|Dr. Kyoko Masuda
Assistant Professor, Laboratory of Immunology, Institute for Frontier Life and Medical Sciences, Kyoto University
|B.S., Department of Chemistry, Kwansei Gakuin University
|April 2005–March 2007
|Research Fellow, Japan Society for the Promotion of Science
|Ph.D., Graduate School of Biostudies, Kyoto University
|April 2007–December 2008
|Postdoctoral Fellow, Department of Genetics, University of Georgia
|January 2009–September 2012
|Research Fellow, RIKEN Research Center for Allergy and Immunology
|October 2012–September 2016
|Assistant Professor, Institute for Frontier Medical Sciences, Kyoto University
|Assistant Professor, Institute for Frontier Life and Medical Sciences, Kyoto University
Cancer immunotherapy has long been proposed as a therapy in addition to the three main cancer treatments: surgery, radiation, and chemotherapy. The efficacy of immunotherapy, however, has long remained insufficient to allow it become a standard cancer treatment. In the recent years, cancer immunotherapy has progressed significantly. Recently, the field of cancer immunotherapy has been made great progress because immune checkpoint blockades, such as anti-CTLA-4 and anti-PD-1 antibodies, have demonstrated their efficacy against some solid cancers in several clinical studies [1, 2, 3], and some of these antibodies have been approved for the treatment of several types of cancer. Nonetheless, there have been challenges; the efficacy rate is still only 20% to 30%, and adverse reactions related to autoimmune diseases occur with a high frequency (~50%). The immune checkpoint antibodies are used to block the inhibitory mechanism of the immune system, thereby the cytotoxic T lymphocytes (CTLs) show the continuing activation and the cytotoxic effect against cancer cells. However, these antibodies enhance the immune system in an antigen-independent manner, provoking autoimmune diseases as mentioned above.
The cancer immunotherapies using CTLs, nonetheless, have been developed. For example, Dr. S. A. Rosenberg and colleagues at the U.S. National Institutes of Health (NIH) treated melanoma patients with tumor-infiltrating lymphocytes (TILs) that had been expanded ex vivo .
Another one is the T cell receptor (TCR) gene therapy. Patients’ peripheral T cells were transduced with TCR gene specific for a cancer antigen and infused to the patients. This method had a certain effect for specific types of cancers [5, 6, 7]. Notably, though, this is a gene therapy and has the risk of tumorigenesis by the engineered T cells. Also, in both the TIL therapy and the TCR gene therapy, patients basically underwent these therapy as a autologous transfusion. It means that good state (not an anergic or exhausted) of T cells are expected to be obtained from patients. In either method, the shortage of T cells for the treatment has remained problematic because it is not easy to expand T cells ex vivo.
As described above, the CTL therapy has faced the difficulty of generating a sufficient number of antigen-specific CTLs. To overcome this issue, we utilized the reprogramming technology to T cell cloning (Figure 1). Pluripotent stem cells such as ESCs or iPSCs, can be established by reprogramming antigen-specific T cells. The genome structure of rearranged TCR genes is inherited in the pluripotent stem cells. When T cells were differentiated from these pluripotent stem cells all T cells will express the same TCR as the original ones. We could generate fresh T cells as many as possible because cells can be expanded almost unlimitedly at the pluripotent stem cell stage.
We first established iPS cells from CTLs specific for the MART-1 antigen uniquely expressed in melanoma cells (MART1-T-iPSCs)  (Figure 2A-C). The source of cells used in this experiment was a T cell line that had been isolated from a melanoma patient and maintained at NIH. Next, T cells were differentiated from MART1-T-iPSCs (Figure 2D). On day 35, CD4/CD8 double-positive cells (DP cells) were generated (Figure 2E). At this point, TCR was stimulated by adding an anti-CD3 antibody to the culture medium. Six days later, CD8 single positive T cells were generated. Almost all regenerated T cells expressed TCR that can recognize the MART-1 antigen (Figure 2F).
Figure 1. Regeneration of antigen-specific T cells using the iPS cell technology
The concept of the strategy. When ES/iPS cells are established from cancer antigen-specific T cells, the genomic structure of rearranged T cell receptor (TCR) genes is inherited in the ES/iPS cells. When ES/iPS cells are differentiated into T cells, all regenerated T cells will express the same TCR as the original ones.
Figure 2. Regeneration of MART-1 antigen-specific T cells using the iPS cell technology
A. iPS cells were established from MART-1 antigen-specific T cells (MART1-T-iPSCs).
B. Flow cytometric profile of original MART-1 antigen-specific CTL cell line.
C. The morphology of the MART1-T-iPSCs.
D. MART-1 antigen-specific T cells were regenerated from MART1-T-iPSCs. The iPSCs were seeded on OP9 cells and co-cultured for almost 2 weeks. Loosely adherent cells were collected and transferred to OP9/DLL1 cells for the T cells differentiation. On day 40, an anti-CD3 antibody was added to the cells to stimulate TCR.
E. Flow cytometric profile of cells on day 40. CD4/CD8 DP cells were generated.
F. Flow cytometric profile of cells after TCR stimulation.CTLs were generated after 6 days of culture in the presence of the anti-CD3 antibody. Most CD3-positive cells were MART-1 antigen-specific CTLs.
The method introduced above is for an “autologous transfusion” method in which the original cells are collected from the patients, regenerated ex vivo, and transfused back to the patients. However, the autologous transfusion has several disadvantage; “costly” and “time consuming” for the quality control of cells. We therefore decided to apply our method towards “allogeneic transfusion.”
In our concept, for example, T-iPSCs are established from T cells specific for a tumor antigen from healthy volunteers, and CTLs are regenerated from them. We can evaluate the quality of iPSCs or function of CTLs in advance and cryopreserved them if they fulfill the criteria. These cells can be thawed and immediately used to treat HLA-matched patients whose cancer express the same tumor antigen. iPSCs will not be needed to be established for each patient, thereby greatly reducing costs. Also, it is possible to evaluate the function of CTLs in advance, and therefore the efficacy of transfused cells can be ensured. The greatest advantageous point is that the patients can receive off-the-shelf T cell therapy immediately after a diagnosis is made.
Further, the risk of tumorigenesis of transfused cells can be avoided by allogeneic transplantation. This is because the transfused cells are non-self cells and immune system of patients will finally reject them. In many field of regenerative medicine, the regenerated tissues are expected to be functional for the rest of life after transplantation. However it is not necessary the case in immunotherapy field, but it is rather the advantageous point that regenerated cells are finally rejected after they functioned for a certain period.
The WT-1 antigen is a cancer antigen frequently expressed in leukemia and various types of solid cancers . By aiming for allogeneic transfusion therapy using our method, we tried to establish T-iPSC from WT-1 specific CTLs . CTLs of a healthy volunteer’s peripheral blood were stimulated by WT-1 peptide for six weeks in culture, and WT-1 antigen specific CTLs were expanded to reach almost a half of CD8-positive population (Figure 3A). These WT-1 specific CTLs were used to establish iPSCs (Figure 3B). When T cells were regenerated from WT1-T-iPS cells, almost all cells were WT-1 antigen-specific CTLs (Figure 3C).
When we reported our study described in Figure 2 in 2013, regenerated T cells were not the conventional type of CTLs although they had a cytotoxic activity against target in an antigen-independent manner. While conventional type of CTLs express heterodimers consisting of CD8α and CD8β (CD8αβ), innate type of CTLs, such as intestinal CD8-positive T cells and a part of γδT cells, express CD8α homodimer (CD8αα). CD8αβ functions as a coreceptor that enhances TCR signal. CD8αα co-receptors, however, do not bind to MHC molecules, resulting in reduced TCR signal. The regenerated CTLs in our report were the CD8αα type. Therefore, we tried to improve the culture method. In our newly developed method, DP cells were first separated from CD4/CD8 double negative (DN) cells and were then stimulated. Then, CD8αβ-type T cells were efficiently induced (Figure 3D). We also found that CD8αβ-type T cells were not generated when whole cells were stimulated because DP cells were killed by stimulated DN cells.
The WT-1 antigen-specific CTLs regenerated by the improved method had same cytotoxic effect as the original CTLs (Figure 3E). The regenerated CTLs were also highly cytotoxic to a leukemia cell line expressing WT-1 antigen. In addition, the therapeutic effect of regenerated CTLs administration was demonstrated in a leukemia mouse model, in which human leukemia cells had been inoculated into immunodeficient mice before treatment (Figure 3F).
To evaluate the safety of regenerated CTLs, we investigated whether normal tissues were damaged in vivo, and whether the inoculated cells underwent a tumorigenic transformation. In this case, only the regenerated CTLs were inoculated into immunodeficient mice. During six months of observation period, no tissue damage was seen and transleered cells did not develop leukemia. These results indicate that regenerated CTLs are safe.
The regenerated CD8αβ-type CTLs could be expanded in culture by over 10,000-fold in 2–3 months. Approximately 106 CD8αβ-type CTLs can be generated from DP cells in one 10cm dish, and these CTLs can be expanded to over 1010 CTLs. 1010 CTLs will be sufficient enough for the cell therapy, thinking that 108-109 cells are usually used in T cell therapy.
Figure 3. Regenerated WT-1 antigen-specific CTLs showed therapeutic effect in a leukemia mouse model.
A. Flow cytometric profile of peripheral blood of a healthy volunteer before and after simulation by WT-1 peptide. WT-1 antigen-specific CTL population in CD8-positive cells was detected after stimulation for 6 weeks.
B. The morphology of WT1-T-iPSCs.
C. Flow cytometric profile of cells generated from WT1-T-iPSCs. WT-1 antigen-specific CTLs were regenerated using the same method as in Figure 2D. Nearly all regenerated CTLs were WT-1 antigen-specific CTLs.
D. Flow cytometric profile of conventional type of CTLs regenerated from WT1-T-iPSCs. CD8αβ-type CTLs were efficiently regenerated using a newly developed method.
E. Comparison of the cytotoxic activity between the original CTLs and the regenerated CTLs. The original cells were cryopreserved. The effector cells were coclutured with a B lymphoblastoid cell line used as a target cells at different concentration of WT-1 peptide. E:T ration was fixed at 3:1. The percentage of dead target cells was measured after 6 hours.
F. A Kaplan-Meier curve depicting the percentage of survival with or without the treatment in the mouse model. 2 x 104 cells of HL60, a human leukemia cell line expressing WT-1 antigen were intraperitoneally inoculated into immunodeficient mice. From the next day, regenerated CTLs (5 x 105 cells) in phosphate-buffered saline (PBS) were intraperitoneally inoculated into the mice once a week for a total of four times. The same volume of PBS was intraperitoneally injected into the mice in the control group.
I have thus far described the T-iPSC method in which iPSCs are established from T cells. However, this method is still has some problems: we need to select the best iPSC clone, which are heterogeneous in term of T cell generating potential, and TCR affinity differ very much between clones. Such selection process would be time- and money-consuming. To address this issue, we are currently developing an alternative method, TCR-iPSCs. TCR-iPSCs was established by transfecting TCR genes into iPS cells by lentivirus. In most cases, the transfected TCR genes are expressed at early stage of their differentiation process, and their expression is expected to block the rearrangement of the endogenous TCR genes via the so-called “allelic exclusion” mechanism.
We cloned the TCR genes from the regenerated CTLs derived from WT1-T-iPS cells shown in Figure 4 and transfected them into iPS cells of non-T cell origin (WT1-TCR-iPS cells) (Figure 4A). WT1-TCR-iPS cells were successfully differentiated into CD8αβ-type CTLs. The regenerated CTLs showed antigen-specific cytotoxicity comparable to that of CTLs regenerated from WT1-T-iPS cells (data not shown).
If this method could be applied to allogeneic transplantation, the versatility of our approach would greatly increase. As for the TCR gene, it is desirable to use a TCR gene the efficacy and safety has been demonstrated in the clinical study. In addition, the quality of iPS cells would be ensured by use of the homozygous HLA haplotype iPS stock, provided by the Center for iPS Cell Research and Application of Kyoto University, which would facilitate the safety assurance process (Figure 4B).
It should be noted, however, that the TCR-iPSC method is regulated by two points that it is a gene therapy and a regenerative medicine using stem cells. As such, how to ensure safety will be an important task.
Figure 4. The TCR-iPSCs strategy
A. WT-1 antigen-specific TCR genes were transfected into iPSCs to establish WT1-TCR-iPSCs. Regenerated CTLs were differetiated from these iPSCs.
The regenerated CTLs derived from TCR-iPSCs showed antigen-specific cytotoxicity as same as the ones from T-iPSCs (data not shown).
B. The Center for iPS Cell Research and Application of Kyoto University has been ongoing a project to stock iPS cells generated from HLA haplotype-homozygous volunteers. High-quality of TCR-iPSCs cells can be established if quality-guaranteed iPS cell lines were transduced with a cancer antigen-specific TCR genes that their safety and efficacy was ensued in the clinical study. Our strategy could achieve “Off–the-shelf CTLs” against various type of cancer for most people.
This review describes the potential of a regenerative medicine against cancer. The WT-1 antigen described here is expressed in a wide range of solid cancers. By targeting other cancer antigens or neoantigen that arise tumor specific mutation, most type of cancers could be a target of this treatment in the future.
The TCR-iPSC method is a good tool to cover such neoantigens.
On the other hand, the antigen-specific CTLs therapy can cover only patients who has a certain HLA types. Currently targeted haplotypes are HLA-A*24:02 or HLA-A*02:01. Since approximately 70% of the Japanese people has either of HLA-A*24:02 or HLA-A*02:01, the antigen-specific CTLs therapy will be available to a large number of people. In the future, other HLA types will be targeted so that the treatment will be available for most people.
Also, we have to think about the HLA type of the iPSC itself when our method will be applied to an allogenic transfusion setting. As of December 2017, the iPS cell stock lines provided by the Center for iPS Cell Research and Application of Kyoto University are homozygous for the two most common alleles in the Japanese population, representing 24% of the total. The stock will be expanded in the order of allele frequency and will cover 70% to 80% of the Japanese population in the next 6–7 years.
Although there are many challenges to putting our CTL therapy into a bedside, it would be a breakthrough for cancer therapy. We also hope that the concept of “off-the-shelf T cell therapy” will be applied not only to cancer, but also to any diseases related to immunity, including infectious diseases, autoimmune diseases, and allergy.