Parkinson’s IPS cell trial in Japan switching to allogeneic

Jun TakahashiWhat’s better for stem cell trials such as for vision loss or Parkinson’s Disease: allogeneic or autologous cells?

In a major shift earlier this year, the induced pluripotent stem (IPS) cell trial in Japan for treatment of macular degeneration (MD) switched gears from using the patients’ own cells (called “autologous”) to using banked cells from other people, termed “allogeneic”.

Dr. Masayo Takahashi, the leader of this MD trial indicated the main reason was due to regulatory changes related to stem cells in Japan. This decision has delayed the clinical study, but there is hope it will restart soon. It appears for some as yet unknown reason the Japanese government has decided to only allow the use of allogeneic, matched IPS cells from cryobanks.

Now a second clinical study in the works in Japan also using IPS cells but as the basis for treatment of Parkinson’s Disease (PD) appears to be following suit. The PD trial, run by Dr. Jun Takahashi (pictured above; spouse of Masayo Takahashi, making them the world’s stem cell power couple), reportedly will also switch to focus on allogeneic cells.

The advantages of allogeneic cells include the fact that they can be validated and batch prepared in advance. In theory in this allogeneic system there might be no waiting period for patients while their own cells are turned into IPS cells. However, finding matches from a bank of IPS cells may prove somewhat difficult for allogeneic use for some patients. Even with major HLA type matching, minor mismatches could lead to some level of rejection. This could necessitate the use of immunosuppression. By contrast, autologous use of IPS cell-based products would likely require no immunosuppression after transplantation.

Together these changes in IPS cell clinical plans suggest a significant, broader shift in the field potentially toward allogeneic use of IPS cells. It’s not clear if other groups with IPS cell-based therapies in the translational pipeline, including in other countries, will follow suit or stick with the originally hoped for autologous focus.

21 thoughts on “Parkinson’s IPS cell trial in Japan switching to allogeneic”

  1. Universal Cells technology page shows one way to eliminate all HLA expression.

    They use rAAV homologous recombination to knockout beta-2-microglobulin to eliminate HLA class I (b2m is obligate heterodimerizing partner), and knockout RFXANK to eliminate all HLA class I (transcription factor). They can also introduce specific alleles of whatever molecules they want, including single chain class I.

  2. This is an excellent issue raised regarding the ability of VSELs to form teratoma and chimera by pluripotent stem cells. I will request you to refer to the link for detailed reply to this query.
    Briefly – VSELs are endogenous pluripotent stem cells with ability to differentiate into all cell types (this has been shown in vitro also), are mobilized in response to stress, but they neither expand in culture nor form teratoma nor integrate in a developing embryo to form a chimera. This is because Mother Nature has provided them with unique property of quiescence – if they behave in the body like ES/iPS cells behave in a Petri dish – our body will form tumors all the time.

  3. I’ve never heard of VSELs until today but I’m surprised to be reading another debate about whether a particular cell type is a stem cell. I read about definitive tests for pluripotency, such as teratoma and chimera formation. Have VSELs passed these tests?

    1. First the link to Mystery population paper
      VSELs are endogenous pluripotent stem cells, we do not need to grow them in a Petri dish at all for being available as a cell line for therapy. This is how they are different from ES/iPS cells. More research needs to be done to learn how to manipulate these endogenous VSELs in vivo. We are realizing that various age related diseases are more to do with a compromised niche than the stem cells. We observed VSELs to be 6 folds increased in a streptozotocin treated mouse pancreas by flow cytometry. Thus we do not need to administer more stem cells – to cure diabetes – need to develop another strategy. Diabetes is a disease of the niche … but more groups need to be convinced and get involved – and move the field forward. We recently restored spermatogenesis in a busulphan treated testis
      (doi: 10.4172/2157-7633.1000216) – something not yet achieved using ES/iPS cells and our recent paper (DOI: 10.1186/s13048-015-0199-2).

  4. Use of allogeneic iPS cells for regenerative medicine is equivalent to using hES cells. Then why use iPS cells at all? Human ES cells should suffice. iPS technology got Nobel prize because it offered overcoming immunological issues associated with hES cells and was a more ethical approach as it did not involve embryos. Somehow second point is irrelevant as hES cell lines are immortal and embryos are not required on regular basis. In fact we have to accept the fact that the field of hES/iPS cells has not progressed as expected. Main obstacle is that they tend to differentiate into their fetal counterparts and thus may be of little value to regenerate adult organs. hES/iPS cells are differentiated into tissue specific progenitors and it is hoped they will mature further upon transplantation. But this has not happened at least not in the field of pancreas – some beneficial effect is seen on transplanting pancreatic progenitors in diabetic mice but it is not long lasting as reported by various groups. Against this background – we have demonstrated a role for VSELs to regenerate all cell types in adult mouse pancreas after partial pancreatectomy ( and also argued that dysfunction of endogenous VSELs with age results in T2DM and pancreatic cancer (

  5. Hi Jeanne,
    with respect to my suggestion that genome edited cells to generate “universal” aka “off the shelf” cells, you wrote,

    “I forgot to mention- genetic engineering of cells would add a huge complication. Anything you hit with CRISP/R would need to be completely sequenced to make sure there weren’t any introduced deleterious off-target effects. ”

    Are you going to whole genome sequence every one of your autologous products before they go into your patients?

    CRISPR/Cas9 was published in August 2012 – lots of understanding of rules, and a lot of improvements have happened in 3+ years and a lot more are possible. TALENs not much before that, ZFNs only about a decade old. For that matter, iPS and direct reprogramming is not that old and is getting better every year (serum free, small molecules, for examples). I expect (ok, hypothesize) that cells from newborn – such as from cord blood – will be the way to go.

    DNA sequencing, with high accuracy, is cheap. Turnaround is now a couple of days (illumina HiSeq + parallel processing); Kits are available to preparing genomic samples with phasing. The genomic revolution will help QC both autologous and allogeneic/universal cells.

    The T-cell world our lab [Cooper lab, MD Anderson Cancer Center, ties to Ziopharm and Intrexon] lives in has labs and companies going autologous, and others going allogeneic/universal. We’ll know within five years the costs and efficacies. Last week’s news was Cellectis use of TALENs to treat an infant with genome modified CAR T-cells,

  6. With respect to Dr. Deepa Bhartiya’s two references on Very small embryonic-like stem cells (VSELs), I recommend reading Irv Weissman’s lab 2013 publication (open access) Miyanishi et al 2013 Do pluripotent stem cells exist in adult mice as very small embryonic stem cells? Stem Cell Reports,
    Here is the abstract:
    Very small embryonic-like stem cells (VSELs) isolated from bone marrow (BM) have been reported to be pluripotent. Given their nonembryonic source, they could replace blastocyst-derived embryonic stem cells in research and medicine. However, their multiple-germ-layer potential has been incompletely studied. Here, we show that we cannot find VSELs in mouse BM with any of the reported stem cell potentials, specifically for hematopoiesis. We found that: (1) most events within the “VSEL” flow-cytometry gate had little DNA and the cells corresponding to these events (2) could not form spheres, (3) did not express Oct4, and (4) could not differentiate into blood cells. These results provide a failure to confirm the existence of pluripotent VSELs.

    1. I had reacted to Miyanshi paper in 2013 in Nature News ( Irv Weissman came very close to discovering VSELs in 1998 when he reported a ‘mystery population’ in mouse bone marrow but it was left to Mariusz Ratajczak’s group in 2006 to publish detailed characterization of these stem cells. We are now reporting VSELs in azoospermic testicular biopsy of cancer survivors and in ovary surface epithelium in the links provided. Field of making gametes from ES/iPS stem cells has not moved whereas VSELs spontaneously differentiate into oocytes and sperm ( Correct science will not stop when one group publishes negative results (although it slows down). VSELs are very small in size, exist in very few numbers and are lost when cell suspensions are spun at 12-1500 rpm. VSELs pellet down only when cells are spun at 1000G. This is the underlying cause for discrepant data in literature.

      I sincerely thank Paul to provide us a forum for healthy discussion on stem cells.

  7. Hi Paul – thx for the news update on the intended switch to Allo for Dr. Jun Takahashi’s Pakinson’s trial.

    I had a couple of questions for Jeanne on the topic of immuno-suppression & US Allo banking.

    As I understand the use of immuno-suppression drugs in the Asterias & Ocata hESC cell transplantation trials to-date discontinued their use after a short period post surgery (60-90 days I believe).

    Do you think that Allo iPS cell therapies will be any different than the early trial designs for hESC in that regard? Also will the Auto trials need any as a result of the reprogramming?

    Also on this topic is the CNS area a relatively safer place to administer pluripotent cells than say into organs or other non-CNS areas given the often referred to semi immuno-privileged nature of the CNS or is it really only the eye that is somewhat protected? For example in Phase II for the eye they will be exploring different immuno-suppression regimes – including none at all.

    Thanks for your feedback and review of the genetic drift & HLA topic – important issues. Also on gene editing.

    There has been some doubt about there not being a perfect match in Allo, as Paul mentioned. Can you comment on the report that CDi have found 2 “superdonors” which can provide a partial HLA match to 19% of the US population. Is this work in the area a possible developing solution to create a viable Allo bank over time, similar to the Japanese iPS bank project?

    Auto as you say is preferable but if the Allo programs gain traction will there be a selective option for patients that may prefer Auto if their payer plans cover that versus a more universal approach?


  8. @ Dr. Loring – thanks for your insight regarding production costs of allogeneic vs autologous cell therapies. I’m finding it difficult to find informed opinion on this and even more so to find numbers – will you publish your findings?

    Such information could impact strongly on the choice of indication chosen by researchers and start-ups looking downstream at the global market value vs.COGS.

  9. Very interesting discussion. iPS and SCNT technologies were both developed to grow patient specific ES cells and now use allogeneic iPS cells for therapy is being discussed. We still believe that endogenous pluripotent ES-like stem cells in adult organs will be the ultimate candidates for regenerative medicine. You may wish to visit following two links to observe these stem cells to believe in them;

  10. I forgot to mention- genetic engineering of cells would add a huge complication. Anything you hit with CRISP/R would need to be completely sequenced to make sure there weren’t any introduced deleterious off-target effects. And, the longer you culture the undifferentiated cells, the more likely there will be selection for those that grow faster, which are probably not the ones you want to treat patients with. Keep in mind that differentiation puts selective pressure on the cell population, too, so that also needs to be watched.

    And, I want to emphasize, we do pick clones out of a set of at least three iPSC lines per patient, but not because they vary a lot in their ability to differentiate. When the protocols are good, the variability between people with different genetic backgrounds becomes less of a problem.

  11. Paul,

    I talk to a lot of people involved in organ transplantation, and everyone agrees that minor antigens are important. How important? I think that there is enough concern that even with well-matched donors, the recipients receive immunosuppression- they don’t want to take chances.

    About our progress: we’ve just started communications with the FDA, and should get their unofficial assessment of our plans early next year. We’ve been focusing on reproducibility among our different patient iPSC lines, optimizing differentiation protocols while developing and applying rigorous quality control assays. So far all of our funding has been from private donors, and we hope to obtain additional grant funding to speed development.
    No glitches, yet…remarkable for research!

  12. Hi Paul and all:

    I think everyone agrees that using autologous cells for transplant is the best option, if it’s possible. Immune surveillance for nascent tumors would be intact, and immunosuppression (at a cost of $80,000 a year for life) should not be necessary. We don’t want to knock out HLA.

    The only rationale for using allogeneic cells is cost- cost of generation and QC of the cells, then cost of getting approval. In Japan, the government has paid for the costs of identifying individuals who are homozygous for HLA alleles, and generating and banking their iPSCs. This has established a bank of iPSC cells that match, in the major haplotypes, most of the Japanese population.

    It is far more difficult in the US to obtain homozygous HLA cells because we are so ethnically diverse. In our lab we’ve found only one, a native of Kenya, who is homozygous for any of the major HLA types. But the alleles are not the dominant ones in the US population, so their use is limited.

    So, in the US it is currently impossible to achieve the same matching using allogeneic cells – I haven’t estimated the cost of locating enough homozygous individuals, getting their cells and informed consent, and reprogramming them. But it is gigantic.

    Another shortcoming for any allogeneic bank is the challenge of scaling up cell populations without genetic drift. And genetic drift can not only increase the chances of tumors (the best surviving cells thrive) but also may change the ability of the cells to differentiate using existing protocols.

    For Parkinson’s disease, we need only a few million cells per patient- the costs are minimal. And with efficient QC, the costs can be held down without compromising the safety of the cells.

    So, even after calculating the costs, my group is focusing on patient-specific iPSCs for Parkinson’s disease for now. Someday, there may be an allogeneic alternative. But not now.


    1. Thanks, Jeanne. Very interesting and valuable to hear what’s going on with your IPS cell work for Parkinson’s. One thing I’ve heard mentioned (and need to learn more about) is that even with major HLA matching that sometimes allogeneic cell transplants can be rejected. Is that a significant worry in the IPS bank arena or not so much? Can you comment on where things are at with your IPS translational work in terms of pre-IND, IND, etc.?

  13. Hi Paul,
    Knocking out (or down) HLA-A simplifies matching. See Torikai et al 2013 Blood, for example ( … disclosure, I now work in Cooper lab). that work used ZFNs with Sangamo, Cellestis is doing similar things with TALENs, and of course we now live in the age of the CRISPR/Cas9 revolution (and don’t forget about RNAi). Knocking out (or down) the other key loci, HLA-B, HLA-C, HLA-DR, is technically possible, though would risk donor cells being targeted by Natural Killer cells (NK cells), whose main mission is to identify and kill cells with “missing self”. There may be ways to engineer dummy HLAs, and/or install KIRs, to protect against NK cells.
    These may not be the exact numbers, but another option is identifying homozygous HLA donors (for -B, -C, -DR). HLA-A knockout from these will result in ~80 cell lines covering 95% of the Japanese population, and ~800 cell lines covering 95% of US population (in both case, ‘asymptotic long tails’ to get to 99%, 99.9% etc, so probably more practical to use immunosuppression for the long tailers). 80 and 800 cell lines is ‘industrial scale’, that is doable. Since these would likely be iPScell lines, could be QC’d in culture.
    Not mutually exclusive, another option is to identify “the best” iPS cell line, or “top ten” lines, with respect to functionality (differentiation to RPE or PD neurons, for examples), and rip out (or knockdown) the endogenous HLA’s, and install other alleles — homologous recombination is getting better every year.
    As for, “minor mismatches could lead to some level of rejection.” – minor histocompatibility loci (aka coding SNPs) are a minor problem compared to HLAs and KIRs.

Leave a Reply