Comparing the Benefits and Drawbacks of Stem Cell Therapy Based on the Cell Origin or Manipulation Process: Addressing Immunogenicity

Abstract

Mesenchymal stem cells (MSCs) are effective in treating autoimmune diseases and managing various conditions, such as engraftment of allogeneic islets. Additionally, autologous and HLA-matched allogeneic MSCs can aid in the engraftment of human allogeneic kidneys with or without low doses of tacrolimus, respectively. However, HLA alloantigens are problematic because cell therapy uses more HLA-mismatched allogeneic cells than autologous for convenience and standardization. In particular, HLA-mismatched MSCs showed increased Ag-specific T/B cells and reduced viability faster than HLA-matched MSCs. In CRISPR/Cas9-based cell therapy, Cas9 induce T cell activation in the recipient's immune system. Interestingly, despite their immunogenicity being limited to the cells with foreign Ags, the accumulation of HLA alloantigen-sensitized T/B cells may lead to allograft rejection, suggesting that alloantigens may have a greater scope of adverse effects than foreign Ags. To avoid alloantigen recognition, the β2-microglobulin knockout (B2MKO) system, eliminating class-I MHC, was able to avoid rejection by alloreactive CD8 T cells compared to controls. Moreover, universal donor cells in which both B2M and Class II MHC transactivator (CIITA) were knocked out was more effective in avoiding immune rejection than single KO. However, B2MKO and CIITA KO system remain to be controlled and validated for adverse effects such as the development of tumorigenicity due to deficient Ag recognition by CD8 T and CD4 T cells, respectively. Overall, better HLA-matching or depletion of HLA alloantigens prior to cell therapy can reduce repetitive transplantation through the long-term survival of allogeneic cell therapy, which may be especially important for patients seeking allogeneic transplantation.

INTRODUCTION

Mesenchymal stem cells (MSCs) can be obtained from a variety of tissues, such as bone marrow, adipose tissue, and umbilical cord blood, and organs, such as, the spleen, thymus, and kidney, with adipose-derived mesenchymal stem cells (ADSCs), bone marrow-MSCs (BM-MSCs), and umbilical cord-MSCs (UC-MSCs) being the most commonly used (1). In addition to their immunomodulatory effects, MSCs are multipotent progenitor cells that can differentiate into a variety of cell types and have been shown to be therapeutic in a variety of diseases (2,3,4,5,6). However, the HLA-ABC alloantigens expressed on allogeneic MSCs are highly immunogenic (7,8,9), which is a common issue in cell therapy utilizing allogeneic cells (10). Mechanistically, in alloantigen recognition, T cells of the recipient immune system can recognize and act on HLA alloantigens through direct or indirect pathways (Supplementary Fig. 1) (11). Ab-mediated rejection (ABMR) by alloreactive B cells is a widely recognized cause of allograft rejection (11,12).

CRISPR/Cas9 is the most widely used method for epigenome editing, owing to its efficiency and convenience (13). Specifically, cell therapies involving MSCs and induced pluripotent stem cells (iPSCs) use CRISPR/Cas9 for genome editing (14). However, it is a system of bacterial origin, which raises concerns regarding its potential immunogenicity (15). The expression of foreign Ags from genome-edited cell therapy can trigger an immune response in the recipient's immune system, which can lead to a decrease in therapeutic efficacy and accumulation of sensitized immune cells (16,17). These foreign Ags can be recognized by T cells via class I and II MHC pathways on Ag presenting cells (18).

This study first examined the various efficacies of MSCs for disease treatment and whether there were differences based on their origin. Second, cell therapy-induced immunogenicity and its causes were investigated in detail. Third, we investigated the benefits and disadvantages of HLA-matched allogeneic cell therapies. Fourth, we discuss the ability of the β2-microglobulin knockout (B2MKO) and universal cell therapies system to evade recipient immune responses to HLA alloantigens. We also review the currently available options for suppressing Ag-specific recipient immune responses. This review of immunogenic-cell therapy will contribute to the identification of the most effective therapeutic approaches while minimizing their adverse effects.

THE UTILITY OF MSCs AS A TREATMENT FOR A VARIETY OF DISEASES

MSCs have therapeutic effects in various diseases, including autoimmune diseases such as systemic lupus erythematosus, severe refractory gingivostomatitis, and oral lichen planus (OLP) (Table 1; reports from 2016 to 2022). Specifically, feline chronic gingivostomatitis (FCGS), an autoimmune disease similar to OLP, is a painful and debilitating inflammatory disease of the oral mucosa that requires lifelong treatment with antibiotics and corticosteroids; however, when MSCs were used to treat FCGS, 4 out of 7 animals (57%) reported clinical improvement (Table 1). HLA-mismatched MSCs (cured by 12–20 months) were also effective in the treatment of FCGS, but less effective than autologous MSCs (cured by 3–9 months) (concentrations of serum IFNγ and neutrophil counts, both p=0.057) (19). In addition, MSCs contributed to improved treatment of ankle non-union in diabetic patients (improving non-unions, p=0.04; reduced infection rate, p<0.01), engraftment of allogeneic islets (improving graft survival, p<0.01), knee osteoarthritis (p<0.05), and radiation-induced xerostomia (increase of 0.06 ml per minute, p=0.0009) (Table 1).

Table 1. Effects of MSCs on the treatment of various diseases.

Authors		Trial types/Object/Period	MSCs (subject)/Dose/Number of doses/Injected site	MHC matching/Serum types	Results
Effects of cell therapy on autoimmune diseases
	Arzi et al. (2016) (20)	• Animal trial (feline)	• ADSCs (n=7), control (n=6)	• Autologous	• Therapeutic effect: 1) reduction of circulating CD8+ T cells 2) normalization of the CD4/CD8 ratio 3) reduction of neutrophils.
		• Treatment of SRG (similar disease to human OLP)	• 5×106 cells/kg	• FBS
		• 6 months	• 2 times
			• IV
	Arzi et al. (2017) (19)	• Animal trial (feline)	• ADSCs (n=7)	• MHC mismatching	• Clinical improvement by ADSCs treatment (4 of 7 [57%]).
		• Treatment of SRG	• 5×106 cells/kg	• FBS
		• 6 months	• 2 times
			• IV
	Kamen et al. (2022) (21)	• Human clinical trial	• UC-MSCs (n=6)	• HLA mismatching	• Allogeneic MSCs is effective for refractory SLE patients (phase I trial).
		• SLE	• 1×106 cells/kg	• HSA	• Effective in reducing Lupus Impact Tracker in SLE patients (52 wk, p=0.007) (22).
		• 12 months	• 1 time		• Autologous MSCs from SLE patients were not effective.
			• IV
Effect of cell therapy in various disease treatment
	Hernigou et al. (2015) (23)	• Human clinical trial	• BMSCs (n=86), control (n=86)	• Autologous	• Therapeutic effect: BMSCs (82.1%), SIBG Control (62.3%) (p=0.04).
		• Ankle non-unions in diabetic patients	• 61,000±18,000	• No use of serum	• Infection rate: BMSCs (1%), SIBG control (20%) (p<0.01).
		• 12 months	• 1 time		• Skin necrosis: BMSCs (1%), SIBG control (11%) (p=0.01).
			• Ankle non-union
	Oliveira et al. (2017) (24)	• Animal trial (mouse)	• ADSCs (n=40)	• MHC mismatching, syngeneic	• Survival of allogeneic ADSCs: 100% on day 7(6 out of 6 mice), 33.3% on day 14, and 0% on day 28.
		• Engraftment of islets	• 2×105 cells	• FBS	• Rejection time of allogeneic islet: Prolonged survival up to 19 days in the islet + allogeneic MSCs group (up to 13 days in the islet alone group).
		• 30 days	• 1 time
			• Subcutaneous injections at the kidney
	Chen et al. (2021) (4)	• Human clinical trial	• ADSCs (n=17), control (n=8)	• HLA mismatching	• Reduction of pain scores by ADSCs treatment (wk 12, p=0.0026).
		• Knee osteoarthritis	• 16×106 cells	• Unknown	• Improvement of KSCRS score (objective knee indicator, symptoms, and functional activities) (wk 48, p=0.0234).
		• 96 wk	• 1 time
			• IAI
	Lynggaard et al. (2022) (3)	• Human clinical trial	• ADSCs (n=10), control (n=10)	• HLA mismatching	• Improvement of UWS flow rate by ADSCs treatment: increase of 0.06 ml/min (p=0.0009).
		• Radiation-induced xerostomia	• 5×107 cells	• Human platelet lysate	• Improvement of XQ score: 22.6 units reduced (p=0.0004).
		• 4 months	• 1 time		• ADSCs (Off-the-Shelf)
			• Submandibular and parotid gland

HSA, human serum albumin; IAI, intra-articular injection; KSCRS, Knee Society Clinical Rating System; UWS, unstimulated whole saliva; XQ, xerostomia questionnaire.

MSCs IN ENGRAFTMENT OF HUMAN KIDNEY TRANSPLANTS: AUTOLOGOUS & HLA-MATCHED ALLOGENEIC CELLS

In human organ transplants, calcineurin inhibitors (CNI; cyclosporine [1980s], tacrolimus [1990s]) provide better protection against acute rejection (25). However, the long-term use of these inhibitors increases the risk of infection and malignancy, which is an obstacle to long-term survival; therefore, MSCs treatment has been utilized as an alternative to CNI (Table 2; reports from 2016 to 2022). Human MSCs can reduce or replace the use of CNI in the engraftment of HLA-matched allogeneic kidneys and suppress rejection for a long time. Specifically, autologous BM-MSCs were able to prevent rejection after the tacrolimus withdrawal, and HLA-matched allogeneic BM-MSCs were able to prevent rejection with a low dose of tacrolimus for two years (Table 2).

Table 2. Effect of autologous- or HLA-matched allogeneic-MSCs on engraftment of human kidney transplantation.

Authors	Object/Period	MSCs (subject)/Dose/Number of doses/Injected site	HLA matching/Serum types/Injection point	Results
Pan et al. (2016) (26)	• Improvement of kidney transplant	• BMSCs (n=32)	• Matched allogeneic (donor-derived MSCs)	• The combination of low-dose tacrolimus and MSCs was as effective as standard dose tacrolimus (at after 2 years): urea, urine protein, urinary RBC, urinary WBC, 24-h urine protein, and creatinine clearance rates.
	• 24 months	• 5×106 cells (1st), 2× 106 cells/kg (2nd)	• No use of serum
		• 2 times	• Unknown
		• 1st: kidney artery, 2nd: IV
Dreyer et al. (2020) (27)	• Improvement of kidney transplant	• BMSCs (n=10)	• Matched allogeneic	• BMSCs injection prevented kidney rejection even with the use of low-dose tacrolimus (at 12 months).
	• 12 months	• 1.5×106/kg	• FBS
		• 2 times	• 6 months after KTp
		• IV
Meucci et al. (2021) (5)	• Improvement of kidney transplant	• BMSCs (n=27), control (n=27)	• Autologous	• BMSCs replaces tacrolimus without kidney rejection
	• 24 wk	• 1–2×106/kg	• Unknown	• BMSCs ameliorates cardiovascular complications by discontinuation of tacrolimus (at 24 wk).
		• 2 times	• 6–7 wk after
		• IV
Reinders et al. (2021) (28)	• Improvement of kidney transplant	• BMSCs (n=29), control (n=28)	• Autologous	• BMSCs replaced tacrolimus without kidney rejection and was safe.
	• 24 wk	• 1.5×106/kg	• No use of serum	• BMSCs increased the number of regulatory T cells compared to the control group (at 24 wk, p=0.014).
		• 2 times	• 6–7 wk after KTp
		• IV
Večerić-Haler et al. (2022) (29)	• Treatment of Ab-mediated kidney graft rejection	• BMSCs (n=3)	• Autologous	• Even standard therapy combined with autologous BMSCs did not improve AMR in KTRs.
	• 12 months	• 3×106 cells/kg	• Human serum
		• 3 times	• 5 wk after KTp
		• IV

AMR, antibody-mediated rejection; KTR, kidney transplant recipient.

STEM CELL THERAPY IMMUNOGENICITY

Causes of immunogenicity: HLA-alloantigens, xenogeneic molecules, and foreign Ags from genome editing system

Cell therapy is promising for the treatment of various diseases; however, immunogenicity remains challenging (Table 3; reports from 2015 to 2023). Thus, understanding the causes and impact of immunogenicity is essential for the safe use of cell therapy. Kol et al. (30) reported that HLA-mismatched MSCs induce an immune response in alloreactive Abs or cytotoxic CD8 T cells. These results are supported by reports that HLA-ABC alloantigens expressed by MSCs induce alloreactive CD8 T cell activity primarily through a direct pathway (Supplementary Fig. 1) (8). Interestingly, rejection of HLA-mismatched MSCs can be analyzed both in vivo and in vitro approximately 3 weeks after injection, compared to mixed lymphocyte responses in donors and recipients, which are detectable at approximately 7 days (8,31,34,41). These results correlate with MSCs having immunomodulatory activity and relatively low expression of HLA alloantigens, which may be a hallmark of MSCs, leading to a later onset of rejection (11). However, T cells specific for alloantigens on allogeneic MSCs do not respond to immunosuppressants (36,38). Večerić-Haler et al. (29) also reported that autologous MSCs used for immunomodulation were ineffective as inhibitors of chronic ABMR, suggesting that, even if MSCs have immunomodulatory effects, they are limited in suppressing Ag-specific rejection (Tables 1–4) (8). These results are also supported by reports that HLA-mismatched allogeneic cell therapy results in a lower survival rate than autologous cell therapy (Table 3). In iPSC cell therapy, it has been reported that HLA-matched iPSC-neurons have a higher survival rate than mismatched cells in the brain, a tissue known to have a low immune response, proving that the same applies to allogeneic cell therapy (17,48).

Table 3. Cell therapy immunogenicity.

Authors		Trial types/Object/Period	MSC types/Dose/Number of doses/Injected site	MHC matching/Serum types	Results
Immunogenic cell therapy
	Kol et al. (2015) (30)	• Animal trial (horse)	• ADSCs/BMSCs	• Mismatched allogeneic	• No organ toxicity or systemic inflammatory response.
		• Safety and lymphocyte response tests	• 25×106 cells	• No use of serum	• Repeated injections of allogeneic MSCs may result in a cytotoxic response with an increase of circulating CD8+ T cells.
		• 35 days	• 3 times
			• IV
	Owens et al. (2016) (31)	• Animal trial (horse)	• ADSCs/BMSCs	• Mismatched allogeneic	• Anti-MSCs Abs were detected in 7 out of 19 horses (37% of the study horses).
		• Detection of alloreactive Ab	• 2.5–8×107 cells	• FBS	• Abs to MSCs may develop 3–4 wk after MSCs injection.
		• 600 days	• 4 times
			• 4 sites including IV
	Kawamura et al. (2016) (32)	• Animal trial (monkey)	• iPSC-CM (GFP)	• Matched/mismatched allogeneic	• Matched iPSC-CM grafts had better survival compared to mismatched those (p<0.05).
		• Immunogenicity evaluation for MHC matched iPSC-derived CM	• Aggregation or sheet type: 3.3×106 cells (3 sheets/animal)	• Both autologous and xeno-free serum (33)
		• 2 months	• 1 time
			• Anterior wall of the left ventricle, Backs of recipient animals (for sheets)
	Morizane et al. (2017) (17)	• Animal trial (monkey)	• iPSC-neurons	• Matched/mismatched allogeneic	• Rate of allogeneic iPSC-neuron induced inflammation: MHC-matching < MHC-mismatching (at 3 months, p=0.005).
		• Immunogenicity for MHC matched/mismatched iPSC-derived neurons in brain	• 4.8×106 cells (8×105 cells/tract, 6 tracts)	• Both autologous and xeno-free serum (33)	• More iPSC-neurons survived in MHC-matched grafts compared to MHC-mismatched grafts (at 4 months, p=0.004).
		• 4 months	• 1 time
			• One side of the putamen
	Chang et al. (2020) (34)	• Ex vivo (human)	• ADSCs	• Mismatched allogeneic	• Allogeneic ADSCs induce alloreactive CD8 T cell-mediated cytotoxicity (8).
		• Immunogenicity evaluation for human MSCs	• 1×103 cells/well	• Autologous serum (comparison with FBS)	• Allogeneic ADSCs cause significant production of memory-CD8 T cells in allogeneic-Ag stimulation (34).
		• 3 wk	• 2 times
			• Allogeneic-Ag stimulation
Immunogenicity for xenogeneic molecular
	Joswig et al. (2017) (35)	• Animal trial (horse)	• BMSCs	• Autologous, mismatched allogeneic	• Autologous BMSCs exposed to xeno-serum (FBS) induced a significant adverse response in joints and increased synovial total nucleated cell counts (p=0.0007).
		• Osteoarthritis therapy	• 10×106	• Both FBS and autologous serum
		• 36 days	• 2 times
			• Intra-articular
	Wagner et al. (2019) (36)	• Ex vivo (human)	• Using Cas9 whole protein	• Human PBMCs	• CRISPR/Cas9 for gene-editing induces activity of preexisting reactive CD4/CD8 T cells.
		• Evaluation of T cells reactive to Cas9	• 1×107 PBMCs	• Human serum	• Reactive T cells responding to virus-related gene therapy are not well regulated by immunosuppressive agents such as anti-CTLA4 and low dose prednisone (37).
		• 5 days (response with PBMC and Cas9 protein)	• 1 time		• Cas9-specific reactive Treg are effective in regulating these reactive T cells.
			• Proliferation assay for Cas9 protein
	Chang and Park (2019) (8)	• Ex vivo (human)	• ADSCs	• Mismatched allogeneic	• ADSCs grown in xenogeneic medium cause faster T cell-mediated cytotoxicity through a direct pathway.
		• Evaluation of alloreactive memory T cells	• 1×103 ADSCs/well	• Autologous serum (comparison with FBS)
		• 3 wk	• 2 times
			• Allogeneic-Ag stimulation
	Deuse et al. (2019) (38)	• Ex vivo (human)	• iPSC/iEC	• Autologous	• Autologous iPSCs and their derivatives also induce immune activation.
		• Mechanism of autologous iPSC rejection	• 4×105 iEC	• FCS	• Gene reprogramming has up to 9 times higher mutation rates than conventional culture conditions (39,40).
		• 90 h (response with effector cells and target cells)	• 1 time		• Mutations in the mitochondrial DNA of iPSCs cause the production of neoepitopes.
			• In vitro T cell-mediated rejection

Table 4. Comparative analysis of immunogenicity for autologous-versus allogeneic-cell therapy.

Authors		Trial types/Object/Period	MSC types/Dose/Number of doses/Injected site	MHC matching/Serum types	Results
Comparison of autologous and MHC-mismatched allogeneic MSCs
	Hare et al. (2012) (42)	• Human clinical trial	• BMSCs	• Autologous, mismatched allogeneic	• Both allogeneic and autologous cells showed their safety by reducing infarct size (p<0.001) in patients with ICM.
		• Functional safety & adverse effects	• 2×107 (1st), 10×107 (2nd), 20×107 (3rd)	• Human serum albumin (Pa, 0), FBS (Pa, 1)	• More than 30% of patients tested showed sensitization to HLA Ags (8 of 27) at baseline.
		• 13 months	• 3 times		• A majority of the sensitized patients (7 of 8 [87.5%]) showed sensitization at all time points with alloreactive Ab.
			• Myocardium
	Pigott et al. (2013) (43)	• Animal trial (horse)	• BMSCs	• Autologous, mismatched allogeneic, and xenogeneic	• The number of perivascular cellular cuffs and vascular microthrombi was increased in all MSC-injected joints (p=0.036).
		• Immune response to MSCs from various sources	• 15×106 cells	• FBS	• No presence of any MSCs was confirmed in the synovium at day 60.
		• 120 days	• 1 time
			• IAI
	Isakova et al. (2014) (44)	• Animal trial (monkey)	• BMSCs	• Autologous, mismatched allogeneic	-Primary injection
		• Comparison of autologous and allogeneic MSCs	• 2.5×106 cells (1st), 1×106 cells (2nd)	• Unknown		• A significant increase of circulating leukocytes, neutrophils and eosinophils was observed only at 10 or 30 days in the allogeneic group (p<0.05).
		• 7 months	• 2 times			• Levels of circulating CD3+ T cells and NK were found to be significantly elevated at 60 days in the allogeneic group (p<0.05, p=0.0005, respectively).
			• 1st: ICI, 2nd: SCi			• B cells were always significantly increased in the allogeneic group (p<0.05).
					-Secondary injection
						• The number of lymphocytes showed a significant increase at 7 days after the second transplant in the allogeneic group (p<0.05).
	Arzi et al. (2017) (19)	• Animal trial (feline)	• ADSCs (n=7)	• HLA mismatching	• Allogeneic ADSCs have been shown to have lower clinical efficacy compared to autologous ADSCs.
		• Treatment of SRG	• 5×106 cells/kg	• FBS
		• 6 months	• 2 times
			• IV
	Oliveira et al. (2017) (24)	• Animal trial (mouse)	• ADSCs	• Syngeneic, mismatched allogeneic	• Survival rates of allogeneic GFP+ ADSCs: day 7 (100%, 6 out of 6 mice), day 14 (33.3%), day 28 (0%)
		• Islet transplantation	• 2×105 cells	• FBS	• Rejection time of allogeneic islet: 19 days in islet plus allogeneic MSC group (13 days in islet alone group).
		• 30 days	• 1 time
			• SCi at the kidney
	Colbath et al. (2020) (45)	• Animal trial (horse)	• BMSCs	• Autologous, mismatched allogeneic	• Lymphocytes increased significantly only in the allogeneic group (at 24 h after injection, p=0.04).
		• Comparison of autologous and allogeneic MSCs	• 1×107 cells	• FBS
		• 2 wk	• 1 time
			• IAI
	Hwang et al. (2020) (46)	• Animal trial (mouse)	• ADSCs	• Syngeneic, allogeneic, and xenogeneic	• Infiltration of CD45+ Leukocytes: xenogeneic (43.1%), allogeneic (23.3%), and syngeneic (2.8%) (p<0.001 vs. xenogeneic).
		• Comparison of immune responses to MSCs of various origins	• 2×105 cells	• FBS	• Infiltration of CD8 T cells: allogeneic (5.1%), xenogeneic (0.09%), and syngeneic (0.02%) (p<0.001, allogeneic vs. xenogeneic).
		• 7 days	• 1 time
			• Left caudate putamen injection
Immunogenicity for the treatment of repeated allogeneic cell therapy
	Joswig et al. (2017) (35)	• Animal trial (horse)	• BMSCs	• Autologous, mismatched allogeneic	• Even when using autologous serum, repeated IAI injections of allogeneic MSCs induce an alloimmune response (p=0.0009).
		• Clinical response to repeated injection of allogeneic versus autologous MSCs	• 10×106	• Both FBS and autologous serum
		• 4 wk	• 2 times
			• Intra-articular injection
	Rowland et al. (2021) (47)	• Animal trial (horse)	• BMSCs	• Autologous, matched/mismatched allogeneic	• The repeated injection increased peri-articular edema and synovial effusion in the mismatched group (all p<0.05; matched vs. mismatched).
		• Effects of repeated injection	• 10×106	• FBS	• The mismatched group consistently induced an increase of IFNγ in the joint (p=0.01; matched vs. mismatched).
		• 36 days	• 2 times		• Mismatched group are susceptible to donor-specific anti-MHC Abs and complement-mediated cytotoxicity.
			• IAI

Pa, passage; ICM, ischemic cardiomyopathy; ICI, intracranial injection; SCi, subcutaneous injection.

Other causes of immunogenicity include foreign Ags and xenogeneic molecules used in the genetic manipulation and culture of cell therapy, respectively. First, the Cas9 protein from CRISPR/Cas9, which is used to generate iPSCs, can induce the activity of CD8 T cells as well as Ag-specific CD4 T cells (36). This is further supported by recent reports that plasmacytoid dendritic cells (DCs), as well as conventional DCs, which are known to be potent cross-presenting dendritic cells, can cross-present extracellular Ags to Class I MHC (18). However, it is encouraging that Wagner et al. (36) have shown that Ag-specific Treg cells are effective in suppressing the activity of Cas9-specific T cells. Second, fetal bovine serum, commonly used for the culture and proliferation of MSCs, promotes cytotoxic CD8 T cell activation and kills HLA-mismatched MSCs more rapidly than human serum (8,35). These results correlate with reports that albumin can bind to a variety of tissues and cells and be absorbed intracellularly (49,50). This means that fetal bovine serum must be replaced with human serum or xenofree material before use (8,51).

Comparison of immunogenicity in autologous– versus HLA-mismatched allogeneic–cell therapy

Data comparing the immunogenicity of allogeneic and autologous cells in MSCs therapy have been examined (Table 4; reports from 2012 to 2023). The results showed that HLA-mismatched MSCs induced an increase in B cells and the production of alloreactive Abs compared to autologous MSCs (42,44). In addition, HLA-mismatched MSCs induce significant T lymphocyte infiltration compared to autologous MSCs (44,45,46). Ultimately, the immunogenicity of HLA-mismatched MSCs resulted in a significant decrease in the survival rate compared to that of autologous MSCs (8,24). Interestingly, even immunogenic HLA-mismatched MSCs were effective in prolonging the survival of allogeneic islets without serious adverse effects (Table 1) (19,24). These results suggest that HLA-mismatched MSCs may partially contribute their immunomodulatory effects, even if they trigger Ag-specific immune responses in the recipient immune system. However, immunogenic-cell therapy may eventually be eliminated by rejection by the recipient’s immune system, which may cause routine repeated administration and result in adverse effects (Tables 3 and 4).

BENEFITS OR DEFECTS OF HLA-MATCHED ALLOGENEIC-CELL THERAPY

Autologous-derived cell therapies may be the most effective choice from an immunological safety and functional standpoint. However, they have the disadvantage of not being immediately available. In particular, patient-derived autologous cell therapy may not be functionally effective (21). Thus, HLA-matched allogeneic-cell therapies may be an alternative to autologous cell therapies. Table 5 compares the advantages and disadvantages of HLA-matched cell therapy based on Tables 1–4. Most importantly, HLA-matched allogeneic cells can lower the induction of alloreactive memory T/B cells in recipients compared to mismatched allogeneic cells (47). In addition, the long-term survival of HLA-matched allogeneic cells reduces the number of repeated cell treatments (24,47,52,53,54). Although HLA-matched allogeneic cells have disadvantages in terms of convenience, they have the potential to be used effectively once national/international cell banking systems are well established, as their immunogenicity can be minimized (Fig. 1) (55).

Table 5. Benefits and defects of HLA-matched allogeneic cell therapy.

Contents	Benefits of HLA-matching	Descripts
Convenience	Medium	• Mismatched allogeneic cells > matched allogeneic cells.
Major causes of immunogenicity in cell therapy	Y (minimization of defect)	• Allogeneic cells: HLA-alloantigens.
		• Genome-edited cells: foreign Ags produced by genome-editing system.
		• Xenogeneic Ags: FBS.
Production of alloreactive memory T cells	Y (minimization of defect)	• Mismatched allogeneic cells > matched allogeneic cells.
Prospects for repeat allogeneic cell therapy	Y (minimization of defect)	• Faster removal of allogeneic cells as the number of treatments increases (Tables 3 and 4).
		• Major causes of increased allograft rejection: HLA-mismatching and immune cells pre-sensitized to an alloantigens.
Verified cell therapy	Y	• Matched or mismatched allogeneic cells > autologous cells.
		• Cell banking is effective for the use of matched allogeneic cells (Fig. 2).
Long-term survival	Y	• Autologous cells > matched allogeneic cells > mismatched allogeneic cells.
		• Effective in reducing the number of repeat treatments: autologous and matched allogeneic cells.

Y, yes.

Figure 1. Overview of effective HLA matching in preparing allogeneic cell therapy. (a-c) Cultivation of allogeneic cell therapy (replacing FBS with human serum) and manipulation using genome editing system. (d) For better HLA-matched cell therapy, national/international cell banking systems may be effective (17,55). (e) Preparation of autologous cell therapy. (f) Treat patients with ready-made cell therapy. This figure was created using BioRender.com.

DIFFERENCES IN THE IMMUNOGENICITY OF ALLOGENEIC– AND GENOME-EDITED–THERAPEUTIC CELLS

Both allogeneic and genome-edited cell therapies cause immunogenicity in patients upon transplantation (Tables 3 and 4); however, there are differences in the consequences of their immunogenicity. Table 6 presents the differences in immunogenicity according to the origin of the cell therapy and genetic manipulation based on Tables 3–5. Specifically, HLA-matched therapeutic cells with edited genomes are primarily associated with cytotoxic responses and the accumulation of memory cells in response to foreign Ags rather than the rejection of alloantigens (14,36,38,53,56,57). In other words, the effect of immunogenicity due to foreign Ags is limited to the cell populations associated with these Ags. In contrast, cell therapy using allogeneic cells sensitizes memory T/B cells to HLA-alloantigens, leading to their accumulation (36,58,59). Thus, immunogenicity problems caused by HLA alloantigens may be associated with a much broader range of adverse effects than those caused by foreign Ags.

Table 6. Comparison of immunogenicity differences between allogeneic- and genome edited-therapeutic cells.

Contents on immunogenicity	GE cell therapy using autologous cells	Cell therapy using allogeneic cells
Types of Ags	• Foreign-Ags induced by CRPSPR/Cas9 and inserted gene/vector.	• HLA-alloantigens.
Types of Ag-specific T cells	• Develop both CD4 T and CD8 T cells to Cas9 Ag via cross-presentation.	• Development of alloreactive CD8 T cells to predominantly against mismatched HLA-A,B,C.
	• Cas9 also induces the production of Tregs (36).
Impact of treatment of repeated cell therapy	• Immune rejection is faster for identical genome-edited cells, but HLA Ags are not the targeted.	• Faster rejection dependent on number of cell therapy: HLA-mismatched cells > HLA-matched cells.
		• The use of HLA-matched allogeneic cells is effective in reducing the number of treatments via long-term survival.
Comparison of immunogenicity levels	• GE-mismatched-allogeneic cell > GE-matched-allogeneic cell > GE-autologous cell.	• Mismatched allogeneic cells > matched allogeneic cells > autologous cells.

HOW TO AVOID HLA-ALLOANTIGEN SENSITIZATION 1: B2MKO CELL THERAPY

Recently, attempts have been made to eliminate class I MHC as a major cause of allogeneic rejection. B2MKO iPSCs/iMSCs lacking class I HLA presentation were able to evade rejection by alloreactive CD8 T cells compared to wild-type controls (60). However, cells lacking class I HLA may have been targeted by NK cells (61,62), thus increasing their susceptibility to NK cell lysis (p<0.001) (60). NK cell lysis in B2MKO cell therapy could be suppressed by knock-in of HLA-E or HLA-G at the B2M locus, which would not induce allogeneic responses (62,63). However, B2MKO cell therapy may be fundamentally vulnerable because it is difficult for the recipient's immune system to function in the event of tumorigenesis or exposure to infection. Defects in B2M gene expression have been reported as the basic mechanism of drug resistance in patients with lung cancer (64,65). Thus, B2MKO cell therapy may be considered for the application of a safety system, such as a suicide system introduced into CAR-T cells, to control the occurrence of serious adverse effects (66). Three types of suicide systems have been reported:1) HSV-1 thymidine kinase suicide genes with immunogenicity (67,68), 2) human origin of the iCaspase 9 suicide system (69,70), and 3) uridine monophosphate synthetase-knockout cell line system for auxotrophy to uridine; suicide switches without transduction may avoid immunogenicity issues (71).

HOW TO AVOID HLA-ALLOANTIGEN SENSITIZATION 2: UNIVERSAL DONOR CELL THERAPY

CIITA is the master transcription factor for class II MHC genes, and it has been reported that ablation of class II MHC can alleviate rejection by CD4 T cells (72). Universal donor cells have been proposed that remove both class I and II MHC, the greatest source of immunogenicity in allogeneic transplantation, and include MSCs (73), iPSCs (74,75), and CAR-T (66,76). Wang et al. (74) demonstrated in a monkey xenogeneic model that human universal iPSCs with dual KO of B2M and CIITA can further reduce infiltration of T/B lymphocytes than single KO. Dexamethasone was also able to increase the expression of prostaglandin E-2 (PGE-2), indoleamine-2,3-dioxygenase (IDO), and HLA-G in MSCs, which is interesting because it may contribute to improved function of the cells as well as reduced NK lysis susceptibility (77). However, safeguards such as suicide systems may also be required to control these cells, which are outside the host’s immune system.

DISCUSSION

Because of their immunomodulatory and progenitor cell capabilities, MSCs are attractive therapeutic agents for treating various diseases (Tables 1 and 2). Notably, in human kidney transplantation, autologous MSCs replaced tacrolimus to reduce the adverse effects of CNI, and HLA-matched allogeneic MSCs suppressed rejection at lower concentrations of tacrolimus (Table 2). However, HLA-mismatched cell therapy, used for convenience and standardization, causes rejection owing to the activation of alloantigen-specific T/B cells preexisting in the recipient’s immune system (Tables 3 and 4). Fortunately, there are few reports of recipient rejection of these cell therapies directly causing serious adverse effects (78,79). However, these results may lead to the repeated use of cell therapy from allogeneic sources without a deeper understanding of the trade-offs between adverse effects and therapeutic benefits. A second repeat transplantation of allogeneic MSCs into horses or monkeys has been associated with the development of adverse clinical effects thought to be due to adaptive immunity compared to controls (35,44,47). These results are supported by reports showing that HLA-mismatched allogeneic cells cause an increase in alloreactive memory CD8 T cells, an increase in Ag-specific Abs, decreased survival, and faster immune rejection (44) (Tables 3 and 4). Thus, these results suggest that reducing alloantigen-sensitized T/B cells through an HLA-matching design is the most important factor in improving engraftment and reducing the repeated use of any cell therapy that uses allogeneic cells (Fig. 1) (14,35,53).

The removal of class I MHC from the surface of cell therapy is being attempted as another way to avoid rejection due to the use of HLA-mismatched cell therapy. The neoantigen peptides bind to a class-I MHC molecule consisting of an α chain and a β chain of B2M, which is presented on the cell surface and recognized by the TCR of CD8 T cells (Supplementary Fig. 1) (65). Thus, B2MKO cell therapy was able to function while avoiding rejection by alloreactive T cells, compared to wild-type controls (60). However, since the B2MKO system avoids Ag recognition by CD8 T cells, it may be difficult to deal with the development of adverse effects, such as tumorigenicity or infection of genome-edited cells (64,65). Thus, this problem may be solved by applying a suicide system to prevent the adverse effects caused by the excessive immune response of CAR-T cells (66,67,69,70,71).

Allogeneic CAR-T cells are the most successful cell therapies approved by the FDA, but they are not free from rejection due to immunogenicity (66). However, universal–CAR-T, –MSCs, and –iPSC cell therapies in which both B2M and CIITA are knocked out have been shown to effectively evade immune rejection (66,73,74,75,76). Even more encouraging is that a number of universal cell therapies are being evaluated as functionally normal or well differentiated (66,74,75,76,80). These results show that universal cell therapy has the potential to be an effective treatment that overcomes the disadvantages of autologous or HLA-matched cells. Additionally, these results may provide hope for MSCs, which have failed to achieve efficacy in several clinical trials due to tissue source, donor heterogeneity, and heterogeneous manufacturing (81). However, the control of these cells outside the host’s immune system may require safeguards such as artificial suicide systems. In addition, HLA-KO cell therapies are expected to suppress acute rejection, but immunogenicity to foreign Ags, which is expected to be relatively weak, remains (17,32,36,48).

CONCLUSION

The use of autologous or HLA-matched allogeneic cells in cell therapy may have disadvantages in terms of convenience but may reduce repeated use due to the long-term survival of therapeutic cells. In addition, the use of these cell therapies is expected to increase curative opportunities and success rates in patients who are candidates for allogeneic transplantation (24,47,52,53,54). These results are attributed to the fact that although HLA-mismatched allogeneic cells such as MSCs and iPSCs have immunomodulatory effects, they are ineffective in suppressing the activity of Ag-specific memory cells in the recipient's immune system (Tables 3 and 4) (29). To prevent the rejection of immunogenic-cell therapy, the introduction of Ag-specific Tregs effectively suppresses the recipient's immune response to Cas9 and alloantigens (36,82,83). However, to produce Ag-specific Tregs, there are obstacles, such as high-level technology that has not yet been established, cost, and time consumption (84). Thus, B2MKO- and universal-cell therapies, with proven safety, as well as validated HLA-matched cells, has the potential to be an effective way to reduce immunogenicity. However, the number of times a patient should receive a validated HLA-matched-, B2MKO-, or universal-cell therapy compared with autologous treatment to minimize adverse effects remains to be evaluated. In addition, the use of xenogeneic-free serum benefits the long-term survival of therapeutic cells by reducing unnecessary activity of the recipient’s immune system (8,35,51). This suggests that immunogenic cell therapy may have an opportunity cost in terms of the limited number of treatments between adverse effects and therapeutic benefit, so efforts to eliminate the causes of immunogenicity need to be sustained.

Abbreviations

ABMR

antibody-mediated rejection

ADSC

adipose-derived mesenchymal stem cell

B2MKO

β2-microglobulin knockout

BM-MSC

bone marrow-MSC

CIITA

class II MHC transactivator

CNI

calcineurin inhibitors

DC

dendritic cell

FCGS

feline chronic gingivostomatitis

GE

genome edited

GFP

green fluorescent protein

IAI

intra-articular injection

iEC

differentiation of iPSCs into endothelial cell

iPSC

induced pluripotent stem cell

IV

Intravenous

KTp

kidney transplantation

MSC

mesenchymal stem cell

OLP

oral lichen planus

SIBG

standard iliac bone graft

SLE

systemic lupus erythematosus

SRG

severe refractory gingivostomatitis

UC-MSC

umbilical cord-MSC

SUPPLEMENTARY MATERIAL

Supplementary Figure 1

Recognition pathway of recipient T cells against HLA-alloantigen. (A) A direct pathway is the recognition of HLA-alloantigens presented on the surface of donor cells by recipient T cells. This pathway is primarily associated with acute allograft rejection (S1,S2). (B) In the indirect pathway, HLA-alloantigens from donor cells are recognized by recipient T cells by presentation of recipient APCs. This pathway is involved in chronic allograft rejection (S1,S2). This figure was created using BioRender.com.