Protecting human amnion and chorion matrices during processing: Performance enhancement in a diabetic mouse model and human co-culture system

Recent evidence suggests that protecting human amnion and chorion matrices (HACM) during processing enhances the performance of HACM for wound repair and tissue regeneration. We utilised a diabetic (db/db) delayed wound healing mouse model. Treatment of db/db full-thickness excisional wounds with HACM, processed with a polyampholyte preservative accentuated the proliferative phase of wound healing that decreased the time necessary to heal wounds. Polyampholyte protection improved the preservation of growth factors and cytokines during room temperature storage following E-beam sterilisation and improved its function in wound healing applications. Our findings indicate protected HACM tissue up-regulated MIP2, NF-kB, TNF-α , KI-67

macrophage plasticity including CLC7, CD209, CD36, HSD11B1, ICAM1, IL1RN, IL3RA, ITGAX, LSP1, and PLXDC2 (adj. p-value < 0.05). The polyampholyte alone group demonstrated statistically significant down-regulation of four genes ADRA2, COL7A1, CSF3, and PTGS2 (adj. p < 0.05). The HACM alone group up-regulated four genes ATG14, CXCL11, DNMT3A, and THBD, but the results were not statistically significant. Biomechanical measurements indicated that wounds treated with polyampholyte-protected HACM had more tensile integrity compared with wounds treated with HACM alone. These findings indicate that better protection of HACM during processing stabilises the HACM matrix, which may lead to improved wound healing outcomes. responses, such as cell proliferation and angiogenesis, which can enhance healing of impaired wounds. 6,7 HACM also has a preponderance of exosomes and endosomes, which contain intrinsic signalling components (including proteins and nucleic acids), to critical to the wound repair and regeneration process. 8 Much of the improved healing with HACM in clinical studies has resulted from advances in processing of these tissues to better preserve and to protect the important growth factors and signalling molecules in these matrices. [9][10][11] Continued innovation and advances in the HACM tissue processing to increase storage stability and the off the shelf availability while also preserving the biologically active components is paramount to improving its clinical effectiveness.
Several processing variations of the full-gestation foetal membranes have been developed to establish unique and patentable improvements of the HACM. Early methods of processing entailed separation of the amnion from the chorion. However, with emerging literature on the unique properties of the interface of the two biologic layers in endosomal exchange, 12 immune tolerance, risk surveillance, and inflammation modulation by the placental tissues, it is now considered that the separation of the amnion and chorion may not be as effective. Initial iterations of processing of intact HACM were based on dehydration of the placental tissues at 35 C-50 C for 30-120 min. 13 Improved preservation techniques were then built on the dehydration concept, by using polar organic solvents, such as ethanol or DMSO (dimethyl sulfoxide), followed by lyophilization at À50 C to 80 C. 14 Contemporary processing has integrated other cryoprotectants besides DMSO to improve tissue preservation with a goal of retaining subcellular bioactive components, such as exosomes, that are known to play a crucial role in many aspects of wound healing and angiogenesis. [15][16][17][18][19] A large body of work has been developed to address the need for cryoprotection of cells and tissue during longterm storage. Cryoprotection not only improves cell viability and phenotypic identity but also retains sufficient plasticity for stemness in cells, such as mesenchymal stem cells (MSC). 20,21 More recently, cryopreservation of cells and tissues has been performed using a polyampholyte material that contains poly amino acids with the combination of monomeric subunits of positive and negative charges, which are less toxic than DMSO and requires no washing after thawing the cells and tissues. One example of a polyampholyte is ε-poly-L-lysine blocked with succinic anhydride as describe by Matsumura et al. 20 However, less understanding exists on the ability of a polyampholyte cryoprotectant to preserve function, such as cytokine release and exosomes in the HACM, after lyophilization, or freeze-drying.
This study was designed to determine if a polyampholyte (PolyA) cryoprotectant with an appropriate (proprietary) ratio of amino and carboxyl groups with higher cryopreservation efficiency and lower cytotoxicity could be used to protect HACM during processing in order to preserve the molecular composition of native placental tissues that could benefit its performance in treating delayed wounds.
This study was further guided by a general awareness that all wounds vary not only in size and depth but also in the genomic and metabolic potential of a healing response. Thus, a mouse wound model of delayed healing in db/db mice was used that allowed for a meaningful comparison of wound healing with HACM with and without the additional polyampholyte protection during manufacturing. To further understand the effect of early induction of gene expression by HACM with or without the polyampholyte protection, treated co-cultures of human macrophages and fibroblasts were profiled using NanoString with a custom code set of 282 genes involved in the different phases of wound healing, fibrosis, and macrophage plasticity. These in vitro and in vivo models were chosen to reflect the complexity of the epidermal/dermal stratification of the tissue and to permit meaningful comparison in aspects of tissue integrity that might translate to improved clinical outcomes.

| HACM preparation
Human placental membranes, including both the amnion and chorion layers, were sourced from full-term caesarean section with donor consent. The membranes were aseptically cleaned and processed per standard protocol to remove excess blood and to separate the foetal membranes from the placental decidua. 22 The amnion and chorion layers were left intact such that the intermediate layer, or spongy layer, remained. The tissue was divided into two equal portions. One portion was kept in PBS, and the second was immersed in a polyampholyte cryoprotectant and allowed to soak for 1 h at ambient temperature and atmospheric pressure. Both HACM groups were then lyophilized using a Vivex proprietary process and sterilised with electron beam (e-beam) irradiation (25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35). The total number of membranes processed for the study was 53 per 2 donors. HACM sheets were prepared ( Figure S1) and were first characterised in vitro to determine (1) exosome distribution and microRNA composition as described below, and (2) protein content using an antibody array to characterise the inflammatory, proliferative, and angiogenic factors typically expressed in tissue repair and regeneration. The HACM sheets were then used to treat (1) dorsal full-thickness wounds in the db/db mouse model to evaluate the wound healing response, and (2) in vitro co-cultures of human macrophages and fibroblasts to characterise changes in gene expression afforded by the protection of HACM with polyampholyte. Dragovic et al. 23 The instrument measures the Brownian motion via laser light scattering to count and to size the nanoparticles per millilitre (mL) in a liquid suspension. The measured change in location within a certain time interval provides a specific diffusion coefficient for each individual particle. Using the Stokes-Einstein relation, the ZetaView then calculates the hydrodynamic particle diameter in nanometres (nm) of each particle. Samples were diluted in distilled water (1:500-1:2500) resulting in 100-400 vesicles per frame. For the Zeta-View, the sensitivity was set to 85, the shutter to 150, and the frame rate to 30 FPS. Data were analysed with the ZetaView software (ver-

| Proteomic analysis of the HACM products
Proteomic evaluation of the factors released from HACM and HACM with the polyampholyte was conducted using the Quantibody ® Human Kiloplex Array (RayBioTech, Norcross, Georgia) as described by McQuilling et al., 7 with two matched donors per group. Tissue samples (2 cm Â 2 cm) were incubated at 37 C, 5% CO 2 for 24 h in 4 mL of DPBS. After incubation, the supernatant was centrifuged at 18,000 relative centrifugal force for 10 min. Aliquots were stored at À80 C until the day of the assays. A limit of detection (LOD) for each protein was provided by RayBioTech. Average values of each protein were determined with all values below the LOD considered as zero protein presence.

| In vivo mouse wound model
A diabetic mouse wound healing model was used to compare HACM with polyampholyte protection during processing with untreated HACM as a control. Forty-eight diabetic mice were randomised into this 20-day study. The groups of the study included HACM, HACM + polyampholyte, untreated (Àcontrol), and platelet-derived growth factor (PDGF)-β + transforming growth factor-alpha (TGF-α; +control). The goal was to characterise the effect of HACM preservation on in vivo performance over a 20-day period. Figure S2  The dorsal skin was shaved and cleaned with isopropanol wipes. A sterile 8-mm punch biopsy was used to form a circular full-thickness wound on each animal. Wounds were rinsed with saline. Wounds were treated with either HACM or HACM that had been preserved with polyampholyte (labelled as HACM + PolyA), and then they were covered with a sterile Tegaderm™ dressing. Dressings were applied on Days 0 (immediately after wound creation), 7, and 14 post wounding, and wounds were monitored daily. On the day of the dressing changes, the HACM or HACM + PolyA was replaced in addition to the Tegaderm cover dressing. Negative controls consisted of wounds that were washed with saline alone as a standard procedure and covered with Tegaderm™. Positive controls received a topical application of PDGF-β (10 μg/mL) and TGF-α (1 μg/mL) prepared in 0.5% hydroxypropyl methylcellulose. A 100 μL dose was given to each animal for the first 7 days of the study in the positive control group.
Gross photographic images were recorded at each dressing change while the animals were under inhalational isoflurane anaesthesia. The effect of the HACM on wound healing was evaluated by measuring wound closure. Wound area measurements were calculated using ImageJ to define and measure the wound bed area. 27 Table S1, which includes the thermal cycling parameters. The CFX Maestro software 3.1 was used to quantitate the relative gene expression and the p values for statistical significance. F I G U R E 1 Exosome size distribution in the two human amnion and chorion matrices (HACM) groups using ZetaView Particle Tracking Analyser. For HACM + polyampholyte, the exosome count was higher than for HACM alone for a matched tissue donor sample. The exosomes in the HACM group had a median diameter of 141.6 ± 4.8 nm (from 11 measurements) while the HACM + polyampholyte group had a median diameter of 113.9 ± 4.4 nm (from 11 measurements), suggesting the polyampholyte preservation of HACM better preserved an abundance of smaller diameter exosomes. HACM tissue groups revealed a greater total concentration of particles/mL present in HACM with the polyampholyte cryoprotectant than in HACM alone.

| Immunohistochemistry and histology
F I G U R E 2 Exosomes were analysed using next generation sequencing. Analysis of miRNA content showed that human amnion and chorion matrices (HACM) + polyampholyte had similar miRNA cargo, including miRNA-148a and miRNA-21 as HACM alone.

| Biomechanical measurements
The animals were euthanized at 20 days and a 1 cm

| Genomic profiling of human macrophages and fibroblasts
Co-cultures of human macrophages and fibroblasts treated with HACM preserved with polyampholyte compared with HACM or polyampholyte alone were tested to better understand the effect of early induction of gene expression by treatments. Twelve-well cell culture plates were adapted to allow for co-culture of human fibroblasts and macrophages ( Figure S3). A hole was drilled with 3 /4 00 drill bit between wells in a 12-well plate. The chamber in between the plates was sealed at the edges with food grade silicone to prevent leaking from the inner chamber. Two millilitre of purified human monocytes (Creative Biolabs-IC-LX032) were added to a co-culture well at a seeding density of $2.7 Â 10 5 cells/mL and incubated overnight.
Once settled, the cells were activated with IFN-γ (100 ng/mL) for , F I G U R E 3 Heat map of proteomic composition of the human amnion and chorion matrices (HACM) and HACM +polyampholyte (labelled as HACM + PolyA) determined using an antibody array. The protein composition was similar among the inflammatory, proliferative, and angiogenic biomarkers examined. Colour Key (pg sample/cm 2 : Blue 0-100, dark green 100-200, light green 2000-5000, orange 5000-10,000, and red>10,000. proprietary Wound Healing Codeset that has 282 genes that cover the phases of wound healing (haemostasis, inflammation, proliferation, and tissue remodelling), fibrosis and macrophage plasticity.

Detailed bioinformatics analysis is provided in the Supplementary
Methods S1 section.

| Statistical analysis
Statistical analysis was performed using GraphPad Prism 9.2 (San Diego, California) with the differences between the negative control group and the treatment groups analysed by two-way analysis of variance, p < 0.05 considered as statistically significant. Results were expressed as mean ± standard deviation.         important roles in wound re-epithelialization, angiogenesis, and macrophage differentiation, including mir-21, and mir-148a. [35][36][37][38][39] Proteomic analysis of the HACM and preserved HACM samples demonstrated that the level of inflammatory, proliferative, and angiogenic markers were similar (pg/cm 2 ). However, more work will be required to determine whether the polyampholyte protection leads to better integrity of these proteins and perhaps to a slower release profile over the time course of healing, attributable to complex matrix binding that is known to be modulated by intrinsic membrane ligands such as integrins that guide migration, homing, macrophage modulation, and effect release and angiogenesis across not only species but across foetal-maternal barriers. 40 Genomic profiling data in vitro indicates that polyampholyte better protects the HACM, with consequent effects on macrophage plasticity. We noted 12 genes were significantly up-regulated by the polyampholyte-treated HACM. These genes primarily play a critical role regulating the plasticity of macrophages (CCL7, CD209, CD36,   HSD111B1, ICAM1, IL1RN, IL3RA, ITGAX, LSP1, PLXDC2, SPI1, and   TLR2). Specifically, CCL7 is a chemokine that is known to increase macrophage chemotaxis potential, which is consistent with the enhanced proliferative phase of wound healing in our animal model. 40 CD209 is a C-type lectin that functions in cell adhesion and pathogen recognition and ITGAX (CD11c) is the integrin alpha X chain protein, both of which are known as pro-M2 phenotype biomarkers. 41,42 CD36 is a well-characterised macrophage receptor that binds and phagocytoses apoptotic cells and other oxidised lipids. 43  It has been demonstrated that a knockout or deficiency in HSD111B1 results in inflammatory angiogenesis. Toll-like receptor (TLR) family play a critical role in pathogen recognition and activation of innate immunity. The ratio of TLR4/TLR2 is known to effect macrophage plasticity. 47 Up-regulation of TLR4 switches macrophages to the proinflammatory state and TLR2 promotes the switch the anti-inflammatory phenotype. The proteins LSP1, PLXDC2, and SPI1 play an important role in macrophage chemotaxis and motility. [48][49][50] Our findings raise multiple important questions including, (1) does size of the exosome matter in terms of whether the recipient cells that uptake small versus large exosomes could be different, which could lead to differential wound inflammatory and healing responses, macrophages also secrete high levels of PDGF, which has been shown to promote angiogenesis. 54 Biomechanical measurements of the healed tissue indicated that polyampholyte preserved HACM had better integrity, and when compared with the untreated wounds, it was statistically significant ( p < 0.05). From the perspective of wound closure, the preserved HACM appear to promote the healing of diabetic mouse wounds, but additional samples will be required to reach statistical significance.
Both DFU and VLU are prone to recurrence at staggering rates of 20%-85%. 55

ACKNOWLEDGEMENTS
The preclinical studies were performed in collaboration with Alpha Preclinical LLC, Grafton, Massachusetts. IRB review was not required for laboratory research on human cells obtained from tissue repositories/banks.

CONFLICT OF INTEREST STATEMENT
T. Ganey, WB Martin, H. Kaliada, S. Littlejohn are current or former employees of Vivex Biologics.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.