ORIGINAL RESEARCH published: 27 May 2020 doi: 10.3389/fimmu.2020.00942 Cord Blood Platelet Rich Plasma Derivatives for Clinical Applications in Non-transfusion Medicine Dinara Samarkanova 1,2,3*† , Steven Cox 4,5† , Diana Hernandez 4,5 , Luciano Rodriguez 1 , Ricardo P. Casaroli-Marano 1,6 , Alejandro Madrigal 1,4,5 and Sergio Querol 1,2 1 Banc de Sang i Teixits, Barcelona, Spain, 2 Transfusional Medicine Research Group, Vall d’Hebron Research Institute (VHIR), Barcelona, Spain, 3 Department of Medicine, Universidad Autónoma de Barcelona (UAB), Barcelona, Spain, 4 Anthony Nolan Research Institute, Royal Free Hospital, London, United Kingdom, 5 UCL Cancer Institute, Royal Free Hospital, London, United Kingdom, 6 Department of Surgery, School of Medicine & Hospital Clinic de Barcelona, University of Barcelona, Barcelona, Spain Cord blood platelet rich plasma (CB-PRP) derivatives have been investigated as potential therapeutic agents for the treatment of diverse conditions including ocular surface disease and skin ulcers. We have developed processes for the formulation of several Edited by: CB-PRP preparations, which have different composition and attributes. Here we describe Helena Stabile, Sapienza University of Rome, Italy the molecular characteristics of these preparations and we make recommendations as Reviewed by: to their most appropriate clinical application based on functional and immunomodulatory Jose Luis Bueno, profiles. We show that incubation of adult peripheral blood mononuclear cells (PBMCs) Hospital Universitario Puerta de Hierro Majadahonda, Spain with all three preparations dramatically reduced the production of INFγ and the Thierry Burnouf, expression of NKG2D and CD107a in NK, NKT, and T cells thus diminishing their Taipei Medical University, Taiwan activation, we propose that the likely mechanism is the high levels of soluble NKG2D *Correspondence: ligands present in plasma. Of the three preparations we investigated, CB platelet Dinara Samarkanova dsamarkanova@bst.cat lysate (PL) and platelet releaseate (PR) have higher concentrations of trophic and † These pro-angiogenic factors, CB platelet poor plasma (PPP) has the lowest concentration of all authors have contributed equally to this work analytes measured. Based on these finding we propose that CB-PR is the most suitable raw material for skin wound patches, while CB-PL and PPP can be used to prepare eye Specialty section: drops for severe ocular surface pathologies and inflammatory conditions such as corneal This article was submitted to Cytokines and Soluble Mediators in ulcers or severe dry eye disease, respectively. Immunity, Keywords: cord blood, platelets rich plasma, plasma, platelets, ocular surface disease, skin ulcer, a section of the journal immunomodulation Frontiers in Immunology Received: 18 October 2019 Accepted: 22 April 2020 INTRODUCTION Published: 27 May 2020 Citation: Human platelet rich plasma (PRP) has been recognized as a rich source of trophic factors Samarkanova D, Cox S, Hernandez D, (1, 2), which aid healing in several indications including some ocular surface disorders (3, 4), Rodriguez L, Casaroli-Marano RP, such as severe dry eye disease and corneal ulcers, skin wounds (5) and orthopedic applications. Madrigal A and Querol S (2020) Cord Initially, autologous PRP was used to treat corneal injuries (6) but more recently allogeneic Blood Platelet Rich Plasma Derivatives for Clinical Applications in preparations have been favored (7–9). PRP can be obtained readily from both adult peripheral Non-transfusion Medicine. and cord blood (CB). Comparisons of PRP obtained from these two sources suggest that Front. Immunol. 11:942. they differ both in the type and in the amount of factors they contain (10). These and doi: 10.3389/fimmu.2020.00942 other studies argue that CB-PRP has therapeutic advantages over adult PRP as the latter Frontiers in Immunology | www.frontiersin.org 1 May 2020 | Volume 11 | Article 942 Samarkanova et al. CB-PRP Derivatives contains more pro-inflammatory molecules which can aggravate MATERIALS AND METHODS certain conditions, while CB-PRP contains higher levels of anti- inflammatory molecules [reviewed in (11)]. Ethical Considerations CB-PRP contains soluble molecules of NK Group 2, member PRP derived samples from individual CB units were obtained D (NKG2D) ligands (sNKG2DL), which are in part responsible from BioBank (BST). The informed consent for the use of for NK, NKT, and T cells suppression and may represent samples for research was obtained in accordance with the a mechanism of fetomaternal tolerance (12). Further work requirements of the Declaration of Helsinki and local laws at elucidated the differential roles of the various types of sNKG2DLs the time of collection. This study was approved by the BST local and revealed soluble unique long 16 binding proteins (sULBP1) Research Ethics Committees (HCB/2017/0785). as the most abundant soluble ligand in CB-PRP that directly downmodulated NK cell cytotoxicity in a dose-dependent CB Unit Collection manner (13). Additionally, we showed that polymorphisms CB units were collected in authorized maternity hospitals. Before in Major Histocompatibility Complex Class I-related Chain delivery, mothers signed an informed consent for donation that A (MICA) which encode either valine (Val) or methionine allows the use of these samples for research and validation (Met) at residue 129 affected the ability of soluble MICA purposes. Qualified health care professionals collected CB units (sMICA) to suppress NK cell function. Met is known to while placenta was still in utero using validated procedures (19). bind with strong affinity to NKG2D whereas Val has weaker The collection bag contains 25 mL of citrate phosphate dextrose binding (14). (CPD) as anticoagulant. Depending on the processing protocols used to separate the PRP from whole blood, a variety of preparations can be CB Unit Verification made which have different properties. An initial low speed Full serological and microbiological analysis of the CB units centrifugation of whole CB units yields CB-PRP which can was carried out to exclude infection. Tests were performed be further processed. CB-PRP contains a variety of growth for the presence of Hepatitis A, B, and C viruses, human factors (GF), cytokines and other immunomodulatory factors immunodeficiency virus, Treponema Pallidum (Syphilis), that can act on the proliferation, differentiation and function Trypanosoma cruzi (Chagas disease), and cytomegalovirus as of both immune and other cells such as fibroblasts. CB- well as aerobic and anaerobic bacterial and fungi (BacTAlert, PRP also contains albumin, transferrin, fibronectin, fibrinogen, Biomerieux, INC. Durham). Quarantine was applied as a minerals, adiponectin, vitamin A and E, several fatty acids minimum for 2 weeks. Only units, which were non-reactive, and platelets (15), which contribute as trophic factors in were further processed. wound healing mechanisms of epithelial cells of the cornea and its stromal tissue (16). These elements have also been Preparation of CB-PRP Derivatives implicated in cutaneous scarring formation (17). Thus, the Briefly, a total of 20 whole CB (WCB) units were used for availability through CB banking and the therapeutic benefits the experiments. Units were treated individually and from each of the factors contained in CB-PRP derived products make unit we derived all 3 preparations described. Each unit was CB an ideal product source for the treatment of a variety centrifuged at low speed (210 g) for 10 min, the supernatant of diseases and inflammatory conditions. These benefits can collected constitutes the PRP. The PRP was then centrifuged at be enhanced by processing the PRP both by concentration high speed (2,000 g) for 15 min to obtain two fractions, a PPP and of the platelets and their lysis, which releases platelet-derived a platelet pellet, which was resuspended in PPP to obtain a platelet growth factor (PDGF), basic fibroblast growth factor (bFGF), concentrate (CB-PC) in range of 800–1,200 × 109 platelets/L in epithelial growth factor (EGF), vascular endothelial growth 10 (±5) mL and then stored at −80◦ C. CB-PC fraction was frozen factor (VEGF), and transforming growth factor beta (TGFβ) for quarantine. Only microbiology and serology results negative among others, all of which have been shown to enhance tissue samples of CBPC were used for preparation of platelet lysate (CB- regeneration (18). PL) and platelet releaseate (CB-PR). Figure 1 shows a schematic The purpose of this study was to compare the profile of the processing of CB to obtain the different preparations tested. of GFs, cytokine and immune elements of CB-PRP derived preparations to determine which ones would be best suited CB-PPP Preparation for clinical applications. We formulated, from CB-PRP, three The remaining PPP was frozen at −80◦ C for future analysis; this different preparations: (a) platelet poor plasma (CB-PPP), (b) product is CB-PPP. platelet concentrate lysate (CB-PL), and (c) platelet concentrate releaseate (CB-PR). CB-PL Preparation The concentrations of various GFs, cytokines and soluble CB-PC samples underwent 3 freeze (∼-80◦ C)/thaw (∼37◦ C) NKG2D ligands, together with their immunomodulatory effects cycles (20) to lyse platelets and release GFs followed by a on Natural Killer (NK) cells and CD8+ T cells were determined centrifugation step at 5,000 g for 15 min. The collected GF rich in the different CB-PRP preparations. Taken together, the results supernatant free of intact platelets was subsequently stored at can help us determine the optimal CB-PRP derived preparations −80◦ C for future use. This CB-PL, diluted with Plasmalyte for different clinical applications for ocular surface diseases and (Baxter, USA) (1:1, v:v) is used to prepare CB eye drop skin wound healing. (CBED), which was used in clinical trial [ClinicalTrails.gov ID Frontiers in Immunology | www.frontiersin.org 2 May 2020 | Volume 11 | Article 942 Samarkanova et al. CB-PRP Derivatives FIGURE 1 | CB-PRP derivatives processing scheme. NCT03084861]. For this study was used only CB-PL, without TNFα; (plate 2) TGFβ1; (plate 3) HGF; (plate 4) MMP2/9; dilution as the active ingredient. (plate 5)TIMP1-4. CB-PR Preparation CB-PC underwent one freeze/thaw cycle (which in our PBMC Functional Analysis Peripheral blood mononuclear cells (PBMCs) were isolated from experience is insufficient to cause major platelet lysis and whole blood (WB) from consenting, healthy adult male or allows the storage of this fraction ready for calcium gluconate female volunteer donors by density-gradient centrifugation using activation), the resulting fraction termed platelet releasate was Lympholyte R -H solution (Cedarlane, ON, Canada). PBMCs treated with 10% calcium gluconate (Braun R , Terrassa, Spain) were then used for co-culture with the various CB-PRP derived (1:10, v:v) and incubated for 1 h at room temperature. This preparations for functional analysis. For assessment of cellular jellification step forms the basis for the clinical skin patch, function of CD3- CD56bright/dim NK cells, CD3+ CD56+ NKT which has been used for clinical trial (ClinicalTrails.gov ID cell and CD3+ CD56– T cells, healthy adult donor PBMCs NCT02389010). For the purpose of the analyte measurement were plated in RPMI (Lonza, Slough, UK) containing 10% in this study, the releasate was also treated with 10% calcium heat-inactivated fetal calf serum (FCS) supplemented with 1% gluconate (as above) but in the presence of heparin at 0.61 IU/mL penicillin and streptomycin (complete media) and containing to prevent clotting and thus the jellification of the plasma and human IL-2. Test cultures using CB-PRP preparations: CB- trapping GFs, which would have confounded the measurements PPP, CB-PL or CB-PR were diluted with an equal volume of factors in this preparation. After a centrifugation step (5,000 g of complete media containing 200 IU IL-2 per well (200 µl for 15 min) the supernatant free of intact platelets (checked volume) in 96-well U-bottom plates. Results were normalized by automatic counter-Beckman-Coulter) was used for analyte against PBMCs cultured with complete media containing 200 measurements or stored at −80◦ C until needed. IU IL-2 per well and expressed as percentage of maximum expression (for measurement of NKG2D, CD107a or IFN- CB-PRP Derived Products Cytokines and γ expression). All cultures were incubated at 37◦ C with 5% GF Quantification CO2 for 48 h before Phorbol 12-myristate 13-acetate (PMA) CB-PRP preparations of CB-PPP, CB-PL, and CB-PR from and ionomycin stimulation. Duplicate cultures were carried out single units were analyzed using magnetic beads of Luminex for each experiment without PMA and ionomycin stimulation kits (R&D Systems, Abingdon, UK) according to manufacturer’s to assess levels of NKG2D and baseline CD107a expression. recommendations, to determine concentration of following NKG2D expression on the relevant cells was determined using analytes: (plate 1) PDGF AB/BB, EGF, bFGF, VEGF, IL1/6/10, the unstimulated cultures and CD107a background levels in Frontiers in Immunology | www.frontiersin.org 3 May 2020 | Volume 11 | Article 942 Samarkanova et al. CB-PRP Derivatives unstimulated cultures were subtracted from the values obtained AGCMTCCAACAAC-3′ ). Cycle sequencing was performed for equivalent stimulated cells. Three independent experiments using standard protocols and exon 3 forward sequencing primer were carried out using four different, healthy PBMC donors, (5′ - CCCTGGGCTGAGTTCCTC-3′ ) and reverse sequencing each tested with CB-PPP, CB-PL, CB-PR, or media only and data primer (5′ - ATAGCACAGGGAGGGTTT-3′ ). points represent donor means for each replicate test sample. Statistical Analysis Flow Cytometry Results are shown as mean with standard error of the mean Briefly, cells were labeled with fluorochrome-conjugated (SEM) and were evaluated using Graphpad Prism 8 (Graphpad antibodies in Phosphate-buffered saline (PBS) with Bovine Software, CA, USA. Datasets were analyzed using both paired and Serum Albumin (BSA) (0.5%) for 10 min at 4◦ C. Antibodies unpaired non-parametric statistical analysis as the sample size is (BD Biosciences, Oxford, UK) were as follows: CD3 (SK7), too small to assume normal distributions. As we had 3 categories CD56 (B159), CD107a (HA4A3), NKG2D (BAT221, Miltenyi of sample, we used Friedmans’s test for the paired comparisons Biotec, Bisley, UK). Viability was assessed using Annexin V and and a Kruskal-Wallis for the unpaired analysis. Significance levels 7-AAD. For quantitation of CD107a, cells were re-suspended in are indicated as p < 0.05 (∗ ), p < 0.01 (∗∗ ), p < 0.001 (∗∗∗ ), complete media containing 100 ng/ml PMA, 1 µg ionomycin and p < 0.0001 (∗∗∗∗ ), unless the exact p-value is given. For and 0.1% 2-mercaptoethanol (stimulated) or complete media comparisons between genotypes only we used Mann-Whitney. with 0.1% 2-mercaptoethanol (non-stimulated) for 2 h at 37◦ C. Fluorescence minus one (FMO) controls (where samples are stained sequentially with all antibodies except one) were used RESULTS to set gates. Acquisition was performed using a Fortessa II flow cytometer (BD Biosciences, Oxford, UK) and analysis Comparison of Cytokine and GF was carried out using FlowJo version 10.5.0 (Tree Star Inc., Concentration Among the Different OR, USA). The gating strategy used for analysis of lymphocyte Preparations of CB-PRP Derivatives subtypes was performed as described in our previously study The prepared CB-PC units contained 1,004 ± 76 × 109 platelets (12) and is shown in Supplementary Figure S1. (PLT)/L (n = 10) and were within the standard range of 800– 1,200 × 109 PLT/L, which was used to prepare CB-PL and CB-PR. Soluble NKG2DL and IFN-γ Detection and The levels of trophic and wound healing factors (group I), pro- Quantification Assays angiogenic molecules (group II), pro-inflammatory (group III.A) Soluble MICA/B (DY1300/DY1599) and ULBP1 (DY1380) were and anti-inflammatory (group III.B) GFs and cytokines for all quantified in CB-PRP derivatives directly and IFN-γ (DY285) three preparations were analyzed and compared, the results are was measured in PBMC stimulated supernatants using Duoset shown in Table 1, Figure 2, we have included statistical analysis ELISA kits (R&D Systems, Abingdon, UK), according to of results both for paired and unpaired samples on the table, but manufacturer’s instructions. the graph shows only significance for the paired samples. For comparative purposes the levels in tears of some of these factors MICA-129 (Valine or Methionine) SNP are as follows: EGF (200–3,000 pg/mL), TGFβ (2–10 ng/mL), Genotyping HGF (200–500 pg/mL), VEGF (2–19 pg/mL), PDGF (90–1,700 Genomic DNA for MICA genotyping was obtained from WCB pg/mL) (3). In CB-PL preparations the concentration of most using QIAamp R DNA blood minikit (Qiagen, GmbH, Hilden, factors tested was higher than that observed in CB-PPP, however Germany). The SNP rs1051792 (G/A) causing substitution of Val this did not reach significance in HGF, MMP-2, TIMP-1, 2, and (G) for Met (A) at position 129 of the MICA gene was genotyped 4 or IL6 and was lower for MMP-9. There were no statistically using a TaqMan R assay (Applied Biosystems, Foster City, CA, significant differences in concentration in any of the analytes USA). PCR amplification conditions were as recommended by examined between CB-PR and CB-PL, except for HGF, MMP- the manufacturer and reactions contained the forward primer 2, and TIMP-2 and 4, which were higher in CB-PL (Table 1, 5′ -GCTCTTCCTCTCCCAAAACCT-3′ and reverse primer 5′ - Figure 2). CGTTCATGGCCAAGGTCTGA-3′ and the two allele-specific As expected, for processed CB-PPP samples, that had 8 ± dye labeled probes FAM-5′ -AATGGACAGTGCCCC-3′ and 3 × 109 PLT/L in 24 ± 4 mL (n = 10), all measured analytes VIC-5′ -AATGGACAATGCCCC-3′ . Results were obtained and had lower concentrations, with the exception for HGF, TIMP- interpreted using the CF96 real-time PCR system (Biorad, 2 and 4 which were no different and MMP-2 and 9, which Hercules, CA, USA). Confirmatory typing was performed using were higher in CB-PPP than in CB-PR (Table 1). Pro-angiogenic DNA extracted from WCB and polymerase chain reaction (PCR) factors showed low concentrations, except for MMP-9. Both amplification and sequencing of exon 3 of the MICA gene. pro and anti-inflammatory cytokines have significantly higher Results were used to determine the presence of “A,” “G” or both concentration in CB-PL than CB-PPP, but are similar between encoding Met or Val at residue 129 (ATG or GTG, respectively). CB-PL and CB-PR (Table 1, Figure 2). In adult serum the Exons 2–6 of the MICA gene were amplified as previously concentrations of these factors are as follows: EGF (500–1,000 described (21) using a forward primer located in intron 1 pg/mL), TGFβ (6–50 ng/mL), HGF (100–1,000 pg/mL), VEGF (5′ - CACCTGTGATTTCCTCTTCCCCAGAGC-3′ ) and reverse (1,000–5,000 pg/mL), PDGF (30–100 ng/mL) (3). While levels of primer in the 3′ untranslated region (5′ -CTAACAATTTGC EGF, VEGF, HGF are comparable between CB-PL, CB-PR, and Frontiers in Immunology | www.frontiersin.org 4 May 2020 | Volume 11 | Article 942 Samarkanova et al. CB-PRP Derivatives TABLE 1 | Comparative composition of CB-PRP derivatives. Paired Unpaired Analyte Sample Concentration p-Value CB-PPP p-Value CB-PPP p-Value CB-PR p-Value CB-PPP p-Value CB-PPP p-Value CB-PR mean ± SEM vs. CB-PL vs. CB-PR vs. CB-PL vs. CB-PL vs. CB-PR vs. CB-PL Group I (trophic and wound healing factors) pg/mL EGF CB-PPP 93.5 ± 5 0.0004 0.0110 >0.9999 0.0002 0.008 >0.9999 CB-PL 858 ± 66 CB-PR 800.6 ± 54 bFGF CB-PPP 179.5 ± 5 0.0004 0.0110 0.9999 0.0001 0.0011 >0.9999 CB-PL 348 ± 26 CB-PR 310 ± 21 HGF CB-PPP 1446 ± 204 0.0553 0.29669 0.0002 >0.9999 0.6702 0.1399 CB-PL 1676 ± 227 CB-PR 1127 ± 116 TGF-β1 CB-PPP 8767 ± 929 0.0002 0.1365 0.1365 0.0003 0.0126 0.8665 CB-PL 58646 ± 3553 CB-PR 49461 ± 3949 Group II (pro-angiogenic factors) pg/mL VEGF CB-PPP 329 ± 17 <0.0001 0.0760 0.0760 <0.0001 0.0038 0.7279 CB-PL 1046 ± 367 CB-PR 798 ± 90 PDGF CB-PPP 3388 ± 535.5 <0.0001 0.1017 0.1017 <0.0001 0.0569 0.0447 AB/BB CB-PL 37267 ± 1817 CB-PR 10276 ± 752 ng/mL MMP-2 CB-PPP 170 ± 8 >0.9999 0.0012 0.0140 >0.9999 0.0905 0.1607 CB-PL 153 ± 9 CB-PR 140 ± 7.7 MMP-9 CB-PPP 101 ± 12 0.0286 0.0005 0.7158 >0.9999 0.8959 >0.9999 CB-PL 88 ± 12 CB-PR 83.4 ± 13 TIMP-1 CB-PPP 157 ± 12 0.0733 0.0005 0.4008 0.0100 0.0003 >0.9999 CB-PL 332 ± 18 CB-PR 372 ± 11 TIMP-2 CB-PPP 48 ± 2.6 0.2969 0.0553 0.0002 0.8156 0.0720 0.0024 CB-PL 48.7 ± 2 CB-PR 38.8 ± 2.5 TIMP-3 CB-PPP 1.1 ± 0.01 0.0029 0.0065 >0.9999 0.0019 0.0033 >0.9999 CB-PL 2.8 ± 0.2 CB-PR 3.0 ± 0.3 TIMP-4 CB-PPP 1.8 ± 0.2 0.2969 0.0553 0.0002 0.8156 0.7047 0.0667 CB-PL 2.0 ± 0.2 CB-PR 1.5 ± 0.2 Group III.A (pro-inflammatory GFs and cytokines) pg/mL IL-6 CB-PPP 77 ± 10 0.0140 >0.9999 0.1017 0.0778 >0.9999 0.4621 CB-PL 98 ± 6 CB-PR 82 ± 5 (Continued) Frontiers in Immunology | www.frontiersin.org 5 May 2020 | Volume 11 | Article 942 Samarkanova et al. CB-PRP Derivatives TABLE 1 | Continued Paired Unpaired Analyte Sample Concentration p-Value CB-PPP p-Value CB-PPP p-Value CB-PR p-Value CB-PPP p-Value CB-PPP p-Value CB-PR mean ± SEM vs. CB-PL vs. CB-PR vs. CB-PL vs. CB-PL vs. CB-PR vs. CB-PL IL-1α CB-PPP 40.5 ± 1.0 <0.0001 0.1017 0.1017 <0.0001 0.0081 0.6702 CB-PL 88 ± 8 CB-PR 69 ± 4 TNF-α CB-PPP 103 ± 1.6 0.0024 0.0024 >0.9999 0.0008 0.0009 >0.9999 CB-PL 123.6 ± 4 CB-PR 122 ± 2 Group III.B (anti-inflammatory cytokines) pg/mL IL-10 CB-PPP 43.5 ± 0.8 0.0004 0.0110 >0.9999 <0.0001 0.0020 >0.9999 CB-PL 72 ± 8 CB-PR 63.8 ± 3 Non-parametric comparison test. n = 10/product type. adult serum, the concentration of PDGF is lower in CB-PR. TGFβ presence of CB-PR improved viability significantly compared is slightly higher in CB-PC derived products and comparable to to CB-PPP and CB-PL again regardless of sample pairing, and adult serum levels in CB-PPP samples. there was no significant difference between CB-PPP and CB- PL in these cells. The viability of CD3+ CD56+ NKT was Effect of CB-PRP Preparations on Donor highest in cells grown in CB-PR but the difference was only Cell Viability slightly significant between CB-PR and CB-PL when samples After measuring the levels of GFs and other factors, we were analyzed paired. For CD3+CD56– T cells, again CB-PR is investigated the effect of the PRP derivatives preparations on significantly better than CB-PL, but no different than CB-PPP immune cells, specifically on NK cells. We and others have whether samples are analyzed paired or unpaired. established that the immunomodulatory effects of plasma leading Overall, incubation of PBMCs in the presence of all three to suppression of NK cells are exerted through both the CB-PRP preparations had a beneficial effect on viability and no high concentration of TGFβ and NKG2D engagement with detrimental or toxic effect. sNKG2DLs. Therefore, we examined the effects of the different preparations on the phenotype and activity of NK, NKT, and T cells. Incubation With CB-PPP, CB-PL, and Initially we examined the effect of incubation of PBMCs in CB-PR Shows Differential Reduction in the presence of the different preparations to ensure they had NKG2D Expression no toxic effects and viability was comparable. We incubated the Generally, incubation of peripheral blood CD3– CD56 dim and cells in standard cell culture conditions, with 50% solution of bright NK cells, CD3+ CD56+ NKT cells and CD3+ CD56– T each PRP preparation, diluted with complete media containing cells with CB-PRP derivatives resulted in reduced expression of IL-2. We then determined the percentage of live lymphocytes NKG2D relative to media only (Figure 4). Significant differences by gating CD3-CD56 bright/dim NK cells, CD3+ CD56+ NKT were observed between the different preparations of CB-PRP cells and CD3+ CD56– T cells cultured with each of the CB- with all cell types, with CB-PL reducing NKG2D expression PRP preparations. The results, shown in Figure 3. show a general the most. While incubation with CB-PPP only reduced NKG2D improvement in viability when cells were cultured with CB-PRP- expression to 94.7% ± 0.25 of maximal expression in CD3– supplemented media compared to media only. For all cell types CD56 dim NK cells, CB-PR reduced it to 78.6% ± 2.05 (p = investigated, viability was significantly improved in the presence 0.0052) and CB-PL to 77.04% ± 2.57 (p = 0.0010). Similar of CB-PR compared to media only. While the presence of CB-PL NKG2D expression patterns were seen in CD3– CD56 bright was only significantly better compared to media alone in CD3– NK cells after incubation with the 3 preparations. Analysis CD56 bright NK cells, CB-PPP was significantly better only in of CD3+ CD56+ NKT cells showed a similar reduction in CD3-DC56 dim NK cells. NKG2D expression after treatment with both CB-PL and CB- The viability of CD3– CD56 bright NK cells was not PR (49.3% ± 2.3% and 54.7% ± 1.5%, respectively) which was significantly different between cells grown in CB-PR or CB-PL significantly more pronounced than culture with CB-PPP (p = when samples were compared either paired or unpaired. Growing 0.0004 and p = 0.0110, respectively in paired analysis). A similar them in the presence of CB-PL or CB-PR increased the viability profile to CD3+ CD56+ NKT cells was observed for CD3+ significantly compared to CB-PPP, again regardless of samples CD56– T cells, though the overall reduction in expression was being paired or unpaired. For CD3– CD56 dim NK cells the less dramatic. Frontiers in Immunology | www.frontiersin.org 6 May 2020 | Volume 11 | Article 942 Samarkanova et al. CB-PRP Derivatives FIGURE 2 | (A) Trophic and wound healing factors in CB-PRP preparations in pg/mL, growth factors analyzed were EGF, bFGF, HGF, and TGF-β1. (B–D) Concentration of the angiogeneic group of factors in CB-PRP derived preparations. Results show concentration in pg/mL and ng/mL of each factor. The analytes measured were (B) VEGF and PDGF in pg/mL; (C) MMP-2 and MMP-9 in ng/mL; (D) TIMP1-4 in ng/mL. (E) Concentration of inflammatory and anti-inflammatory cytokine group of analytes in CB-PRP preparations. Data points represent concentration mean of IL-1α, IL-6, and TNF-α pro-inflammatory and IL-10 anti-inflammatory cytokines. CB-PRP preparations investigated were, cord blood platelet poor plasma (CB-PPP; n = 10), cord blood platelet lysate (CB-PL; n = 10) or cord blood platelet releasate (CB-PR; n = 10). Each sample was measured in duplicate in two separate experiments. Data points represent concentration mean. Statistical analysis was performed using non-parametric one-way ANOVA (Kruskal-Wallis test with Dunn’s post-hoc test for unpaired samples and Friedman’s for paired samples). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Incubation With CB-PL, CB-PPP, and (p = 0.0024). We measured the total IFN-γ produced after CB-PR Reduces Expression of CD107a and stimulation of cultures following the incubation period and results are shown in Figure 6. All CB-PRP preparations resulted IFN-γ in reduced IFN-γ production compared to complete media only, Incubation of all cell types examined with CB-PL resulted not though there were no significant differences between CB-PR only in a significant reduction in the expression of NKG2D, and CB-PL, treatment with both these preparations resulted in but also in a dramatic, and statistically significant, reduction in significantly less IFN-γ production compared to CB-PPP. the expression of CD107a indicating a reduction in activation potential as shown in Figure 5. For example, percentage of maximum expression of CD107a in CD3- CD56bri NK cells CB-PRP Preparations CB-PL, CB-PPP, and incubated with CB-PL was 34.2% ± 3.4 SEM compared to CB-PR Contain sNKG2DLs 72.8% ± 4.01 SEM with CB-PPP (p = 0.002) and 41.4% ± We previously reported that CB plasma contains soluble 3.81 SEM with CB-PR (p = 0.539). Similar differences were NKG2DLs such a sMICA, sMICB, and sULBP1 (12, 22). Using seen in CD3– CD56 dim NK cells and CD3+ CD56+ NKT ELISA, we measured the concentration of these ligands in the cells. Compared with NK and NKT cells, a smaller reduction different preparations of CB-PRP. For this analysis we used two in CD107a expression was seen in CD3+ CD56– T cells and sets of preparations for CB-PL and CB-PR with a total of 20 there was no significant difference between CB-PR and CB- and 19 samples, respectively. One set of CB-PPP was too dilute PL. However, CD107a expression, in the presence of both CB- for the limit of detection of the assay, therefore only 10 samples PL (52.9% ± 4.14) and CB-PR (54.48 ± 3.8), was significantly were utilized. We found that sMICA ligands were present in most reduced compared to incubation with CB-PPP (83.3% ± 2.2) samples of CB-PPP, CB-PL, and CB-PR and all contained sMICB Frontiers in Immunology | www.frontiersin.org 7 May 2020 | Volume 11 | Article 942 Samarkanova et al. CB-PRP Derivatives FIGURE 3 | Viability of CD3- CD56 dim, CD3- CD56 bright NK cells, CD3+ CD56+ NKT cells, and CD3+ CD56- T cells from adult donor PBMCs (n = 3) after incubation with cord blood plasma preparations or complete media only. Results show percentage of live cells (Annexin V negative, 7-AAD negative) for each cell type. Cord blood plasma preparations investigated were cord blood platelet poor plasma (CB-PPP; n = 10), cord blood platelet lysate (CB-PL; n = 10), or cord blood platelet releasate (CB-PR; n = 10). PBMCs were incubated with each preparation diluted 50:50 with media and supplemented with IL-2, or with complete media and IL-2 only for 48 h prior to antibody staining and flow cytometry analysis. Each experiment was repeated with four different PBMC donors and data points represent donor means. Statistical analysis was performed using non-parametric one-way ANOVA (Kruskal-Wallis test with Dunn’s post-hoc test for unpaired samples and Friedman’s for paired samples). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. and sULBP1 (Figure 7A). There was no significant difference of IFNγ is reduced by the presence of the CB-PRP derived in levels of sMICA or sMICB among the different CB-PRP preparations (∼60%), but the reduction is not uniform across preparations but significantly more sULBP1 could be detected in preparations (incubation with CB-PR leads to ∼50% less CB-PL (3.85 ng/ml ± 0.20) compared to CB-PPP (2.62 ng/ml ± INFγ for example) but it is not significantly different within 0.35) (p = 0.0058). preparations for the different genotypes. In agreement with our previous study, we did detect a difference in reduction of IFN- MICA Polymorphisms and Their Effects on γ according to genotype with CB-PPP. Samples typed as V homozygous showed less of a reduction in production of IFN- IFNγ Production γ (mean 52% ± 2.4 SEM) compared to M homozygous and The MICA polymorphism encoding either valine (Val; V) or V/M heterozygous samples (mean 42% ± 1.5 SEM) and was methionine (Met; M) at residue 129 of the MICA protein is statistically significant (p = 0.0074). Overall, 50% of samples thought to influence functional differences in NK cell activation tested were V/V and 50% M/M or M/V genotype. (14). We categorized CB-PRP preparations (CB-PL, CB-PPP, and CB-PR) as either Val homozygous (V/V) or homozygous and heterozygous samples containing Met (M/M and V/M) and DISCUSSION examined the amount of IFN-γ produced by NK cells after PMA and ionomycin stimulation compared to media alone In this study we have assessed and compared the properties of (Figure 7B). As expected per previous results the production 3 different CB-PRP preparations, CB-PPP, CB-PL and CB-PR. Frontiers in Immunology | www.frontiersin.org 8 May 2020 | Volume 11 | Article 942 Samarkanova et al. CB-PRP Derivatives FIGURE 4 | Expression of NKG2D on NK, NKT and T cells from healthy donors is reduced by incubation with, CB-PPP, CB-PL, and CB-PR. Data represents analysis of CD3– CD56 dim, CD3– CD56 bright NK cells, CD3+ CD56+ NKT cells and CD3+ CD56– T cells in adult donor PBMCs (n = 4) after incubation with CB-PRP preparations or complete media only. Results show percentage of maximum expression relative to complete media for each cell type from unstimulated cultures. CB-PRP preparations investigated were cord blood platelet poor plasma (CB-PPP; n = 10), cord blood platelet lysate (CB-PL; n = 10), or cord blood platelet releasate (CB-PR; n = 10). PBMCs were incubated with each preparation diluted 50:50 with media and supplemented with IL-2, or with complete media and IL-2 only for 48 h prior to antibody staining and flow cytometry analysis. Each experiment was repeated with four different PBMC donors and data points represent donor means. Statistical analysis was performed using non-parametric one-way ANOVA (Kruskal-Wallis test with Dunn’s post-hoc test for unpaired samples and Friedman’s for paired samples). *p < 0.05, **p < 0.01, ***p < 0.001. We measured the concentrations of various factors which are Further we looked at the immunomodulatory properties of the thought to contribute to the activity of CB-PRP. We found that 3 preparations, by investigating their effects on the activation of as expected the CB-PPP preparation had lower concentrations NK, NKT, and T cells. We found that all 3 preparations reduced of most analytes, while there was less differences between the expression of activation markers NKG2D and CD107a in all CB-PL and CB-PR. In general, the concentrations of the key cell types investigated to different degrees and further they also analytes were higher in CB-PRP preparations than in tears decreased the expression of INFγ in NK cells. As we used the and comparable to those in adult serum. This is similar to same unit to make all three preparations, and the paired and other reported studies (23, 24), though we measured additional, unpaired analysis give almost identical results, it is most likely the previously not reported analytes. effects are due to the PRP preparation and only some of the effect Frontiers in Immunology | www.frontiersin.org 9 May 2020 | Volume 11 | Article 942 Samarkanova et al. CB-PRP Derivatives FIGURE 5 | Expression of CD107a after PMA and ionomycin stimulation on NK, NKT, and T cells from healthy donors is reduced by incubation with, CB-PPP, CB-PL, and CB-PR. Data represents analysis of CD3– CD56 dim, CD3– CD56 bright NK cells, CD3+ CD56+ NKT cells and CD3+ CD56– T cells in adult donor PBMCs (n = 4) after incubation with CB-PRP derived samples or complete media only. Results show percentage of maximum expression relative to complete media for each cell type following 2 h stimulation with PMA and ionomycin. CB-PRP preparations investigated were cord blood platelet poor plasma (CB-PPP; n = 10), cord blood platelet lysate (CB-PL; n = 10), or cord blood platelet releasate (CB-PR; n = 10). PBMCs were incubated with each preparation diluted 50:50 with media and supplemented with IL-2, or with complete media and IL-2 only for 48 h prior to antibody staining and flow cytometry analysis. Each experiment was repeated with four different PBMC donors and data points represent donor means. Statistical analysis was performed using non-parametric one-way ANOVA (Kruskal-Wallis test with Dunn’s post-hoc test for unpaired samples and Friedman’s for paired samples). *p< 0.05, **p< 0.01, ***p< 0.001, ****p< 0.0001. is due to individual variations between individuals. As we have measure the concentrations of the GFs at different stages reported previously, we believe this reduction is caused by the during the processing and see how the different steps affect the presence of soluble NKG2D ligands (MICA, MICB and ULBP1) final product. in the CB-PRP preparations, but it would be prudent to have The mechanism of action of CB-PRP in terms of a larger sample size to confirm this. Studies in tumor biology immunomodulation has been investigated by us and others do corroborate this idea, as soluble NKG2D ligands produced (12, 22). The anti-inflammatory molecule transforming growth by tumor cells have been reported to downregulate NK cell factor beta (TGFβ), is an obvious candidate to exert such a responses (25, 26). function as it is known to be present in fairly high levels in To understand better the properties of these PRP preparations PRP (27–29). Some studies have looked at the effect of TGFβ and the processing methodology, it would be important to on the induction of T-regulatory (Tregs) cells as a plausible Frontiers in Immunology | www.frontiersin.org 10 May 2020 | Volume 11 | Article 942 Samarkanova et al. CB-PRP Derivatives NKG2D, CD107a, and INFγ in NK cells significantly more than PRP preparations form Val homozygote donors, revealing a functional difference affecting the anti-inflammatory potential of sMICA variants (12). The use of PRP as a treatment for eye (9) and skin lesions (30) has been established for some time. Though initial studies were carried out mainly on autologous PRP preparations, as a means of avoiding transmission of pathogens, the use of allogeneic PRP has been favored in recent years (9, 31). The use of allogeneic PRP has several advantages, and it offers a means of standardizing the preparation and quality control to make dosing regimens more accurate. Amongst the sources of PRP, CB has been studied closely in the last decade or so, as it has all of the advantages of other allogeneic sources, with the extra advantages of being essentially a waste by product which requires very little extra effort to prepare; having special properties and being less likely to have unexpected transmissible viral pathogens, as mothers are checked during pregnancy. Currently we are using CB-PRP preparations on two separate FIGURE 6 | IFN-γ production by PBMCs from healthy donors is reduced by incubation with CB-PPP, CB-PL and CB-PR samples. Results show clinical trials for two indications: an eye drop treatment for percentage of maximum IFN-γ expression in culture supernatants relative to Neurotrophic Keratopathy (NCT03084861) and a gel patch complete media following 2 h stimulation with PMA and ionomycin. CB-PRP for diabetic foot ulcers (NCT02389010). However, there is a preparations investigated were cord blood platelet poor plasma (CB-PPP; n = need to establish the likely differences in preparations and 10), cord blood platelet lysate (CB-PL; n = 10), or cord blood platelet releasate how the biochemical content and functional attributes of each (CB-PR; n = 10). PBMCs were incubated with each preparation diluted 50:50 with media and supplemented with IL-2, or with complete media and IL-2 only preparation may affect their use. We propose that for skin for 48 h prior to antibody staining and flow cytometry analysis. Each ulcers, the most appropriate preparation would be platelet experiment was repeated with four different PBMC donors and data points releasate (PR) in gel form (CB-PG), where the GFs content is represent donor means. Statistical analysis was performed using released gradually (32). This contains the highest concentrations non-parametric one-way ANOVA (Kruskal-Wallis test with Dunn’s post-hoc of both angiogenic as well as trophic and wound healing test for unpaired samples and Friedman’s for paired samples). No significant differences between the different CB-PRP preparations were found. factors and the lowest concentration of inflammatory molecules. *p < 0.05, ***p < 0.001. For skin regeneration several phases are required, first there is a vasoconstriction and platelet aggregation phase to stop bleeding, followed by an inflammatory phase which recruits several cells (mainly neutrophils) into the site followed by mechanism to explain the overall anti-inflammatory properties a proliferative phase where there is new tissue formation of PRP preparations (7). In contrast, in our previous studies we (33). The second phase requires pro-inflammatory molecules were able to demonstrate that CB-PRP preparations contain such as TNFα and INFγ, as well as factors that promote soluble molecules that act as ligands for NK cell receptors collagen degradation, followed by a re-epithelialisation phase and result in reduction of NK cell activation potential (22). where fibroblasts are recruited to the site and with the help These soluble molecules, which have an activating role when of growth factors induce the multiplication of this fibroblast, expressed on the cell surface, interact with NKG2D and directly which in turn deposit new collagen. In chronic wounds there down-modulate cytotoxicity of cells bearing this receptor in a is a clear disruption to this process therefore providing these dose-dependent manner, independent of a similar effect exerted factors exogenously should have a positive effect. Consequently, by TGFβ (12). Further work delineating the differential roles preparations with high levels of TNFα, for example, which of the various types of NKG2D ligands (NKG2DLs) revealed increase inflammation, as well as metalloproteinases and their that soluble ULBP1 (sULBP1) was the most abundant ligand in inhibitors, such as the MMPs and TIMPs, which degrade CB-PRP. Furthermore, we also found that a second NKG2DL collagen, should aid in the induction of the wound healing termed MICA, also supressed NK function, but this ability was cascade (34). High levels of bFGF and EGF consolidate the affected by a genetic polymorphism in the MICA gene itself (12). effect by promoting fibroblast proliferation. The necessary A dimorphism in MICA gives rise to two possible variants of neo vascularisation is also promoted by VEGF, though the the protein with either a methionine (Met) or a valine (Val) at levels of this factor are modest in the CB-PR preparation. residue 129. This difference has been found to have functional Importantly the CB-PR preparation reduces the expression of consequences, with the affinity for the receptor being higher NGK2D activating receptor as well as CD107a in NK cells only on the Met containing protein (14). Indeed, investigations modestly, thus the necessary immune cell activation and IFNγ found that PRP preparations from donors who had at least production is not blocked by this preparation. Previous work one Met allele were able to down regulate the expression of using platelet lysate preparations in in vitro models of wound Frontiers in Immunology | www.frontiersin.org 11 May 2020 | Volume 11 | Article 942 Samarkanova et al. CB-PRP Derivatives FIGURE 7 | CB-PRP preparations (CB-PPP, CB-PL, and CB-PR) contain immunosuppressive soluble NKG2D ligands. (A) ELISA analysis of soluble NKG2D ligand content (sMICA, sMICB, and sULBP1) in CB-PPP (n = 10), CB-PL (n = 20), and CB-PR (n = 19) revealed similar quantities of sMICA and sMICB. Significantly higher concentration of sULBP1 was found in CB-PL compared with CB-PPP. (B) MICA genotyping was performed with DNA obtained from whole CB used to prepare CB-PL (n = 20), CB-PR (n = 19), and CB-PPP (n = 12). MICA-129 valine homozygotes (V/V) were compared with heterozygote and homozygote MICA-129 methionine genotypes (V/M or M/M) for each preparation. Statistical analysis was performed using non-parametric one-way ANOVA (Kruskal-Wallis test with Dunn’s post-hoc test) or Mann Whitney test. **p < 0.01. healing demonstrated that the combination of factors present of lubrication (39), the CB-PPP preparation may offer an in platelet lysates was able to promote wound healing in a optimal topic treatment, though here we analyzed the raw dose dependent manner (35). CB-PC derivatives (lysate and material, the topical eye drops used for clinical application releasate) are rich in GFs, and are attractive sources for many are diluted in Plasmalyte, which may dilute the fibrinogen clinical applications in degenerative conditions, such as joint and and the pro-thrombin present in the CB-PRP avoiding their cartilage pathologies (36, 37), where injection of this product can possible negative effects. So far, studies have not shown any be beneficial. safety issues caused by the presence of fibrinogen which is For the treatment of ocular conditions, preparations need present in autologous PRP preparations too (40, 41), however, to be appropriate to the indication being treated. On the one further studies are necessary to measure this more accurately. hand, ocular ulcers and other conditions that have damaged This preparation contains low levels of pro-angiogenic and the ocular surface where regeneration is required need special wound healing factors, it also reduces the activation of NKT preparations (38). These should contain sufficient quantities of cells by about 50%, which may arrest the progression of EGF, bFGF, and a certain quantity of inflammatory cytokines disease caused by the recruitment of Th17 cells. Previous to activate the regenerative process. However, they must not studies in animal models of severe dry eye disease, showed promote the excessive proliferation of fibroblasts leading to that activated NK and NKT cells up-regulated IL-6 and IL-23 scarring, as it would impair the function of the cornea. The CB- and created an environment where dendritic cells could skew PL preparation has moderately high levels of EGF, bFGF, HGF, T cell activation to a more pathogenic Th17 phenotype (42). PDGF, and VEGF and fairly high levels of TGFβ (compared In the absence of NK and NKT cells, disease progression was to CB-PPP and CB-PR), mostly in line with those found less severe (43), as no pathogenic Th17 cells were activated in natural tears, except for VEGF which is much higher in (42). Additionally, in the presence of CB-PRP derivatives PL. The CB-PL preparation has a dramatic effect on NK, preparations there is a substantial reduction (∼50%) in the NKT and CD3 T-cells. It reduces the expression of both production of INFγ by immune cells, which would otherwise NKG2D and CD107a substantially indicating a reduction on contribute to the inflammatory process. The CB-PPP preparation the activation potential of these cells, thus likely reducing contains the lowest levels of the pro-inflammatory cytokines overall inflammation and disease progression but without IL-6, IL-1α, and TNFα, and substantial amounts of soluble compromising viability. NKG2D ligands (ULBP1, MICA, and MICB), making it ideal For ocular diseases where the corneal surface is not yet to treat conditions where inflammation has a major role in compromised, such as dry-eye disease, which is usually caused by disease progression. a deficiency of the lacrimal functional unit and is accompanied In conclusion, CB has unique properties and can be used as by ocular surface inflammation and irritation due to lack a source of plasma and platelet to formulate preparations for Frontiers in Immunology | www.frontiersin.org 12 May 2020 | Volume 11 | Article 942 Samarkanova et al. CB-PRP Derivatives regenerative medicine, that can be used for the topical treatment collection. This study was approved by the BST local Research of several eye conditions that affect the ocular surface, such as Ethics Committees (HCB/2017/0785). severe dry eye disease, corneal ulcers and burns. Furthermore, these preparations could also serve as an optimal adjuvant to help AUTHOR CONTRIBUTIONS in skin wound healing and for articular regenerative processes. Through differential processing of CB-PRP production, we have DS and SC designed, carried out experiments, and wrote formulated three blood derived preparations which have different the manuscript. DH analyzed results, designed, and and interesting biochemical and immunomodulatory properties wrote manuscript. RC-M design study, analyzed results, which can be investigated in translational applications. Further and wrote manuscript. SQ, LR, and AM designed study and work in evaluating the safety and efficacy of these preparations reviewed the manuscript. must now be undertaken. Additionally, future studies should include: (a) in vivo safety FUNDING and efficacy studies in models of corneal and skin wounds. (b) the assessment of vitamin A concentration on final preparations, This Project was supported by a competitive internal grant as it is known to be light sensitive (44), but an essential ingredient of Banc de Sang i Teixits (Ref I. 2017.036). SC, DH, in eye drop preparations and (c) the effects of the preparations on and AM receive funding from Anthony Nolan (Charity No. the immune response in general. 803716/SCO38827). DATA AVAILABILITY STATEMENT ACKNOWLEDGMENTS All datasets generated for this study are included in the Thanks to Banc de Sang e Teixit and Anthony Nolan Research article/Supplementary Material. Institute laboratories staff. Thanks to Ms Pauline Dodi for her help in initial phases of manuscript. ETHICS STATEMENT SUPPLEMENTARY MATERIAL Plasma samples from individual CB units were obtained from BioBank (BST). The informed consent for the use of samples The Supplementary Material for this article can be found for research was obtained in accordance with the requirements online at: https://www.frontiersin.org/articles/10.3389/fimmu. of the Declaration of Helsinki and local laws at the time of 2020.00942/full#supplementary-material REFERENCES of allogeneic platelet concentrates from umbilical cord blood. Blood Transfus. (2016) 14:73–9. doi: 10.2450/2015.0122-1 1. Parazzi V, Lazzari L, Rebulla P. Platelet gel from cord blood: 9. Giannaccare G, Versura P, Buzzi M, Primavera L, Pellegrini a novel tool for tissue engineering. Platelets. (2010) 21:549– M, Campos EC. Blood derived eye drops for the treatment of 54. doi: 10.3109/09537104.2010.514626 cornea and ocular surface diseases. Transfus Apher Sci. (2017) 2. Parazzi V, Lavazza C, Boldrin V, Montelatici E, Pallotti F, Marconi 56:595–604. doi: 10.1016/j.transci.2017.07.023 M, et al. Extensive characterization of platelet gel releasate from 10. Buzzi M, Versura P, Grigolo B, Cavallo C, Terzi A, Pellegrini M, et al. cord blood in regenerative medicine. Cell Transplant. (2015) Comparison of growth factor and interleukin content of adult peripheral 24:2573–84. doi: 10.3727/096368915X687471 blood and cord blood serum eye drops for cornea and ocular surface diseases. 3. Drew VJ, Tseng C-L, Seghatchian J, Burnouf T. Reflections on dry eye Transfus Apher Sci. (2018) 57:549–55. doi: 10.1016/j.transci.2018.06.001 syndrome treatment: therapeutic role of blood products. Front Med. (2018) 11. Belderbos ME, Levy O, Meyaard L, Bont L. Plasma-mediated immune 5:33. doi: 10.3389/fmed.2018.00033 suppression: a neonatal perspective. Pediatr Allergy Immunol. (2013) 24:102– 4. Sanchez-Avila RM, Merayo-Lloves J, Riestra AC, Fernandez-Vega Cueto 13. doi: 10.1111/pai.12023 L, Anitua E, Begoña L, et al. Treatment of patients with neurotrophic 12. Cox ST, Danby R, Hernandez D, Alejandro Madrigal J, Saudemont A. keratitis stages 2 and 3 with plasma rich in growth factors (PRGF-Endoret) Functional characterisation and analysis of the soluble NKG2D ligand eye-drops. Int Ophthalmol. (2018) 38:1193–204. doi: 10.1007/s10792-01 repertoire detected in umbilical cord blood plasma. Front Immunol. (2018) 7-0582-7 9:1282. doi: 10.3389/fimmu.2018.01282 5. Martinez-Zapata M, Martí-Carvajal A, Solà I, Expósito J, Bolíbar I, Rodríguez 13. Cox S, Madrigal J, Saudemont A. Three novel allelic variants of L, et al. Autologous platelet-rich plasma for treating chronic wounds (Review) the RAET1E/ULBP4 gene in humans. Tissue Antigens. (2012) 80:390– summary of findings for the main comparison. Cochrane Database Syst Rev. 2. doi: 10.1111/j.1399-0039.2012.01933.x (2016) 25:CD006899. doi: 10.1002/14651858.CD006899.pub3 14. Steinle A, Li P, Morris DL, Groh V, Lanier LL, Strong RK, et al. 6. Ralph RA, Doane MG, Dohlman CH. Clinical experience with Interactions of human NKG2D with its ligands MICA, MICB, and homologs a mobile ocular perfusion pump. JAMA Ophthalmol. (1975) of the mouse RAE-1 protein family. Immunogenetics. (2001) 53:279– 93:1039–43. doi: 10.1001/archopht.1975.01010020815015 87. doi: 10.1007/s002510100325 7. Stenwall PA, Bergström M, Seiron P, Sellberg F, Olsson T, Knutson F, et al. 15. Versura P, Profazio V, Buzzi M, Stancari A, Arpinati M, Malavolta N, et al. Improving the anti-inflammatory effect of serum eye drops using allogeneic Efficacy of standardized and quality-controlled cord blood serum eye drop serum permissive for regulatory T cell induction. Acta Ophthalmol. (2015) therapy in the healing of severe corneal epithelial damage in dry eye. Cornea. 93:654–7. doi: 10.1111/aos.12801 (2013) 32:412–8. doi: 10.1097/ICO.0b013e3182580762 8. Rebulla P, Pupella S, Santodirocco M, Greppi N, Villanova I, Buzzi M, et al. 16. Ljubimov A V., Saghizadeh M. Progress in corneal wound healing. Prog Retin Multicentre standardisation of a clinical grade procedure for the preparation Eye Res. (2015) 49:17–45. doi: 10.1016/j.preteyeres.2015.07.002 Frontiers in Immunology | www.frontiersin.org 13 May 2020 | Volume 11 | Article 942 Samarkanova et al. CB-PRP Derivatives 17. Nagaraja S, Chen L, DiPietro LA, Reifman J, Mitrophanov AY. Computational 32. Javier Sánchez-González D, Méndez-Bolaina E, Trejo-Bahena NI. analysis identifies putative prognostic biomarkers of pathological scarring in Platelet-rich plasma peptides: key for regeneration. Int J Pept. (2012) skin wounds. J Transl Med. (2018) 16:32. doi: 10.1186/s12967-018-1406-x 2012:532519. doi: 10.1155/2012/532519 18. Fréchette J-P, Martineau I, Gagnon G. Platelet-rich plasmas: growth 33. Sorg H, Tilkorn DJ, Hager S, Hauser J, Mirastschijski U. Skin wound healing: factor content and roles in wound healing. J Dent Res. (2005) 84:434– an update on the current knowledge and concepts. Eur Surg Res. (2017) 9. doi: 10.1177/154405910508400507 58:81–94. doi: 10.1159/000454919 19. Sirchia G, Rebulla P, Mozzi F, Lecchi L, Lazzari L, Ratti I. A quality system for 34. MacLeod AS, Mansbridge JN. The innate immune system in acute and chronic placental blood banking. Bone Marrow Transplant. (1998) 21(Suppl. 3):S43–7. wounds. Adv Wound Care. (2015) 5:65–78. doi: 10.1089/wound.2014.0608 20. Strandberg G, Sellberg F, Sommar P, Ronaghi M, Lubenow N, Knutson 35. Barsotti MC, Chiara Barsotti M, Losi P, Briganti E, Sanguinetti F, et al. Standardizing the freeze-thaw preparation of growth factors E, Magera A, et al. Effect of platelet lysate on human cells from platelet lysate. Transfusion. (2017) 57:1058–65. doi: 10.1111/trf. involved in different phases of wound healing. PLoS ONE. (2013) 13998 8:e84753. doi: 10.1371/journal.pone.0084753 21. Cox ST, Hernandez D, Danby R, Turner TR, Madrigal JA. Diversity and 36. Li TY, Wu YT, Chen LC, Cheng SN, Pan RY, Chen YC. An exploratory characterisation of polymorphic 3’ untranslated region haplotypes of MICA comparison of single intra-articular injection of platelet-rich plasma vs. and MICB genes. HLA. (2018) 92:392–402. doi: 10.1111/tan.13434 hyaluronic acid in treatment of haemophilic arthropathy of the knee. 22. Cox ST, Laza-Briviesca R, Pearson H, Soria B, Gibson D, Gomez S, et al. Haemophilia. (2019) 25:484–92. doi: 10.1111/hae.13711 Umbilical cord blood plasma contains soluble NKG2D ligands that mediate 37. Laver L, Marom N, Dnyanesh L, Mei-Dan O, Espregueira-Mendes J, Gobbi A. loss of natural killer cell function and cytotoxicity. Eur J Immunol. (2015) PRP for degenerative cartilage disease: a systematic review of clinical studies. 45:2324–34. doi: 10.1002/eji.201444990 Cartilage. (2017) 8:341–64. doi: 10.1177/1947603516670709 23. Baba K, Yamazaki Y, Sone Y, Sugimoto Y, Moriyama K, Sugimoto T, et al. An in 38. Nugent RB, Lee GA. Ophthalmic use of blood-derived products. Surv vitro long-term study of cryopreserved umbilical cord blood-derived platelet- Ophthalmol. (2015) 60:406–34. doi: 10.1016/j.survophthal.2015.03.003 rich plasma containing growth factors—PDGF-BB, TGF-β and VEGF. J 39. Rodríguez-Pomar C, Pintor J, Colligris B, Carracedo G. Therapeutic inhibitors Cranio Maxillofacial Surg. (2019) 47:668–75. doi: 10.1016/j.jcms.2019.01.020 for the treatment of dry eye syndrome. Expert Opin Pharmacother. (2017) 24. Valentini C, Orlando N, Iachininoto M, Pinci P, Nuzzolo E, Teofili 18:1855–65. doi: 10.1080/14656566.2017.1403584 L. Cord blood platelet lysate: in vitro evaluation to support the 40. Lee JH, Kim MJ, Ha SW, Kim HK. Autologous platelet-rich plasma eye drops use in regenerative medicine. Mediterr J Hematol Infect Dis. (2019) in the treatment of recurrent corneal erosions. Kor J Ophthalmol. (2016) 11:e2019021. doi: 10.4084/mjhid.2019.021 30:101–7. doi: 10.3341/kjo.2016.30.2.101 25. Groh V, Wu J, Yee C, Spies T. Tumour-derived soluble MIC ligands 41. Alio JL, Rodriguez AE, Abdelghany AA, Oliveira RF. Autologous platelet- impair expression of NKG2D and T-cell activation. Nature. (2002) 419:734– rich plasma eye drops for the treatment of post-LASIK chronic ocular surface 8. doi: 10.1038/nature01112 syndrome. J Ophthalmol. (2017) 2017:1–6. doi: 10.1155/2017/2457620 26. Cerboni C, Ardolino M, Santoni A, Zingoni A. Detuning CD8 + T 42. Zhang X, Volpe EA, Gandhi NB, Schaumburg CS, Siemasko KF, Pangelinan lymphocytes by down-regulation of the activating receptor NKG2D: Role SB, et al. NK cells promote Th-17 mediated corneal barrier disruption in dry of NKG2D ligands released by activated T cells. Blood. (2009) 113:2955– eye. PLoS ONE. (2012) 7:1–10. doi: 10.1371/journal.pone.0036822 64. doi: 10.1182/blood-2008-06-165944 43. Chen Y, Chauhan SK, Saban DR, Sadrai Z, Okanobo A, Dana R. Interferon-γ- 27. Okuda K, Kawase T, Momose M, Murata M, Saito Y, Suzuki H, secreting NK cells promote induction of dry eye disease. J Leukoc Biol. (2011) et al. Platelet-rich plasma contains high levels of platelet-derived growth 89:965–72. doi: 10.1189/jlb.1110611 factor and transforming growth factor-β and modulates the proliferation 44. Allwood MC. The influence of light on vitamin A of periodontally related cells in vitro. J Periodontol. (2003) 74:849– degradation during administration. Clin Nutr. (1982) 1:63– 57. doi: 10.1902/jop.2003.74.6.849 70. doi: 10.1016/0261-5614(82)90006-1 28. Qian Y, Han Q, Chen W, Song J, Zhao X, Ouyang Y, et al. Platelet-rich plasma derived growth factors contribute to stem cell Conflict of Interest: The authors declare that the research was conducted in the differentiation in musculoskeletal regeneration. Front Chem. (2017) absence of any commercial or financial relationships that could be construed as a 5:1–8. doi: 10.3389/fchem.2017.00089 potential conflict of interest. 29. Eppley BL, Woodell JE, Higgins J. Platelet quantification and growth factor analysis from platelet-rich plasma: implications for wound healing. The reviewer JB declared a past co-authorship with one of the authors SQ Plast Reconstr Surg. (2004) 114:1502–8. doi: 10.1097/01.PRS.0000138251.07 to the handling editor. 040.51 30. Picardi A, Lanti A, Cudillo L, Cerretti R, Dentamaro T, De Angelis G, et al. Copyright © 2020 Samarkanova, Cox, Hernandez, Rodriguez, Casaroli-Marano, Platelet gel for treatment of mucocutaneous lesions related to graft-versus- Madrigal and Querol. This is an open-access article distributed under the terms host disease after allogeneic hematopoietic stem cell transplant. Transfusion. of the Creative Commons Attribution License (CC BY). The use, distribution or (2010) 50:501–6. doi: 10.1111/j.1537-2995.2009.02439.x reproduction in other forums is permitted, provided the original author(s) and the 31. Everts PAM, Knape JTA, Weibrich G, Schönberger JPAM, Hoffmann J, copyright owner(s) are credited and that the original publication in this journal Overdevest EP, et al. Platelet-rich plasma and platelet gel: a review. J Extra is cited, in accordance with accepted academic practice. No use, distribution or Corpor Technol. (2006) 38:174–87. reproduction is permitted which does not comply with these terms. Frontiers in Immunology | www.frontiersin.org 14 May 2020 | Volume 11 | Article 942