors to the site of angiogenesis. Hypoxia can also promote EC proliferation and sprouting in addition to matrix remodeling [38]. The physiological wound healing process requires both angiogenesis, which means the sprouting of newly formed capillaries from pre-existing blood vessels, and vasculogenesis, which implies the formation of new vessels by recruited progenitor cells. Wound healing failure is linked with aging and diabetes, as well as their synergistic negative effects. Several studies have demonstrated that HIF-α expression is decreased in a diabetic mouse model, and that its inactivity in mice exposed to high glucose concentrations could be reversed by treatment with deferoxamine, increasing wound vascularization and healing in db/db mice [39].
Nonenzymatic glycation of intra-and extra-cellular proteins by hyperglycemia may cause tissue damage. In fact, AGEs irreversibly bind to proteins. Several studies have shown that AGEs induce endothelial dysfunction. In particular, they increase oxidant stress, induce permeability, stimulate the expression of adhesion molecules and augment the migration of T cells and macrophages [40]. For this reason, we decided to culture cells in the presence of AGE-BSA to mimic diabetogenic conditions. The results of our experiments revealed that all three dermal substitutes (Integra ® , PELNAC ® and Myriad ® ) were able to support at least 10% of proliferating cells in a diabetes-like condition.
We successively focused our experiments on Integra ® and PELNAC ® because they demonstrated the highest vascularization properties and low variability in the results. The angiogenic drive induced by dermal substitutes was evaluated by RT-qPCR for the main pro-angiogenic factors/receptors and tissue remodeling factors. With the exception of TEK, the ANGPT-1 receptor, both matrices (Integra ® and PELNAC ® ) were shown to activate the expression of all genes under study. Furthermore, we also confirmed, at the protein level, the upregulation of VEGF-A synthesis by PELNAC ® of and the upregulation of PlGF secretion by Integra ® .
Inflammation is one of the physiological responses to wound healing [13]. The formation of granulation tissue is a condition sine qua non to complete restitutio ad integrum. Thus, we performed RT-qPCR with the aim of evaluating the expression of healing-promoting cytokines. In this case, Integra ® was shown to be able to upregulate the gene expression of TNF, IL8/CXCL8, MCP1/CCL2 and IL6. These data were partially confirmed at the protein level, since Integra ® seemed to downregulate the production of TNF-α, but also to increase the production of IL-8/CXCL8. In this case, Integra ® also showed the most promising results by upregulating the genes involved in leukocyte recruitment, e.g., IL-8/CXCL8 and MCP1/CCL2, although this also led to the upregulation of TNF-α and IL-6, cytokines involved in chronic inflammation.
Finally, in vivo experiments were undertaken to confirm our in vitro observations. We assessed the ability of preseeded ADMECs on Integra ® to promote the engraftment, integration and vascularization of clinically employed dermal regeneration scaffolds at the site of application. In vivo experiments to evaluate the efficacy of human cell therapies require the use of immunocompromised mice (in our case NSG mice) [14]. Given that they allow the engraftment of human cells, these mice are widely used in preclinical assessments of human cell-based therapies despite the absence of the complex immune response required for wound healing [41].

Conclusions
In conclusion, Integra ® , PELNAC ® and Myriad ® demonstrated the best performance in terms of EC colonization and proliferation, even under hypoxic and diabetic conditions. The interaction of ECs with both Integra ® and PELNAC ® induced the expression of multiple angiogenetic factors. Integra ® was the most effective scaffold in promoting EC engraftment and vascularization in vivo. Thus, the implantation of dermal ECs in combination with ECMs represents a promising strategy for tissue revascularization. Integra ® remains the "gold standard" ADM for this application.

Institutional Review Board Statement:
The study was conducted according to the guidelines of the Declaration of Helsinki with the requirements of the EU Directive 2010/63/EU, and approved by the Institutional Review Board (ICGEB Animal Welfare Board) of ICGEB (protocol code D441.B18 and date of approval 23/03/2017). Ethical review and approval for patient enrolment was obtained by the Regional Ethical Committee of FVG (CEUR), Udine, Italy (Prot. 0010144/P/GEN/ARCS 2019 and Prot. 444 196/2019).

Informed Consent Statement:
Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The data presented in this study are available within the article and in the Supplementary Materials.