Could induced pluripotent stem cells (iPSCs) be just as good as embryonic stem cells (ESCs) for therapeutic use? Although iPSCs have advantages for research, as they sidestep the ethical problems of ESCs and make useful disease models (see ARF related news story), their value for transplants is less certain. Current reprogramming technologies leave slight epigenetic differences between iPSCs and ESCs, which concerns some researchers. Adding fuel to the fire, scientists led by Yang Xu at the University of California in San Diego reported in the May 13 Nature that iPSCs are more likely to be rejected by the immune system than are ESCs. The authors traced this phenomenon to the expression of specific factors by the cells, which suggests that the problem might be lessened by better reprogramming technologies.

To test immunogenicity, first author Tongbiao Zhao derived ESCs and iPSCs from inbred B6 mice. To make iPSCs, he reprogrammed B6 mouse fibroblasts using either three or four genes. In one method, he delivered the reprogramming factors via a retrovirus, and in the other approach he used an episome, a piece of DNA that does not integrate into the animals’ genome. When Zhao and colleagues implanted ESCs into B6 mice, the cells formed large teratomas containing many differentiated cell types. Recognizing them as “self,” the immune system did not attack these ESC-derived teratomas. On the other hand, the animals’ immune systems mounted a fierce assault on implanted iPSCs that had been reprogrammed with retroviral vectors—even though the cells came from genetically identical animals. Teratomas either failed to form, or contained many infiltrating T cells and a lot of necrotic tissue. The immune response was gentler on iPSCs reprogrammed with non-integrating episomal vectors, suggesting that episomal-derived iPSCs are more similar to ESCs. iPSCs made with episomes formed teratomas easily, but T cells infiltrated the majority of them, and about 10 percent of teratomas became necrotic.

To discover the reason for immune system rejection, the authors compared the gene expression profiles of teratomas derived from ESCs with those from iPSCs made with episomes. They found nine genes that were commonly overexpressed in regressing teratomas formed from iPSCs. To test these genes, Zhao and colleagues expressed each one in ESCs and looked for immune rejection of the resulting teratomas. The genes Zg16 and Hormad1 were most strongly associated with rejection. “The abnormal expression of Hormad1 and Zg16 contributes directly to the immunogenicity of the cells,” the authors concluded.

This implies that reprogramming technologies may need to be optimized to reduce epigenetic differences between ESCs and iPSCs, the authors suggested, pointing out that their teratoma assay could provide a useful screening platform for improving reprogramming. Others in the field are more cautious as to whether this finding indicates a problem for iPSC transplantation, noting that the experiment involved injecting undifferentiated iPSCs. Rudolf Jaenisch at the Whitehead Institute, MIT, wrote to ARF, “In a patient, one would inject, of course, only differentiated functional cells.” (See full comment below.) In previous experiments from his group, Jaenisch said, they injected iPS cell-derived hematopoietic stem cells to correct sickle cell anemia in a mouse model, and saw no immune rejection, suggesting that differentiated cells are not immunogenic (see Hanna et al., 2007). Xu disagrees, writing to ARF that, because the iPSCs differentiate into numerous cell types in the teratomas, “in our immunogenicity assay, we were simultaneously evaluating the immune responses to almost all the differentiated cell types present in the body. There are very few iPSCs in the teratomas, so the immune reactions detected are against the differentiated cell types.”

Mahendra Rao, vice president of stem cells at biotech company Life Technologies, agreed with Jaenisch that the immunogenicity of pre-differentiated cells is the key question. Rao wrote to ARF, “It is too early to suggest that this will change our view of either ESCs or iPSCs in any fundamental way.” (See full comment below.) Many labs are using iPSCs derived from people with various disorders to create models for a range of neurodegenerative diseases (see ARF related series). This work should not be affected by the new findings.—Madolyn Bowman Rogers


  1. This is quite carefully done in the sense that ESCs and two different types of iPSCs were considered for the experiments. The results seem clear that iPSCs appear to be more immunogenic that ESCs, and that integration-free iPSCs seem less immunogenic than cells that carry integration.

    The important result I would have liked to have seen is whether there is rejection of differentiated cells, which would be the cells that would be used in a potential therapy. Given that we can transplant tissues and organs with minor mismatches, it seems to me that this is the critical experiment to determine if this is a major issue. Also, the process of growing cells and perturbing them requires culturing undifferentiated cells differently, and the purity of the iPSCs may be different, and these differences may represent technical issues rather than a fundamental difference in the biology of cells.

    It is too early to suggest that this will change our view of either ESCs or iPSCs in any fundamental way, but I do expect that independent replication of the data will be attempted and we will learn more about the basis of this result.

    View all comments by Mahendra Rao
  2. I think that these data are certainly of interest. However, it is possible that the immune reaction is only directed against the undifferentiated cells and not against functional differentiated cells. In a patient, one would inject, of course, only differentiated functional cells, and would eliminate all undifferentiated cells as they would cause tumors.

    Actually, in our previous experiments where we injected iPS cell-derived hematopoietic stem cells to correct sickle cell anemia, we found no reaction against these cells whatsoever. So these data, not mentioned in this paper, would argue that the immune rejection may indeed be directed against the undifferentiated stem cells rather than the differentiated cells.

    View all comments by Rudolf Jaenisch
  3. I thank Rudolf Jaenisch and Mahendra Rao for their comments on our findings. I would like to respond to Jaenisch’s comment. We are fully aware of his group’s elegant study published in the journal Science showing that hematopoietic stem cells derived from genetically manipulated induced pluripotent stem cells can be transplanted into the mouse model of sickle cell anemia without being immune rejected. However, their experimental design did not allow the evaluation of the recipients’ immune responses to the graft because the recipients were lethally irradiated (totally 11-12 Grays of ionizing radiation), which wiped out their immune systems. This is the primary reason why we did not cite Jaenisch's work in our paper. Including my own work, this bone marrow (BM) transplantation protocol is routinely done in the mouse studies to allow efficient bone marrow graft and prevent graft rejection. With this protocol, even the mismatched bone marrow (most published data involve donors from 129/BL6 mixed background and thus without a defined major histocompatibility complex profile) can be used to repopulate the hematopoietic system of the allogenic recipient without being rejected.

    In the course of revising our paper, we thought a lot about how to address the immunogenicity of hematopoietic stem cells derived from iPSCs. However, it is very difficult to design such an experiment due to the findings that, after sublethal irradiation, the donor BM and recipient BM still can develop hematopoietic chimerism (the hematopoietic cells are derived from both donor and recipient), even when the donor cells are allogenic. Once the stable chimerism is established, the allogenic hematopoietic cells derived from the allogenic donor hematopoietic stem cells are immune tolerated by the recipients, making it impractical to study the immune responses to the donor cells that only harbor minor antigens.

    View all comments by Yang Xu
  4. This paper by Zhao et al. demonstrates that allogeneic embryonic stem cell (ESC) transplantation is more immunogenic than syngeneic ESC transplantation in mice. Likewise, allogeneic induced pluripotent stem cell (iPSC) transplantation is less immunogenic in recipients than syngeneic iPSC due to the graft versus host disease (GVHD) triggered by the recipient's immune response. Furthermore, the methods used to reprogram iPSC (retroviral vs. episomal) contribute to immunogenicity even in syngeneic recipients.

    The research conducted in this paper delivers an important message of how efficacious iPSCs can be in clinical applications. More importantly, the question should be, How safe are iPSCs when used as a therapeutic intervention?

    Beyond immunogenicity-mediated rejection of transplanted iPSCs, the safety and toxicity of iPSCs need to be tested in preclinical settings as proof-of-concept studies. Even such an attempt could be a challenge, given the fact that it is human cells that would be put into animal hosts. Additionally, just like all other pharmaceutical and biopharmaceutical agents, it is important to determine the aspects of absorption, distribution, metabolism, excretion, and toxicity (ADME/T) of iPSCs. Specific questions need to be properly addressed, such as, Are these cells homing into the targeted sites? Do they perform as/differentiate into the desired cell population? Do successfully differentiated cells in the host support normal tissue/cellular function? Do implanted iPSCs generate any cellular metabolites which could subsequently mediate any adverse events, etc.?

    The development of iPSCs as a medical therapeutic is still in its infancy. This paper shines a light on future considerations, which could be to develop autologous iPSCs as a therapeutic agent, and compare their safety and efficacy with adult ESCs and allogeneic iPSCs.

    View all comments by Chi Tarn

Make a Comment

To make a comment you must login or register.


News Citations

  1. Where in the World Are the iPS Cells?

Paper Citations

  1. . Treatment of sickle cell anemia mouse model with iPS cells generated from autologous skin. Science. 2007 Dec 21;318(5858):1920-3. PubMed.

Further Reading

Primary Papers

  1. . Immunogenicity of induced pluripotent stem cells. Nature. 2011 Jun 9;474(7350):212-5. PubMed.