Enhanced Skin Regeneration Using a Novel Amniotic-derived Tissue Graft

Background. Chronic and recalcitrant wounds present a significant therapeutic challenge. Amniotic tissues contain many regenerative cytokines, growth factors, and extracellular matrix molecules including proteoglycans, hyaluronic acid, and collagens I, III, and IV. Dehydrated amnion/chorion grafts are currently used to treat a variety of wounds such as diabetic foot ulcers and burns. Objective. The investigators hypothesized that processing methodologies, dehydration, and hypothermic processing and storage of amniotic tissues would affect overall quality of wound healing; they compared dehydrated amnion/chorion (dHACM) grafts to a novel hypothermically stored amniotic membrane (HSAM) graft in a full-thickness rat wound model. Materials and Methods. Sprague-Dawley rats were anesthetized and prepped for surgery; four 1.5-cm diameter full-thickness wounds were created and treated with either: (1) dHACM, (2) dHACM meshed, (3) HSAM, or (4) wound left ungrafted (sham). After 9 or 21 days, wounds and surrounding areas were collected and stained with hematoxylin and eosin. Blinded quantitative analysis of quality of wound healing was completed by evaluating hair follicle/gland formation, dense/scar-like matrix, and basket-weave matrix. Results. At varying time points following placement of the grafts into full-thickness defects, the authors found that all amniotic-derived tissue grafts appeared to stimulate improved healing over sham wounds, evidenced by more normal-appearing dermal matrix architecture, epidermal structure, and maturity. In addition, the HSAM grafts promoted greater tissue regeneration than the dHACM meshed grafts, as measured by the presence of basket-weave collagen matrix and formation of follicles and glands. Conclusions. In sum, this study builds on the amassing literature supporting amniotic tissues for wound repair and demonstrates the importance of tissue processing on the quality of wound healing. 

In 2012, about 9.3% of the US population had diabetes, and this population has grown rapidly over the past several decades.¹ Patients with diabetes have a 15% lifetime risk of developing foot ulcers. These chronic wounds are attributed to a multitude of factors, including bacterial infection, tissue ischemia, continuing trauma, and poor management. A chronic wound is commonly defined as an open wound of the skin taking more than 8 weeks to heal.² One major factor thought to be responsible for the development of chronic wounds is the impairment of cytokine release by cells in the local wound environment including fibroblasts and inflammatory cells, which may play a role in the downregulation of angiogenesis.³ These nonhealing chronic wounds often require surgical intervention, advanced treatment options, and, in some cases, amputation. 

Diabetic foot ulcers (DFUs) are associated with high mortality rates; for example, a retrospective study by Moulik et al⁴ reported 5-year mortality rates of 44% for patients with new-onset DFUs. Effective treatment of DFUs depends largely on debridement and patient compliance with offloading. Secondarily, there are a variety of types of wound dressings suited to address various local ulcer environments. For example, alginates and foams are often used to adsorb fluid and control moisture in the wound. Hydrogels and silver dressings may be used to control eschar and reduce rates of infection, respectively.⁵ In addition, more advanced biological treatments, including topical enzymes and growth factors, have been used to modulate the wound environment and promote wound healing.⁶

Historically, human amniotic membranes have been used as a wound dressing to manage burns, ulcers, and infected wounds.^7-9 In fact, usage of amniotic membranes as a biological dressing for wounds was first documented in the early 1900s.^10,11 The first successful application of amniotic membrane to treat chronic skin ulcers was observed more than 50 years ago.¹² But despite this advancement, the use of human allografts decreased in the years following due to the increased prevalence of communicable diseases such as human immunodeficiency virus and hepatitis. However, the subsequent development and implementation of robust testing modalities have resulted in safer implantation of human grafts, leading to a resurgence in the popularity of human allograft tissue.

Recently, various commercially available dehydrated placental grafts composed of amnion and/or chorion have been used to treat DFUs clinically.^13-17 In these studies, clinical benefit is evident by the overall positive outcomes using placental-derived grafts to treat difficult-to-heal wounds. These results, along with preclinical data supporting the use of amniotic cells and tissues in wound healing,^18,19 highlight a promising role for these tissues in a range of therapeutic applications.

The goal of the current study is to evaluate the effect of different processing modalities including dehydration, meshing, and hypothermic storage of amnion on wound healing outcomes. The investigators compared 3 uniquely processed human amniotic-derived allografts using a full-thickness rat wound model as described in Bonvallet et al.²⁰ Specifically, they created 4 circular full-thickness wounds per rat (d = 1.5 cm), and for each wound the investigators placed a human placental-derived allograft or alternatively left the wound without a graft (sham).  All wounds were then covered with a nonocclusive, nonadhesive dressing (Cuticerin; Smith & Nephew, Memphis, TN) for the duration of the study. Wound healing was assessed histologically by evaluation of the epidermal layer, abnormal scar tissue formation, basket-weave matrix, and the presence of dermal appendages. The central hypothesis was that amniotic-derived grafts would promote more robust tissue regeneration than sham-treated wounds and that differences in processing of the amniotic grafts would have an effect on the overall quality of wound healing.

Preparation of amniotic-derived grafts
All amniotic-derived products were donated by NuTech, a division of Organogenesis (Birmingham, AL). For human amniotic-derived allografts, all applicable standards of the American Association of Tissue Banks and United States Food and Drug Administration were followed. Processing, dehydration, and sterilization of amnion/chorion (dHACM; NuShield; NuTech, a division of Organogenesis) tissues were completed via proprietary techniques, and tissue grafts were stored at room temperature. In addition, for hypothermically stored amniotic membrane (HSAM; Affinity; NuTech, a division of Organogenesis), a proprietary storage solution was developed by NuTech to preserve cell viability and membrane integrity of the amniotic membrane. Importantly, for HSAM grafts, these tissues were stored for about 20 to 30 days at 1°C to 10°C before use in this study to mimic clinical conditions. All amniotic-derived grafts were implanted using the same orientation, with the stromal or chorion side of the graft in direct contact with the wound bed.

Amniotic graft implantation into full-thickness skin defects
All animals received humane care and all procedures were performed with prior approval from the Institutional Animal Care and Use Committee (IACUC; protocol # 140909448) at the University of Alabama at Birmingham (Birmingham, AL). Amniotic-derived tissues were prepared and cut to size for implantation into full-thickness wounds immediately before placement. For the authors’ studies, Sprague-Dawley rats (n = 8; 4 rats/time point) were anesthetized with isoflurane, and the wound areas were prepared by shaving the hair on their backs with shears. Alternate swabbing of the surgical area with betadine and ethanol was used to prep the area, and followed by the creation of four 1.5-cm diameter full-thickness wounds in the back skin. Each wound was subjected to 1 of the following treatments: (1) dHACM, (2) dHACM meshed, (3) HSAM, or (4) wound left ungrafted (sham). Grafts naturally wicked to the wounds and therefore were not sutured in the wounds. All wounds were covered with nonocclusive, nonadhesive dressing.  At postimplantation days 9 and 21, animals were euthanized with an overdose of carbon dioxide exposure and confirmation of euthanasia by thoracotomy; top-down photographs of the wound surface were taken, and the wounded area and surrounding tissues were harvested.

Samples were formalin-fixed and paraffin-embedded. These samples were then sectioned through the center of the wounds, and the sections were mounted and stained with hematoxylin and eosin.

Imaging and tissue analysis
Whole-section images were taken on a Nikon Ti-E inverted microscope (Nikon, Tokyo, Japan) and stitched together using NIS Elements software (Nikon). For all slide analysis and quantification, a blinded study team member completed analysis with no knowledge of the groups. From whole-section images, the area of abnormal tissue/scar tissue and basket-weave collagen matrix resembling native skin tissue were measured using ImageJ (National Institutes of Health, Bethesda, MD) to quantify the number of pixels in each group. Whole-section images also were used to assess the maturity of the epidermal layer and count the number of newly formed dermal appendages, including hair follicles and glands, within the wound environment. For this analysis, all dermal appendages were counted in an identical fashion without regard to the maturity of the appendage. High magnification representative images also were taken in order to display skin regenerative characteristics such as reepithelialization, blood vessel formation, and follicle and gland formation.

Statistics
For statistical analysis of area of abnormal tissue/dense collagen/early scar and basket-weave matrix, ImageJ was used to analyze the areas of interest for each group, and the number of pixels in each area of interest then was used for statistical analysis. For quantification of follicles/gland formation, whole-section images were evaluated and follicles/glands were counted and recorded. For quantification of all data, one-way analysis of variance (ANOVA) comparisons were performed to evaluate differences between groups at days 9 and 21 using Minitab Statistical Software 17 (Minitab Inc, State College, PA).

Treatment groups
Amniotic membranes exposed to 3 distinct types of processing methods were examined (Table): (1) dHACM, (2) dHACM meshed, or (3) HSAM. All of these amniotic-derived grafts contain extracellular matrix (ECM) molecules and cytokines; however, some of the key differences between the dHACMs and the HSAMs are: (1) HSAM have viable cell content; (2) HSAM have the amnion portion present while dHACM consists of amnion and chorion; and (3) the ECM structure of HSAM is preserved compared with the condensed ECM in dHACM. These differences are presented in a schematic representation of the grafts in Figure 1.

Histologic evaluation
Epidermal layer. As an indicator of wound maturity, the epidermal layer was evaluated for all wounds, and at 9 days it was found that none of the sham wounds had reepithelialized. For the dHACM groups, some of the wounds healed with limited epithelialization. In contrast, the HSAM-treated wounds that reepithelialized at 9 days exhibited a stratified epidermis with finger-like projections extending into the dermal layer, representative of healthy epidermal healing (Figure 2).  At 21 days, all wounds in all groups had achieved reepithelialization across the wound bed, with the HSAM group exhibiting a thinner epidermal layer with robust finger-like projections into the dermis (Figure 3).

Histologic summary. High-magnification images were used to qualitatively assess the wounds, and representative images are shown in Figure 4. Amniotic-derived grafts stimulated more robust healing and wound repair than sham wounds. Interestingly, HSAM-treated wounds displayed early epidermal formation, reconstitution of dermal appendages, and a high degree of basket-weave matrix, thus producing regenerated skin tissue that closely mimics unwounded skin at 21 days. 

Extracellular matrix morphology. An important feature of regenerated skin is the architecture of the collagen matrix. The dermal matrix of normal skin has a loose wavy appearance²¹ described as a basket-weave structure. This structure was observed in many of the samples and was evident when compared with the dense matrix suggestive of scar tissue; therefore, the area of basket-weave matrix and dense tissue was measured and used together as an indicator for quality of wound healing.

At postimplantation day 9, all groups had large quantities of dense matrix present; however, HSAM showed a trend toward decreased quantity of this tissue (Figure 5A). Similarly, at 21 days, while dense matrix decreased in all groups compared with 9 days, treatment with HSAM showed a trend toward less scar-like tissue than other groups. These results inversely corresponded to the quantification of basket-weave matrix present within the wound bed (Figure 5B). The HSAM-grafted wounds exhibited significantly more basket-weave matrix than sham wounds at both 9 and 21 days. Furthermore, at 21 days, treatment with HSAM resulted in significantly greater basket weave formation than dHACM meshed grafts. 

Quantification of follicle and gland formation during wound healing
Whole-section images were used to quantitatively evaluate the formation of dermal appendages. To do this, individual appendages within the wounded area were counted for each group and these data were reported as the average per group. At 9 days, all groups had limited numbers of follicles and glands present within the wound bed; however, at postimplantation day 21, the HSAM-treated wounds had statistically more dermal appendages than the sham and dHACM meshed groups (Figure 6). Interestingly, the dHACM graft showed a trend toward greater follicle formation than dHACM meshed graft, and a similar trend toward increased basket-weave ECM (Figure 5) also was noted in the dHACM sample. While further studies are needed, these data hint that meshing may have negatively influenced graft performance. 

However, regardless of the 2 different processing methods for dHACM grafts, the HSAM grafts appeared to elicit a better healing response than all other groups.

This study, in agreement with other investigations evaluating placental-derived tissues for the treatment of wounds,^2,7,18,22-25 supports the potential efficacy of such tissues in promoting robust wound healing. Interestingly, the results reported herein and others^22-24 have highlighted the important role tissue-processing techniques may play in the overall bioactivity of the tissue and the healing potential of placental-derived allografts. The apparent differences in the wound healing response elicited by HSAM compared with dHACM raise many questions regarding the mechanism underlying enhanced healing. One potential explanation relates to the importance of the ECM in regenerative healing.²⁶ The ECM is a critical regulator of cell phenotype and function, and it also acts as a reservoir for cytokines and growth factors.^27,28 In the samples used for this study, HSAM processing results in an open structure whereas dHACM processing causes a collapse in the ECM due to the removal of water. The open ECM of HSAM may allow for better cell attachment and migration into the graft. Furthermore, supporting research has shown that dehydration of various tissues induces protein denaturation leading to diminished function, damage to the ECM ultrastructure and material properties of the tissue, and reduction in cell infiltration.^26,29,30 As well, compared with cryopreserved tissue, dHACM were deficient in key biological signals²² and have reduced anti-inflammatory²³ and angiogenic activity.²⁴

In addition to the role of ECM integrity, it is possible that the presence of viable cells within HSAM grafts may have enhanced the wound-healing process. While the value of the cellular content of HSAM has not been investigated to date, there have been several studies pointing to the therapeutic effects of a variety of stem cell types, including human bone marrow stem cells, adipose-derived stem cells (ADSCs), bulge hair-follicle stem cells, and amniotic fluid-derived stem cells, in diverse wound-healing animal models.^31-34 An important role for adult mesenchymal stem cells (MSCs) in wound healing was displayed in a study by Sasaki et al.³⁵ In this study, green fluorescent protein (GFP) labeled MSCs were injected into the tail veins of C57BL/6 mice, and it was found that GFP-MSCs migrated to the wound environment and differentiated into keratinocytes, endothelial cells, and pericytes. Interestingly, mice treated with exogenous MSCs also exhibited a 2.4x reduction in wound size at 8 days compared with control mice.³⁵ In addition to direct differentiation of stem cells into wound-specific cell types, the supportive role of viable cells in the wound environment is also of interest. Kim et al³⁶ evaluated the effect of ADSCs on human dermal fibroblasts using coculture. This work established that direct cell-cell contact led to greater fibroblast proliferation, and this study also confirmed paracrine activation of enhanced collagen I secretion and cell migration.³⁶ Taken together, these studies suggest a potentially complex role for viable stem cells within the wound environment. Several groups have evaluated the role of various stem cells in wound healing, and in general these studies have highlighted positive effects on dermal fibroblast proliferation, migration, and production of important matrix components.^37-39 Recent studies evaluating amniotic-derived tissue as a source of stem cells have pointed to several types of stem cells present within the tissue, including amniotic fluid-derived stem cells and stem cells from the amniotic membrane.^40-42 A study directly comparing wounds treated with MSCs from amniotic tissue to adipose-derived MSCs in a diabetic NOD/SCID mouse reported significant improvements in wound closure rates and histological wound scores with MSCs derived from amniotic tissue.²⁵ Mechanistically, the authors pointed to increased paracrine and engraftment effects of amniotic-derived MSCs.²⁵

Despite the robust effects seen in this study and others using transplanted cellular tissues, concerns remain regarding the immunogenicity of nonhomologous graft materials. In this study, the investigators used an immunocompetent rat model and saw no evidence of an immunogenic response. In addition, amniotic tissues have been reported to express human leukocyte antigen-ABC (MHC class I) but lack the expression of human leukocyte antigen-antigen D related (MHC class II), resulting in low immunogenicity after cell implantation.^41-45 Further, some groups have evaluated the immunomodulatory effect of stem cells from the amniotic membrane.^23,46-49 For example,  Alikarami et al⁴⁶ found that MSCs derived from amniotic membrane could suppress the proliferation of  T lymphocytes and significantly decrease the production of interferon gamma by T cells in the absence of blood monocytes.⁴⁶

While this study presents interesting results regarding amniotic tissue processing and how it may effect the quality of wound healing, limitations include the use of a small animal model and lack of mechanistic data. Future research will focus on mechanisms underlying the effectiveness of differentially processed grafts as well as studies of wound healing in larger animal models. Likewise, additional research is needed to elucidate the clinical significance of the histologic findings of this study, which implicate a beneficial effect of HSAM on regenerated tissue morphology.

Current studies point to several critical factors that may contribute to enhanced wound repair with amniotic-derived tissues including ECM, cytokines and growth factors, stem cells, and immunomodulation of the wound environment.^50,51 This study adds to the overall evidence supporting the use of amniotic-derived tissues in regenerative applications by highlighting the importance of processing techniques,and how they influence the quality of wound healing. 

The authors are grateful for the assistance of the University of Alabama Comparative Pathology Core Facility (Birmingham, AL). 

Affiliations: Department of Research and Development, NuTech, a Division of Organogenesis, Birmingham, AL; and Department of Cell, Developmental and Integrative Biology, University of Alabama at Birmingham, Birmingham, AL 

Correspondence:
Susan L. Bellis, PhD
982A MCLM
1918 University Boulevard
Birmingham, AL 35294 
bellis@uab.edu 

Disclosure: Dr. Mowry is employed by NuTech, a Division of Organogenesis (Birmingham, AL). This research was supported by NuTech, a Division of Organogenesis.