Author contributions: A.S.B.: conception and design, wrote manuscript, conceived and prepared figures, and final approval of manuscript; N.P.R.: conception and design, wrote manuscript, and final approval of manuscript; K.O.L.: prepared figures and reviewed manuscript; X.F.: contributed to writing manuscript; Y.X.: conception and design, contributed to writing manuscript and final approval of manuscript. A.S.B. and N.P.R. contributed equally to this article.
Disclosure of potential conflicts of interest is found at the end of this article.
First published online in STEM CELLSEXPRESS March 14, 2012.
Autologous-induced pluripotent stem cells (iPSCs) may eventually be used in cell replacement therapies to treat a wide range of diseases and have been touted as a solution to the vexing problem of immune rejection in this context. Emerging evidence suggests, however, that ostensibly histocompatible iPSCs may be rejected following transplantation. Here, we review the mechanisms that contribute to immunogenicity in iPSCs and forward approaches to permit their acceptance in potential cell replacement therapies. STEM CELLS2012;30:797–803
In 2006, the breakthrough to generate induced pluripotent stem cells (iPSCs), somatic cells reprogrammed with defined factors to a pluripotent embryonic stem cell (ESC)-like state, fuelled hope that iPSCs could act as renewable tissue sources for transplantation to treat a plethora of debilitating, degenerative diseases . Indeed, proof of principle studies in mouse models of sickle cell anemia and liver degeneration has shown the potential clinical application of iPSCs to treat human diseases .
THE UNPREDICTED IMMUNOGENICITY OF SYNGENEIC IPSCS
The crucial, perceived advantage of using patient-derived (autologous) iPSCs in cell replacement therapy (CRT) has been their capacity to bypass immune rejection, a dogma that was recently challenged in a study showing rejection of syngeneic mouse iPSCs following transplantation . This study took advantage of the capability of C57BL6 (B6) iPSCs to form teratomas in B6 mice after transplantation; teratomas, which are surrogate tissues for transplantation, contain cells from all three embryonic germ layers (mesoderm, ectoderm, and endoderm), permitting simultaneous evaluation of the immunogenicity of various differentiated cell types. Using the teratoma assay, it was found that immune destruction was mediated by activating T-cell-dependent immune responses. Interestingly, teratomas formed by transplanting ESCs derived from inbred B6 mice are immune tolerated after transplantation in B6 recipients, suggesting that syngeneic iPSCs are more immunogenic than their ESC counterparts in this context.
This study raises intriguing questions about the cause of immunogenicity in transplanted syngeneic iPSCs. It is unclear, for example, how the developmental stage of the cell or the cell type used to induce pluripotency impacts the immunogenicity of iPSCs. Certainly some cell types (e.g., cord blood) used in reprogramming are innately less immunogenic than others (e.g., adult skin) and perhaps this directly influences the intrinsic immunogenicity of derived iPSC lines through epigenetic mechanisms [4, 5]. The reprogramming method itself also appears to determine iPSC immunogenicity; iPSCs derived via the episomal approach are less prone to immune-mediated attack than those generated using viral vectors . This data suggest that refining reprogramming methodology may mitigate immune destruction of iPSCs.
While some immune rejection was demonstrated in the context of transplanting undifferentiated iPSCs , it is entirely possible that this immunogenicity could increase during differentiation to specific tissues, as has been observed during directed differentiation of ESCs [6, 7]. This issue is particularly worthy of further investigation as only fully differentiated tissues from iPSCs will be used in a clinical setting. To assess the potential impact of immune rejection on iPSCs in clinical transplantation, it will also be necessary to extend these mouse iPSC studies to assess the immunogenicity of fully differentiated cells derived from human iPSCs in humanized mouse models.
GENETIC AND EPIGENETIC ABNORMALITIES AS CONTRIBUTORS TO IPSC IMMUNOGENICITY
Gene profiling studies have shown that ESCs and iPSCs have very similar global gene expression patterns [1, 8–10]. However, recent studies found aberrant methylation patterns and an epigenetic memory of their tissue of origin in iPSC lines [5, 11–13]. Thus, it is relevant to ask whether genetic and epigenetic factors could influence the immunogenicity of iPSCs. While aberrant reprogramming of DNA methylation may indeed present an immunological problem, it should be noted that the phenomenon of epigenetic memory, in certain circumstances, might be advantageous in the generation of tissues from iPSC that are relatively intractable to standard differentiation protocols  or in the context of deriving an iPSC line from a less immunogenic cell source of origin. Epigenetic abnormalities, however, can be extinguished if iPSCs are repeatedly passaged , leaving an epigenetic blank slate from which multiple-differentiation pathways are equally possible and with the potential for reduced immunogenicity if iPSCs are made from cell sources with relatively more potent starting immunogenicity. Yet in the setting of ESCs, duplications of stretches of DNA and chromosomal abnormalities occur after multiple passages in vitro  and, if this phenomenon was observed in iPSCs, it could compound their immunogenicity. In this regard, recent studies have identified reprogramming-associated coding sequence mutations and chromosomal abnormalities in newly generated iPSCs . Perhaps most strikingly, abnormal gene expression was observed in teratoma cells, which were differentiated in vivo from syngeneic iPSCs and found to directly contribute to the immunogenicity of iPSCs .
So how can genetic and epigenetic abnormalities in iPSCs, which lead to enhanced immunogenicity, be neutralized? In this specific context, the choice of cell types used for reprogramming is probably not the answer because the genetic and epigenetic abnormalities are detected in iPSCs reprogrammed from various cell types including keratinocytes, neural stem cells, and fibroblasts [16–18]. While the addition of epigenetic-modifying compounds during reprogramming might eliminate the epigenetic memory and abnormalities in iPSCs, the issue with reprogramming-associated genetic abnormalities may prove harder to resolve. Recent studies have indicated that critical tumor suppressors p53 and Arf repress induced pluripotency [19–23]. Therefore, successful reprogramming may require transient inactivation of these tumor suppressors. Because p53 acts as chief guardian of the genome and is required to maintain genetic stability, transient inactivation of p53 during reprogramming could lead to genomic instability . To resolve this issue, it would be important to understand how p53 suppresses induced pluripotency in order to modify the reprogramming strategy to avoid inadvertent manipulation of the p53 pathway.
PATIENT-DERIVED iPSC-DERIVED TISSUE FOR CRT: WILL THEY BE TRULY AUTOLOGOUS?
Clinical CRT using patient-specific iPSCs will be contingent on transplantation of either (a) iPSCs that have been differentiated to specific cell types or (b) iPSCs that have been rectified for specific gene defects via gene therapy, followed by differentiation to the desired cell type. In both scenarios, substantial manipulation of patient iPSCs is required. This begs the pivotal question of whether specific tissues differentiated from iPSCs and/or disease reversed iPSCs will remain immunologically identical to the undifferentiated iPSC line from which they were derived.
Expression of key transplantation antigens, major histocompatibility complex (MHC) molecules in mouse or human leukocyte antigens (HLAs) in human, has been identified as one of the principal mechanisms facilitating increased immunogenicity during ESC differentiation to different tissues [7, 25] and may be pertinent to the immunogenicity of the differentiated progeny of iPSCs. For example, mouse ESC-derived insulin producing cell clusters (IPCCs) demonstrated higher MHC class I expression than their undifferentiated ESC counterparts . Furthermore, stimulation in vitro with the inflammatory cytokine interferon-gamma (IFN-γ) caused IPCCs to induce MHC class I faster than undifferentiated ESC. In stark contrast to undifferentiated ESC, terminally differentiated IPCCs also expressed MHC class II after challenge with IFN-γ. In vitro differentiation of iPSCs could independently generate expression of novel minor histocompatibility antigens' (mH) differences that may be sufficient to trigger rejection of an otherwise syngeneic or autologous iPSC graft . Overall, these data predict that undifferentiated patient-derived iPSCs may have the potential to become more immunogenic as they mature during differentiation in vitro to specific cell types. The rejection of transplanted syngeneic iPSCs differentiated in vivo to teratomas lends further weight to this hypothesis .
Extending the notion of increased immunogenicity during iPSC differentiation to the context of using disease reversed iPSCs for CRT makes for a potentially more complicated situation. The facility to rectify diseases via genetic manipulation of patient-derived iPSCs, as typified by elegant proof of principle studies in mouse [2, 27], is clearly a tantalizing prospect. However, these studies, as elegant as they are, have yet to directly assess the possibility that the corrected gene or gene products derived from iPSCs may be viewed as a foreign or “non-self” entity, since it was absent during T-cell selection and education in the thymus. Thus, further investigation is warranted into whether tissue derived from gene-corrected iPSCs may be rejected by the host immune system.
In both these autologous settings, the possibility of developing iPSC-derived tumors has been deemed to be greater than for ESCs since they should lack alloimmunity to target any contaminating undifferentiated cells within the graft. However, it is likely that the host adaptive immune response will in fact be able to target those donor cells in an iPSC-derived graft with the propensity to give rise to tumors; T-cell-mediated destruction of undifferentiated iPSC grafts and the recent isolation, from healthy donors, of memory T-cell clones directed against Oct-4 support such a contention [3, 28]. The risk posed by residual expression of embryonic antigens, or maintenance of endogenous reprogramming factor expression within an iPSC-derived tissue graft, could also be negated by using a recently identified approach to remove contaminating cells by cellular selection .
In addition to the immunological considerations outlined, regulatory issues associated with using human products in the clinic and the potentially exorbitant expenses related to generating and painstakingly testing individual cell lines will prevent patient-specific therapy for many. Adding the time it takes to differentiate a patient's own cells into specific tissues will make it difficult to apply patient-specific iPSC therapy, especially to the wealth of acute diseases that could be treated by CRT. Perhaps, therefore, the most suitable option for using iPSCs in transplantation would be to turn to cell banking, a strategy that has also been proposed for ESC-based therapy and is currently in operation for bone marrow and cord blood transplantation to treat hematological malignancies . With the diversity of starting cells available to make iPSCs and the a priori ability to derive them from less immunogenic starting cells, one could envisage that it is considerably easier to generate an immunological match compared with ESCs. Even so, the best possible immunological match from an iPSC bank will most likely be unable to escape immune recognition following transplantation.
THE POTENTIAL FOR HOST IMMUNE SYSTEM-MEDIATED DESTRUCTION OF iPSC-DERIVED TISSUE
The immune system itself has evolved to protect the body against a vast array of pathogens but also plays a significant role in transplantation responses, where the “pathogen” is a non-self transplanted tissue graft. The very act of transplanting tissue to a patient will activate the host or recipient immune response. As suggested by the Zhao et al.  study, the function of transplanting fully differentiated tissues from both autologous and allogeneic iPSCs may therefore be undermined by the host immune response following transplantation.
INNATE IMMUNE RESPONSES TO TRANSPLANTATION OF iPSC-DERIVED TISSUE
The innate immune system functions as a first-line defense against infection and plays a crucial role in inducing the adaptive immune responses by presenting antigen to T cells. In both syngeneic and allogeneic transplantation, the innate immune cell response, comprising effector cells including monocytes, macrophages, neutrophils, and natural killer (NK) cells, is responsible for early inflammatory damage to grafts .
Innate immunity participates in mediating damage to ESC-derived tissue after transplantation [31, 32] and may have a similar role in damaging and impacting the functionality of iPSC-derived tissue. NK cells, for example, may contribute to immunogenicity of iPSC tissue derivatives. Although increasing MHC expression on ESCs and iPSC tissue derivatives may flag grafts for rejection by the adaptive immune response in the allogeneic setting, its expression may be protective to NK cell-mediated attack in both syngeneic and allogeneic transplantation . Moreover, lack of or mismatch of class I MHC expression on ESC-derived grafts has been shown to cause recognition by NK cells . The involvement of NK cells in iPSC graft damage is likely to be dependent on the fine balance between MHC expression on iPSCs and their differentiated progeny.
Neutrophils and macrophages may also contribute to iPSC-derived tissue damage in the initial stages following transplantation. Early after implantation of syngeneic and allogeneic ESC-derived IPCCs, ingress of neutrophils and macrophages to the graft site was noted in association with expression of inflammatory cytokines and chemokines . Given that the innate cell-mediated inflammatory response normally subsides with time in the context of syngeneic transplantation, what will the ramifications be for such damage inflicted to transplanted syngeneic iPSC-derived tissue?
The ultimate destruction of syngeneic iPSCs by T-cell-mediated mechanisms makes a legitimate case against the involvement of innate immunity in the process of immune rejection of syngeneic iPSCs . However, lessons learnt from transplantation of ESC-derived tissue caution against drawing such a simplistic conclusion. Intriguingly, for example, ESC-derived IPCCs were found to be rejected relatively early following transplantation into syngeneic recipients . Knowing that innate cell immunity causes an early inflammatory response that can form the basis for activation of adaptive immune responses , it is tempting to speculate that syngeneic iPSC graft damage caused by innate cell infiltration and inflammatory responses may be vital to their ultimate rejection after transplantation. Thus, limiting innate immune cell-mediated damage in the incipient stages following transplantation of syngeneic iPSC-derived tissues may provide a means to stimulate long-term iPSC graft survival. Further investigations are required to ascertain whether innate and adaptive immunity work together to potentiate rejection of transplanted syngeneic iPSC-derived tissue. As higher infiltration of neutrophils and macrophages has been observed later in ESC-derived IPCCs transplanted in the allogeneic setting [31, 32], it will also be highly relevant to assess the impact and level of cross-talk between the innate and adaptive immune systems for transplantation of allogeneic iPSC-derived tissue.
ADAPTIVE IMMUNE RESPONSES OF TRANSPLANTED iPSC-DERIVED TISSUE
Adaptive immune responses arise through recognition of specific antigen via one of the three allorecognition pathways, leading to stimulation and subsequent activation of recipient T cells by either recipient or donor antigen-presenting cells (APCs) (Fig. 1). There has been much research to determine whether allogeneic ESCs and tissues derived from them will be subjected to immune rejection via such an adaptive immune response. Early studies reported that undifferentiated ESCs expressed low levels of MHC molecules  and that they were unable to stimulate T cells in vitro  or in vivo . However, as alluded to earlier, MHC expression was found to increase as ESC differentiate to specific tissues or when these tissues were exposed to inflammatory signals [7, 25]; the ultimate outcome in the allogeneic setting, and in the absence of immunosuppression or tolerance induction, was graft rejection mediated by CD4 and/or CD8 T cells [32, 36–38]. In assessing transplantation of allogeneic IPCCs differentiated from ESCs, it was memory, rather than naïve, CD8+ T cells that mediate graft rejection  (Fig. 1). These data suggest, perhaps unsurprisingly, that the progeny of allogeneic iPSCs, like those of syngeneic iPSCs, will be subjected to similar mechanisms of graft rejection.
IMMUNOLOGICAL APPROACHES TO OBVIATE REJECTION OF iPSC-DERIVED TISSUES IN CRT
The preceding considerations suggest that if the promise of patient-specific CRT using iPSCs could be fulfilled, their ultimate success clinically would be reliant on sufficient intervention to ameliorate damage or rejection that occurs to transplanted tissue. Importantly, it is also likely to be the case where partial or full immunological matching is achieved from banked, allogeneic iPSC lines. So what options are available to modulate immunity in response to autologous and allogeneic iPSC-based transplantation?
Switching off a transplant recipient's immune system using conventional long-term immunosuppression, with the attendant risk of infection and malignancy, is the most extreme method used to modulate immunity in patients receiving HLA-mismatched organ transplants [39, 40]. Reducing such toxic immunosuppressive regimens is the most imposing clinical objective to attend to in transplantation, which will ultimately have implications to the utility of iPSC-derived tissues in regenerative medicine.
A successful alternative strategy deployed to modulate immunity to organ allograft transplantation in animal models specifically targets T-cell activation by interfering with costimulatory molecules (CSMs), either on the T cell themselves or on the APCs that present cognate antigen . Indeed, blocking these pathways using cytotoxic T-lymphocyte-associated antigen 4-Ig and anti-CD40 ligand (anti-CD40L) antibodies, in conjunction with an antibody against the adhesion molecule lymphocyte function-associated antigen-1, can enable tolerance induction to human ESCs transplanted in the immune privileged testis, and moderately protect cells transplanted into the hearts of immunocompetent mice . This same costimulatory blockade protocol approach also promotes the survival of allografts and xenografts derived from ESCs and iPSCs . In light of Zhao et al.'s findings , it will be important to investigate whether syngeneic iPSC-derived cells can also be rescued from rejection using this protocol. Costimulatory blockade could therefore have utility in clinical transplantation of iPSC-derived tissues, although clinical translation of these protocols in organ transplantation has so far proved troublesome .
FACILITATING iPSC GRAFT SURVIVAL THROUGH INDUCTION OF IMMUNOLOGICAL TOLERANCE
Under normal conditions, the immune system does not mount a response against its own “self” tissues due to a state called self-tolerance controlled by central and peripheral mechanisms (Fig. 2). A long sought after goal in transplant immunology has been to exploit the body's mechanisms for inducing this state to allow an individual to effectively recognize an unrelated donor transplanted graft as its own self tissue . This approach would circumvent the requirement for immunosuppression and enable long-term survival of iPSC-derived tissues.
Harnessing central tolerance mechanisms to re-educate T cells to allow acceptance of donor iPSC grafts would include, among other strategies [44, 45], reprogramming the recipient immune system by induction of hematopoietic chimerism where the recipient's bone marrow is ablated to allow engraftment of donor-specific hematopoietic stem cells [14, 46]. Owing to the toxic conditioning regimens required to induce this state, it could be argued that hematopoietic chimerism is an extreme choice for all but the most severe illnesses that could be treated using iPSC-derived tissue.
We will focus instead on the potential methods to enforce immune tolerance in the periphery for the acceptance of iPSC-derived tissue in CRT . In this regard, regulatory T cells (Treg) capable of suppressing or “regulating” the activation of alloreactive (recognizing non-self) lymphocytes are crucial to the maintenance of peripheral tolerance and have established ability to regulate transplant graft acceptance , including that of ESC-derived tissue [38, 41]. In this manner, the aim of potentiating Treg activity to permit iPSC-derived graft acceptance clinically while also preserving normal T-cell-mediated immunity may be achieved. Using small number of polyclonal or non-antigen-specific Tregs can induce and expand a cohort of antigen-specific Tregs in vivo, leading to a state of tolerance known as “infectious tolerance” [48, 49]. However, derivation of donor antigen-specific Tregs may be a more potent strategy in the setting of transplantation using allogeneic iPSC-derived tissues where there is an obvious need to depress immune rejection. In transplantation of tissues from autologous and/or disease-corrected iPSCs, where more subtle antigenic differences could exist between the “altered-self” cells and the patient they were derived from (e.g., mH antigens), the generation of donor antigen-specific Tregs would also be beneficial to engender acceptance of iPSC-derived grafts.
The ability to induce Treg activity in vitro from normal, nonregulatory T cells by cytokine stimulation to produce so called “inducible or adaptive” Tregs (also known as Tr1 cells) is one potential method to use Tregs in CRT using iPSC-derived tissue. In principle, Tregs can either be induced de novo or isolated from the patient, expanded ex vivo and reinfused before transplantation of iPSC-derived tissue (Fig. 2). In the future, it may also be possible for Tregs to be generated from the same iPSC line that would be used for making tissues for CRT (Fig. 2). To this end, T cells can be generated from ESCs by coculturing them on an OP9 stromal cell layer engineered to express the Notch ligand delta-like 1, which is crucial for proper T-cell selection . When such iPSC-derived T cells could then be manipulated in vitro to become Tregs, they could be infused prior to transplantation of tissues derived from iPSCs for CRT and, in an antigen-specific manner, regulate graft acceptance.
We concede that making T cells in vitro from iPSCs is a laborious task and inducing Treg activity in vitro thereafter a daunting prospect. There are, however, several other indirect approaches that could be deployed to induce and expand “naturally occurring” Treg activity in vivo following transplantation of iPSC-derived tissues. Of potential relevance, here are immunoregulatory cells such as tolerogenic immature dendritic cells (DCs) and myeloid suppressor cells, both of which can be generated from ESC and/or iPSCs [51–53]. Immature DCs are tolerogenic by virtue of their lack of the requisite machinery for T-cell activation, expressing only low levels of class II MHC and CSMs and producing the suppressive cytokines like interleukin (IL)-10. Through these mechanisms, immature DCs could promote graft acceptance of iPSC tissue by leaving a potentially alloreactive T cell in an inactive (anergic) state or, as alluded to above, expanding naturally occurring Treg activity in vivo (Fig. 2). Similarly, recently identified myeloid-derived suppressor cells, which are innate immune cells with a capacity to augment Treg activity, may be exploited in tolerance induction strategies toward iPSC-derived tissue in CRT .
Technology to reprogram somatic cells to an ESC-like state producing iPSCs could have a wide-ranging impact on the nascent field of regenerative medicine. The ability to make iPSCs from individuals afflicted with degenerative diseases has implications in and of itself for modeling disease pathogenesis in vitro and approaches to drug discovery to treat these conditions. Rectifying selected gene defects in iPSCs and directing the differentiation of iPSCs to generate myriad tissues on demand for CRT would be ultimate additions to the clinical armamentarium. Yet there are several issues to resolve if the lofty objective of taking iPSCs from the laboratory to the clinic is to be achieved. Among these, iPSC lines, irrespective of the cell type from which they were derived, should be efficiently and completely differentiated to particular tissues. The known genetic and epigenetic abnormalities of iPSCs must also be resolved as this could lead to adverse clinical effects in transplantation, such as the risk of developing malignancy. Added to these complex issues, emerging evidence suggests that even histocompatible iPSC lines are immunogenic after transplantation. It is therefore of paramount importance to comprehensively assess how the host immune system impacts transplanted syngeneic, autologous, and allogeneic iPSC-derived tissue. In this respect, gauging the immunogenicity of fully differentiated tissue from iPSCs will be of interest. Likewise, it will be necessary to assess whether invoking immune-modulatory approaches will be the key to successful, sustained functionality of iPSC-derived grafts in regenerative medicine.
This work was supported by NIH Grant P20RR018757 (A.S.B. and N.P.R.), Rhode Island Foundation (A.S.B.), BD Biosciences (N.P.R.), a NSFC Grant 30872346 (X.F.), and California Institute for Regenerative Medicine Grant TR1-01277 (Y.X.). Additional support to A.S.B. and N.P.R. laboratories is provided by the administrative and flow cytometry cores within the NIH COBRE at Roger Williams Medical Center.
DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
The authors indicate no potential conflicts of interest.