MSCohi-O lenses attenuate corneal transplant rejection via HLA-G5 driven Treg cell expansion

Abstract

Background

Corneal transplantation is a critical approach for some vision loss patients. However, the rejection of human organs severely affects the survival rate of the cornea. In addition, oral or intravenous medication against rejection may have side effects on patients, such as systemic immunosuppression. Therefore, a promising strategy for ocular local medication needs to be designed. On the basis of these findings, we constructed a novel platform, the mesenchymal stromal cell-coating high-oxygen permeable hydrogel lenses (MSCohi-O), which combines machinery and cells to achieve continuous drug delivery on the ocular surface to attenuate corneal transplant rejection.

Methods

Multiple in vitro experiments, including T-cell proliferation assay, cytokine secretion assay and relevant gene transcription assay have been performed to investigate the mechanisms of UCMSCs inhibiting inflammatory cells. The in vivo experiments using New Zealand rabbit as the corneal transplantation model have been conducted to verified the mechanism and potential of MSCohi-O rescuing transplanted cornea in recipients.

Results

Our in vitro experimental data investigated the mechanisms by which UCMSCs inhibit T cells and facilitate Treg cells through HLA-G5 modulation. In terms of efficacy, the data from the corneal transplantation model demonstrated that MSCohi-O protected the corneas from inflammation, prevented inflammatory cell infiltration in ocular tissues, inhibited proinflammatory cytokines, promoted Treg cell proportions, and ultimately prolonged the survival period of the transplanted corneas.

Conclusions

These findings addressed the rationale of MSCohi-O rescuing transplanted cornea in recipients, and suggested that MSCohi-O has great potential in managing the rejection of corneal transplantation clinically in the future.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12886-025-04546-2.

Background

Umbilical cord mesenchymal stromal cells (UCMSCs) [1] have emerged as a promising therapeutic strategy for preventing and treating immune-mediated transplant rejection. Their immunomodulatory properties, low immunogenicity, and ability to regulate both innate and adaptive immune responses make them ideal candidates for improving graft survival while reducing reliance on systemic immunosuppressants, which often carry risks of infection and malignancy [2, 3]. Multiple lines of clinical evidence have demonstrated the potential of the use of USMSCs for protection of transplanted organs. A pilot study by Wang et al. (2017) revealed that intravenous infusion of UCMSCs (1 × 10⁶/kg) in liver transplant recipients with acute rejection significantly reduced alanine aminotransferase levels and improved histopathological outcomes without adverse effects [2]. A case report highlighted long-term immune tolerance in a living donor kidney transplant recipient treated with autologous MSCs, with stable renal function (serum creatinine ≤ 1 mg/dL) over 9 years [3]. Moreover, in haploidentical hematopoietic stem cell transplantation (haplo-HSCT), sequential UCMSC infusion (8 doses over 3 months) reduces the incidence of severe acute GVHD from 21.9% to 2.1% and that of chronic GVHD from 45.5% to 27.6%, while maintaining a 62.4% 3-year graft-versus-host disease-free survival rate [4].

However, the application of USMSCs in protecting corneal transplantation from rejection and the determination of the underlying mechanisms are still challenging. Topical UCMSC application to the ocular surface offers a compartmentalized immunomodulatory effect specifically, promoting Treg cell-mediated ocular anti-inflammation while preserving the systemic immune surveillance functions essential for posttransplant recovery. However, a key translational barrier lies in the transient retention of UCMSCs on the corneal epithelium. To address this limitation, we engineered MSCohi-O lenses (mesenchymal stromal cell-coating high-oxygen-permeable hydrogel lenses), a biohybrid platform that integrates UCMSCs with silicone hydrogel carriers for sustained ocular delivery.

Our previous study demonstrated significant immunomodulatory potential in managing corneal transplant rejection through targeted therapeutic strategies. Localized delivery of UCMSCs, such as via MSCohi-O, is increasingly favored to minimize systemic immune activation while enhancing ocular surface retention [5]. Various studies have shown that UCMSCs suppress proinflammatory cytokines (e.g., TNF-α and IL-6) [6] and promote regulatory T (Treg) expansion via soluble factors such as human leukocyte antigen-G5 (HLA-G5) and TGF-β, thereby attenuating acceptor-specific immune responses [7, 8].

Given that UCMSCs inhibit inflammation, the pathway through which HLA-G5 affects Treg proliferation has recently emerged as an important new finding. Human leukocyte antigen-G (HLA-G), a nonclassical HLA class Ib molecule, serves as a pivotal mediator of immune tolerance. This molecule engages inhibitory receptors on T lymphocytes, dendritic cells, and monocytes to suppress both innate and adaptive immunity, establishing peripheral tolerance through direct cellular interactions [9, 10]. Its broad immunoregulatory capacity has positioned HLA-G as a therapeutic target in the fields of transplantation medicine, autoimmune disorders, and reproductive health [11–14]. Among the HLA-G isoforms, HLA-G5 exhibits enhanced immunosuppressive activity due to its unique structural conformation, making it a critical regulator within the HLA-G family [15, 16]. While UCMSCs exert immunomodulation primarily via secreted factors, the precise mechanisms remain incompletely characterized [17, 18]. Recent studies have revealed that MSCs not only express membrane-bound HLA-G but also release soluble HLA-G5, which orchestrates key immunoregulatory pathways in vitro [19–22]. Despite these advances, the role of HLA-G5 in mediating UCMSC-driven attenuation of corneal transplant rejection remains poorly understood.

Given these recent findings, we investigated the feasibility of the use of MSCohi-O for attenuating corneal transplant rejection in a corneal transplantation model of New Zealand rabbits. Our results demonstrated that MSCohi-O promoted the survival of transplanted corneas through HLA-G5-stimulated Treg cell activation. Moreover, MSCohi-O inhibited proinflammatory cytokine expression or secretion and enhanced tissue regeneration. Collectively, these preclinical findings position the use of MSCohi-O lenses as a promising translational strategy for rescuing transplanted corneas from rejection.

Methods

UCMSC cell bank

UCMSC banks were established in compliance with the ICH Q5D, USP, and ISO24603 guidelines. Primary (PCB), master (MCB), and working (WCB) cell banks were sequentially generated through standardized protocols for umbilical cord tissue procurement, MSC isolation, and cryopreservation. Quality control assays confirmed that UCMSCs from all banks met predefined specifications for identity, viability, and functionality [5]. The live UCMSCs were cultured with MesenCult™-ACF Plus Medium (# 05445, STEMCELL Technologies, USA) normally and more detailed information was provided in the supplementary materials.

Trilineage differentiation

The multilineage differentiation capacity of UCMSCs was validated via the use of osteogenic, adipogenic, and chondrogenic induction media per the International Society of California and Technology (ISCT) recommendations [23]. The culture medium of the cells was changed every 72 h. Terminal differentiation was confirmed by histochemical staining: Alizarin Red for mineralization, Oil Red O for lipid droplets, and Alcian blue for proteoglycan deposition.

HLA-G5 secretion by UCMSCs stimulated with interferon-γ (IFN-γ)

UCMSCs were plated in 6-well plates and treated with 20 ng/mL IFN-γ for 24 h. The supernatant of the medium was collected, and ELISA analysis was performed with an ELISA Kit (D711536, Sangon, China) according to the manufacturer’s instructions to quantify the HLA-G5 secretion levels in each group compared with those in the untreated controls.

Inhibition of T-cell secretion of inflammatory factors by UCMSCs

Commercialized primary T cells (Chemicalbook, China) at a density of 1 × 10⁶ were cocultured with UCMSCs (2 × 10⁵) in the presence of phorbol ester (10 µg/mL), brefeldin A (4 mg/mL), or ionomycin (1 mg/mL) to activate cytokine release. After 5 h, T cells were isolated, and IFN-γ and TNF-α transcript levels were measured via qPCR. The control groups lacked UCMSCs.

Inhibition of T-cell proliferation by UCMSCs

Primary T cells were cultured in 97.5% ImmunoCult™-XF T-cell expansion medium (# 10981, STEMCELL Technologies, USA) supplemented with 2.5% ImmunoCult™ Human CD3/CD28 T-cell activator (# 100–0784, STEMCELL Technologies, USA). For the T-cell proliferation assay, primary T cells (1 × 10⁶) were cocultured with UCMSCs (2 × 10⁵) in 2 mL of mixed medium. Phytohemagglutinin (PHA) was added to stimulate proliferation. After 96 h, T-cell counts were compared between the test and control (only T cells) groups [24].

Facilitation of T-cell differentiation into Treg cells by UCMSCs

UCMSCs treated with mitomycin C were cocultured with PHA/IL-2-activated primary T cells (1 × 10⁶/well) for 16–20 h. After centrifugation, fresh medium was added, and the mixture was incubated for 48 h. Treg cell populations were quantified via flow cytometry with general Treg cell markers (CD4 + CD127 + CD25+), and CD25 expression was used as the X-axis indicator for the Treg gating strategy in the UCMSC-exposed group versus the PBMC-only group [25].

MSCohi-O lens preparation

MSCohi-O lenses (U.S. patent pending #18/148,421) integrate nonproliferative, irradiated UCMSCs (1.0–2.0 × 10⁵ cells/lens) into silicone hydrogel carriers. Hydrogel synthesis involves copolymerizing organosilicone monomers with hydrophilic components, followed by thermal curing and lathing. UCMSCs were allowed to adhere to lenses, irradiated (15 Gy), and stored in saline.

New Zealand rabbit corneal transplantation model

Fifteen male New Zealand rabbits were purchased from Biocytogen Pharmaceuticals (Beijing, China) and divided into 3 groups (the untreated control group, the MSCohi-O treatment group and the HLA-G5 knockout MSCohi-O treatment group).

Corneal transplantation procedures in New Zealand rabbits adhered to established protocols [26]. Recipient rabbits underwent alkali-burn pretreatment followed by general anesthesia and topical ocular anesthesia with proparacaine hydrochloride. For the alkali-burn pretreatment procedure, an 8-mm filter paper soaked in 2 mol/L NaOH was applied to the cornea of the rabbit’s left eye for 10 s. The 16 days after the alkali-burn pretreatment, fresh corneal grafts were harvested from doner rabbits via 8.5 mm trephination from the peripheral cornea under aseptic conditions. These grafts were secured onto recipient corneas with 10–0 nylon sutures in a circumferential pattern. Postoperatively, the experimental groups received MSCohi-O lenses, whereas the control groups were fitted with blank lenses. Neovascularization progression was quantified on postoperative days 3, 7, 10, and 14 via the standardized formula. On day 14, corneal tissues were excised for histopathological assessment of inflammatory cell infiltration. Concurrently, qPCR analysis was used to evaluate the CD25, CD127, TNF-α, IFN-γ, IL-1β and IL-10 mRNA levels, whereas ELISA was used to quantify the TNF-α and IL-10 protein concentrations. The scoring system for transplant rejection, inflammation indices and corneal opacity indices referred to previous studies by Cheng et al., 2025, Hu et al., 2015 and Kim et al., 2018, respectively [27–29].

Animal sacrifice

When the experiments were finished, the animals were administered an injectable propofol emulsion via intravenous injection at a dosage of 16 mg/kg body weight to achieve general anesthesia. Following confirmation of a surgical plane of anesthesia (loss of pedal withdrawal reflex and stable respiratory rate), euthanasia was humanely performed through abdominal aorta exsanguination under maintained anesthetic depth.

Statistical analysis

The data were analyzed via GraphPad Prism 9. Student’s t test or one-way or two-way ANOVA was applied as appropriate. The results are expressed as the means ± SDs, with *p* < 0.05 considered significant.

Results

The composition and structure of the MSCohi-O lens with UCMSCs were introduced in our previous study [5]. Here, the biomarkers and functions of the UCMSCs used in this study were newly identified. The biomarkers detected included positive expression of CD44, CD90, CD105 and CD73 (Fig. 1A). Trilineage differentiation assays confirmed the multipotent capacity of UCMSCs. Osteogenic induction triggered calcium-rich matrix deposition, as visualized by Alizarin Red staining of mineralized nodules (Fig. 1B, upper). Adipogenic differentiation was evidenced by intracellular lipid accumulation, with Oil Red O-positive vacuoles appearing in the cytoplasmic area (Fig. 1B, center). Chondrogenic commitment produced proteoglycan-rich extracellular matrices, as indicated by intense Alcian blue staining (Fig. 1B, bottom). The transcript levels of IDO, HGF and PGE2 were significantly increased in UCMSCs after gradient stimulation with the corresponding stimulators (Fig. 1C). When cocultured with primary T cells, UCMSCs successfully suppressed the transcription of TNF-α and IFN-γ from primary T cells (Fig. 1D). Moreover, UCMSCs inhibited the transcription of a proinflammatory T-cell (Th17) differential marker (IL-17 A) but promoted the transcription of anti-inflammatory T-cell (Treg) differential markers (IL-10 and TGF-β) in cocultures with primary T cells (Fig. 1D). Furthermore, ELISA revealed that the secretion of soluble HLA-G (sHLA-G) was increased in UCMSCs cocultured with primary T cells (Fig. 1E).

Fig. 1.

Characteristics of UCMSCs. (A) Determination of biomarkers, including CD44, CD90, CD105 and CD73, in UCMSCs. The expression of CD44, CD90, CD105 and CD73 was used as the X-axis indicator, and the SSC shown on the Y-axis did not have comparative significance, which merely indicates that the samples in the gate were single cells rather than clusters. The negative control used irrelevant IgG as the substitute for the primary antibodies. (B) Trilineage differentiation of UCMSCs detected with Alizarin Red staining for osteoblast differentiation (upper panel), Oil Red O staining for adipocyte differentiation (middle panel) and Alcian blue staining for chondrogenic differentiation (bottom panel). (C) IDO, HGF and PGE2 expression in UCMSCs with gradient stimulation with IFN-γ or IL-1β. (D) Expression of TNF-α, IFN-γ, IL-17 A, IL-10 and TGF-β in primary T cells cocultured with UCMSCs. (E) Soluble HLA-G (including HLA-G5) secretion from UCMSCs cocultured with IFN-γ-pretreated T cells was measured via ELISA. N = 3 for each independent experiment

To investigate whether HLA-G5 is one of the key factors by which UCMSCs modulate Treg cell activation, we knocked out HLA-G5 in UCMSCs and then cocultured the UCMSCs with purified Treg cells from primary T cells. The results revealed that when HLA-G5 knockout UCMSCs were cocultured with Treg cells, the transcription of HLA-G5 was negative, and the level of soluble HLA-G5 generally decreased significantly (Fig. 2A). In Treg cells, the transcription of anti-inflammatory factors, including IL-10, TGF-β and IL-35, was reduced when these cells were cocultured with HLA-G5 knockout UCMSCs (Fig. 2B). Moreover, the effects of UCMSCs on the proportions of T-cell subsets were detected with the CFSE label. When stimulated with PHA, the T cells presented activation and proliferation, as the CFSE signal was weaker than that in the positive control groups (Fig. 2C). Coculture of UCMSCs with T cells suppressed the activation and proliferation of T cells; however, HLA-G5 knockout UCMSCs failed to exhibit such inhibition (Fig. 2C). The proportion of Treg cells among the cocultured primary T cells was detected to further investigate the functions of UCMSCs in promoting Treg cells through HLA-G5. When cocultured with UCMSCs, the proportion of bulk primary T cells enriched with Treg cells increased from 4.01% to 22.2%. However, when cocultured with HLA-G5 knockout UCMSCs, the proportion of Treg cells increased from only 4.01% to 5.96% (Fig. 2D).

Fig. 2.

Functions and mechanisms by which UCMSCs modulate the Treg subtype in primary T cells. (A) Transcription of HLA-G5 and secretion of soluble HLA-G in UCMSCs or HLA-G5 knockout UCMSCs. (B) Changes in IL-10, TGF-β and CD35 transcription in Treg cells cocultured with UCMSCs or HLA-G5 knockout UCMSCs. (C) Percent of primary T cells proliferating, as detected via CFSE labeling, under different coculture conditions, including with UCMSCs, with HLA-G5 knockout UCMSCs, or without UCMSCs. (D) Promotion of Treg subtype proportions in primary T cells by HLA-G5 modulation in UCMSCs, as detected with flow cytometry. N = 3 for each independent experiment

Graft rejection in the corneal transplantation model was quantitatively assessed via a fluorescein staining assay and slit-lamp evaluation of corneal injury, transparency and surface integrity. Compared with control and HLA-G5 knockout corneas, control corneas had faster healing rates (Fig. 3A). In addition, the corneas in the MSCohi-O lens treatment groups recovered to clear on day 10, maintained structural clarity until day 14, and presented smooth epithelial surfaces, whereas those in the untreated controls developed opacity by day 3 with persistent surface irregularities (Fig. 3B). The quantitative data of the corneal injury area (Fig. 3C) and graft survival rate score (Fig. 3D) revealed that MSCohi-O contributed to the restoration and survival of the cornea through the HLA-G5 pathway. Furthermore, quantitative scoring revealed significantly lower indices of rejection severity, stromal inflammation, and corneal opacity in the MSCohi-O group than in the control and HLA-G5 knockout groups at postoperative days 7, 10, and 14 (Fig. 3E).

Fig. 3.

Effects of MSCohi-O lenses on a corneal transplantation model. (A) Images of corneal injury (fluorescein staining) and recovery of corneas in different groups from days 3 to 14. (B) Images of corneal opacity and smoothness in different groups detected by a slit lamp from day 3 to day 14. (C) Quantification of the corneal injury area in the different groups over time. (D) Survival curve of transplanted corneas from day 0 (transplantation) to 14. (E) Clinical scoring of the rejection index (left panel), inflammation index (mid panel) and opacity index (right panel) in corneas. N = 5 for each group of animal experiments

To elucidate the molecular mechanisms underlying the MSCohi-O lens-mediated suppression of inflammation and graft rejection, cytokine profiling and histopathological analyses were conducted. Histopathological evaluation revealed markedly reduced inflammatory cell infiltration in the MSCohi-O group compared with the untreated and HLA-G5 knockout groups (Fig. 4A). Furthermore, the qPCR results indicated that the number of Treg cells (CD25 + and CD127-) was greater in the corneas of the MSCohi-O-treated group than in those of the HLA-G5-KO MSCohi-O-treated group. qPCR analysis revealed significant downregulation of the expression of proinflammatory mediators, including TNF-α, IFN-γ and IL-1β, in the corneas of MSCohi-O-treated mice, which were normalized to those in the untreated controls, compared with those in the heroin-HLA-G5-knockout groups (Fig. 4B). However, the expression of the anti-inflammatory factor IL-10 in MSCohi-O-treated corneas was obviously lower than that in corneas treated with HLA-G5-knockout MSCohi-O (Fig. 4B). ELISA quantification of cytokines corroborated these findings: the level of TNF-α secreted by T cells was decreased, but the level of IL-10 secreted from Treg cells was increased in MSCohi-O-treated corneas compared with those in corneas in the HLA-G5 knockout MSCohi-O treatment groups (Fig. 4C). The flow cytometry results revealed that the proportion of Treg cells in corneas increased from 5.37% to 7.11% in the MSCohi-O-treated groups, whereas the proportion of Treg cells in the MSCohi-O-treated groups treated with HLA-G5-knockout UCMSCs did not significantly increase (Fig. 4D).

Fig. 4.

Mechanisms by which MSCohi-O rescues graft survival in a corneal transplantation model. (A) Detection of infiltrated inflammatory cells (dark blue of nuclear staining) via HE staining of corneal tissues from different groups on day 17. (B) The relevant mRNA (CD25, CD127, TNF-α, IFN-γ, IL-1β, and IL-10) expression levels in each group were normalized to those in the untreated group. (C) The secretion levels of TNF-α (upper panel) and IL-10 (bottom panel) detected in the aqueous humor. (D) Treg proportions in the corneas of different groups, including the MSCohi-O-treated group, the MSCohi-O with HLA-G5-knockout UCMSC-treated group and the untreated group. N = 5 for each group of animal experiments

Discussion

Although UCMSCs have been utilized for managing various inflammation-related or rejection-related diseases [2–4], their potential for rescuing ocular tissue is still a challenge. One of the two major difficulties is that the ocular barriers, including the blood‒retinal barrier, blood‒aqueous humor barrier, and cornea, weaken the efficiency of cytokine or drug transport from the blood to the eyes, which prevents oral or intravenous medication from having the expected effects [30–32]. The other major difficulty is the long period during which drugs, cytokines or cells on the ocular surface are maintained.

To address this challenge, we engineered a biohybrid delivery platform, MSCohi-O lenses, which enables prolonged localized UCMSC administration through mechanical integration. UCMSCs possess multifaceted immunoregulatory capabilities mediated through both paracrine signaling and direct cellular interactions. Central to their therapeutic efficacy is the orchestrated secretion of immunosuppressive cytokines (e.g., TGF-β and IL-10) coupled with the suppression of proinflammatory mediators. Mechanistically, our in vitro experiments demonstrated that HLA-G5 secreted by UCMSCs contributed to the inhibition of T-cell proliferation and simultaneously polarized T-cell differentiation toward regulatory phenotypes, such as Treg cells (Figs. 1 and 2). For the appearance that some inhibition of T-cells in the “T + CFSE + PHA + UCMSCHLA-G5-/-” group in Fig. 2C, we considered that it would be induced by some other undiscovered bypassing pathways. However, from the results, the mechanism of HLA-G5 was still the main pathway in this finding. Our experimental data further revealed that UCMSCs have the capacity to downregulate a series of proinflammatory cytokines produced by T cells (TNF-α and IFN-γ), which can critically exacerbate inflammatory tissue damage. Moreover, UCMSCs also upregulated the transcription of anti-inflammatory cytokines (TGF-β and IL-10) by Treg cells. This triple-axis modulation of (1) Treg polarization in T cells, (2) proinflammatory cytokine inhibition, and (3) anti-inflammatory cytokine promotion collectively establishes an immune-privileged microenvironment conducive to rejection tolerance.

More intriguingly, the obstacle of the blood-eye barrier in this type of therapy has become a tool for preventing unadhered or dead UCMSCs on lenses from escaping the blood from entering the body. Consequently, the anatomical uniqueness of the blood-ocular barrier [33] further enhances treatment specificity. This specialized vascular system selectively restricts the systemic translocation of cells and macromolecules from ocular tissues [34]. This has brought minimal risk of unintended allogenic dissemination.

The results of the corneal transplantation experiments demonstrated the mitigation of corneal transplant rejection through coordinated immunomodulation and tissue preservation. By maintaining corneal structural integrity, these bioengineered lenses create a microenvironment conducive to graft survival, effectively reducing pathological opacity and surface irregularities associated with immune-mediated damage. MSCohi-O successfully suppressed multiple key proinflammatory mediators, including TNF-α, IFN-γ and IL-1β, via the HLA-G5-induced pathway and disrupted critical signaling cascades driving inflammatory cell recruitment and infiltration, thereby attenuating stromal inflammation (Figs. 3 and 4).

The limitations of this study were mainly including the difficulty of controlling UCMSCs properties and the existing of unclear bypassing pathways of HLA-G5 signals. Since the activity of UCMSCs was affected by many conditions like medium, cell density and conditions of cell thawing or seeding on lenses, it was uneasy to ensure that every batch of UCMSCs had consistent qualities. Some previous studies had discussed other known anti-inflammatory and anti-angiogenic factors produced by MSCs that may promote wound healing / graft survival, while the relationships among those factors and HLA-G5 were still unclear. These unresolved questions put more obstacles on the way of ascertaining the role of HLA-G5 with its potentially relevant bypassing pathways in anti-rejection field.

We considered to make the soluble HLA-G5 into the eye drops, since soluble HLA-G5 really had its advantages such as easier to make solutions, more controllable concentration and more stable properties comparing to cells. However, the eye drops could not remain on the ocular surface for long time. Therefore, we are trying to use MSCohi-O lenses to solve the problem of continuously inhibiting corneal transplant rejection at the present stage.

Overall, these findings position MSCohi-O lenses as a paradigm-shifting strategy in transplant ophthalmology, synergizing stem cell biology with biomaterial engineering to achieve compartmentalized immune tolerance. With the rational of discovering that HLA-G5 functions in MSCsohi-O, our future studies may explore combinatorial approaches that integrate HLA-G5-enriched UCMSCs with sustained-release antifibrotic agents to further optimize graft acceptance rates.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

ANOVA

Analysis of Variance

CD3

Cluster of Differentiation 3

CD4

Cluster of Differentiation 4

CD25

Cluster of Differentiation 25

CD35

Cluster of Differentiation 35

CD28

Cluster of Differentiation 28

CD44

Cluster of Differentiation 44

CD73

Cluster of Differentiation 73

CD90

Cluster of Differentiation 90

CD105

Cluster of Differentiation 105

CD127

Cluster of Differentiation 127

CFSE

Carboxy fluorescein Succinimidyl Ester

ELISA

Enzyme Linked Immunosorbent Assay

GVHD

Graft Versus-Host Disease

haplo-HSCT

haploidentical Hematopoietic Stem Cell Transplantation

HGF

Hepatocyte Growth Factor

HLA

Human Leukocyte Antigen

HLA-G

Human Leukocyte Antigen-G

HLA-G5

Human Leukocyte Antigen-G5

HLA-G5-KO

Human Leukocyte Antigen-G5 Knock-Out

ICH

The International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use

IDO

Indoleamine 2,3-dioxygenase

IFN-γ

Interferon-γ

IL-2

Interleukin-2

IL-10

Interleukin-10

IL-17A

Interleukin-17 A

IL-1β

Interleukin-1β

IL-6

Interleukin-6

IL-35

Interleukin-35

ISCT

The International Society of California and Technology

ISO

International Organization for Standardization

MCB

Master Cell Bank

MSCohi-O

Mesenchymal Stromal Cell-coating High-oxygen-permeable Hydrogel

MSCs

Mesenchymal Stromal Cells

PBMC

Peripheral Blood Mononuclear Cell

PCB

Primary Cell Bank

PGE2

Prostaglandin E2

PHA

Phytohemagglutinin

Q5D

Quality of Biotechnological/Biological Products: Derivation and Characterization of Cell Substrates Used for Production of Biotechnological/Biological Products

SDs

Standard Deviations

sHLA-G

Soluble Human Leukocyte Antigen-G

TGF-β

Transforming Growth Factor-β

Th17 cell

Type 17 T helper cell

TNF-α

Tumor Necrosis Factor-α

UCMSCs

Umbilical Cord Mesenchymal Stem Cells

USP

United States Pharmacopeia

WCB

Working Cell Bank

Author contributions

Gene Chu and Haoyu Zeng conceived and designed the project. Yun Li, Yaqi Cheng, Yanchun Lin, Sihua Hong, Yuanyue Liu, Ting Fu and Jiexing Chen performed the experiments and the data analysis. Tian Guan and Shiqi Ling wrote the first draft of the manuscript. Gene Chu and Haoyu Zeng approved the final version to be submitted.

Funding

The work was supported by the Guangzhou Huangpu International Science and Technology Collaboration Project, grant number “2021GH08” (to HZ).

Data availability

The datasets used and/or analyzed in the present study are available from the corresponding author upon request.

Declarations

Ethics and consent to participate

All of the protocols for the animal experiments followed the principles of the Laboratory Animal Welfare Act and were approved by the animal ethics committee of Zhongshan Ophthalmic Center Animal Care and Ethics Committee (approval number: O2023048).

Competing interests

The authors declare no competing interests.