Stem cell transplants: hIPSC and hESC behave similarly in brain & often fuse with host cells

What happens following pluripotent stem cell transplants into the brain? Are human IPS and ES cells going to function similarly in this context?

We recently published a new translational paper on the behavior of human pluripotent stem cells when transplanted into the adult mouse brain in collaboration with my great UC Davis colleague, Dr. Veronica Martinez-Cerdeno.

Martinez-Cerdeno Figure 6
Martinez-Cerdeno Figure 6

Martinez-Cerdeno V, Barrilleaux B, McDonough A, Ariza J, Yuen B, Somanath P, Le C, Steward C, Horton K, Knoepfler P. Behavior of xeno-transplanted undifferentiated human induced pluripotent stem cells is impacted by microenvironment without evidence of tumors. Stem Cells Dev. 2017 Jul 10. doi: 10.1089/scd.2017.0059.

We are excited about this paper for a number of reasons. One of the things that makes it pretty unique is that we in parallel transplanted undifferentiated human iPSCs and human ESCs (hIPSC, hESC), whereas most studies pre-differentiate pluripotent stem cells. Even so we did not observe teratomas in the transplant recipients’ brains, which surprised us. It was also unexpected that we found that the mice did not strongly reject the transplants despite the animals being immunocompetent. The end result is that the mice survived without tumors and we got reasonable engraftment of both cell types and could compare their behavior.

You can see Figure 6 of our paper below, showing the fate of transplanted hIPSC where green staining highlights human cells and red staining is for the different indicated fate markers. Most human stem cells took on the appearance of NeuN+ neurons, while in contrast we did not observe human astrocytes. The numbers of human cells (and their fates) present in transplant recipient mice varied by region of the CNS although they were all injected stereotactically in the same location in the lateral ventricle, suggesting that microenvironment impacts transplanted cell behavior in key ways.

Why no tumors? In theory, immune rejection could have explained the lack of tumors. The mouse immune system perhaps killed the specific subpopulation of cells within the injected pluripotent stem cell cultures that had tumorigenic potential. It is also possible and I think more probable, since we observed a surprisingly high rate of fusion of the transplanted hIPSC or hESCs with host mouse brain cells, that this fusion eliminated tumorigenic potential.

We theorize that after human cell fusion with mouse cells, that nuclear fusion quickly followed since we did not observe multi-nucleated cells. The cell fusion we report is something that everyone studying cell transplants should look for in their studies. In theory transplanted stem cell fusion may not be entirely a bad thing as it could be exploited for drug delivery if we can better understand the mechanisms of fusion, but fusion can also potentially complicate apparent findings, especially if it isn’t on your radar screen.

Just because we didn’t observe tumors, doesn’t mean that undifferentiated human pluripotent stem cells would be safe to use clinically (I don’t think they would be) and other groups have reported teratoma in similar although not identical translational studies. In addition, we found in contrast to our human cell results that transplantation of mESCs did lead to teratoma or similar kinds of tumors such as one that appeared like a teratocarcinoma. It remains unclear why the mouse cells formed tumors while the human did not but xenotransplantation may lead to other effects such as the aforementioned cell fusion that impact biological outcomes.

There are a few limitations to our study on pluripotent stem cell transplants. For instance, it isn’t clear how a wider panel of hIPSC and hESC would behave, and some may be more predisposed to form teratoma. Also, we don’t know if the xeno nature of our transplants catalyzed the fusion and that leaves it unclear how much human stem cells would fuse with host differentiated cells following transplants. The clearest indicator of the rate of engraftment was by staining with a human-specific antibody and we also got some nice data by FISH, but while we tried a quantitative approach using qPCR to more precisely measure human gDNA present in the transplanted murine brain, this assay only worked in certain conditions. Most likely the qPCR assay sensitivity was limited by various factors including using fixed brain tissue DNA as starting material.

Overall, hIPSC and hESC behavior was quite similar in our study, further supporting the growing notion that reprogramming produces human cells generally equivalent to hESC, but a unique thing here in our study is that we found this similarity even within the transplant setting context in the brain (not just in a dish).

Many interesting follow ups could build on this work’s foundation including examining the impact of brain injury on transplanted undifferentiated hIPSC and/or hESC behavior. I’d be interested in any thoughts you have regarding the data in our paper on pluripotent stem cell transplants.

8 thoughts on “Stem cell transplants: hIPSC and hESC behave similarly in brain & often fuse with host cells”

  1. Hi Paul,
    Thank you for all your posts, so interesting and helpful.
    I guess one of the first disease to be targeted by hIPSC or hESC transplantation in the brain shall be Huntington disease. Are you aware of clinical trials in process, beside the MIG-HD trial which does not use stem cells? Has nuclear fusion occured in such non-stem cells grafts? What can we expect?

  2. Jeanne A. pawitan

    Hi Paul,
    I think it will be interesting if you do fluorescent in situ hibridization, to see whether the cell nuclei Interphase nuclei) contain human chromosomes. May be using multicolor chromosome painting method for several certain human chromosomnes?

    Jeanne A. Pawitan

    1. Hi Jeanne, We did do FISH. We did it with both mouse and human probes together and separately. There were a lot of cells that had both mouse and human DNA in the same nucleus. Can you access the paper? Take a look at Figure 7.

      1. Jeanne A. pawitan

        Thanks for the information, but unfortunately I can not access the paper, as it is not open access. So it is proven then, that the cells fused.

  3. Thanks Paul, this is even more fascinating than I had imagined! Even if the fused cells (with fused nuclei) can’t proliferate, might they still be capable of making some human-type proteins that have a functional impact on the mouse? One wonders how long the fused-fused cells might persist in the mouse? (Persistence seems relevant to any possible therapeutic application — like patching up defective somatic cells by adding some functional genes.)

    Or am I making an extrapolation that is just too wild?

    1. Hey Brian, Yes, it’s a cool area and not well understood. Scientists have mainly studied fusion as a means to do experiments such as SCNT, hybridomas, etc.

      It may be that cells both in natural conditions in the body, in dishes in the lab, and after cellular transplants in vivo fuse much more often than we realize.

      Especially with cells of the same species and perhaps even of the identical genetic background, how would we even know if they fused?

      We can measure DNA content, we can look for binucleated cells, etc. such as by FACS, but it’s an imperfect system in many ways. How would we know if a tetraploid cell was a result of fusion or just spontaneous tetraploidy?

      Cells may also fuse and then unfuse prior to cell division.

      It may not be exactly considered fusion, but cells also sort of do a mini-fusion dance as well via sharing cellular constituents like exosomes.

      In our study, we saw evidence of human cells in the mouse brains long term so I think they are there permanently most likely and some of them are fusions.

      Yep, there are many therapeutic implications, both positive and negative potentially.

  4. Hi Paul,
    I’m both interested and a bit confused. My understanding of the definition of cell fusion is that it would cause two uni-nuclear cells to combine so that the fused cell was multi-nuclear. The concept of “nuclear fusion” is something that I know a bit about with regards to physics — eg hydrogen fusing to give helium and helium fusing to give Be, C, O — but I have to say that I’m totally confused as to what “nuclear fusion” might mean for the nuclear material of biological cells. Presumably it does not mean the production of a whole new type of nuclear material that is neither mouse nor human? My guess is that it would mean the dismantling of one nucleus in favour of the other?

    By analogy with
    I wonder if your discovery might be more useful for delivering useful cellular materials as opposed to delivering “drugs”? By useful cellular materials I mean the things that occur naturally within the cell that is to be helped. Or am I confused by the modern fashion of labeling seemingly everything as a drug?

    1. Hey Brian,
      Really good questions.
      In this case by “nuclear fusion” we meant that after two cells fuse and presumably a transient period where there is now one chimeric human-mouse cell that is binucleated, that then soon the two nuclei fuse (this occurs when the nuclear lamina bind and then become one continuous membrane encompassing all the chromatin from both cells) to form one chimeric nucleus.

      We reasoned that this happened because we found many chimeric human-mouse individual cells in the brains of transplanted mice, but we did not find amongst them any binucleated cells. Because the transplanted human cells apparently fused mostly with differentiated mouse brain cells, we think that the resulting chimeric cells each with one nucleus (but mixtures of chromosomes from human and mouse) could no longer proliferate and the mixture of differentiated and pluripotent cellular factors likely by default leads to a differentiated cell identity. This is probably one reason why we didn’t see tumors.

      However, it is also possible that instead of there being nuclear fusion, after cellular fusion the hybrid cell does go on to divide once or twice, and during mitosis the nuclear lamina dissolves in the hybrid cell, and then after S phase and during/after cytokinesis nuclear laminas (lamini?) reform in each of the daughter cells around all the segregated chromosomes now in each separate cell, which are mixtures of human and mouse (I don’t think cells can differentially segregate human and mouse chromosomes). But there are past reports of nuclear fusion after cell fusion so we thought that is more likely. If you can access this paper see Figure 7 (

      Does this all make more sense? I should have explained it more in the post.

      Yes, as to your final question, our finding may be relevant for delivery of endogenous cellular materials too such as important proteins that can impact cell behavior.

Comments are closed.