Transdifferentiation does not make iPS cells obsolete

Do we still need iPS cells if transdifferentiation is so cool?

Wait, how did we get to the point that people are asking me that extreme question?

When iPS cells were first introduced to the world by Shinya Yamanaka, who I think will win the Nobel Prize sometime in the next few years, some folks went off the deep end. For example, opponents of human embryonic stem cell research (hESC) pronounced the need for hESC to be gone.

Others like Dr. Oz, to put it lightly, did not understand the nuances of the stem cell field and made gross misstatements. Oz had his train wreck of an appearance on Oprah saying so wrongly “the stem cell debate is over” with hero Michael J. Fox present. It is painful to even think about that stupid stunt by Oz so I’m not even going to link to the video.

A half a dozen years later, we still need hESC even though iPS cells are a quite powerful technology.

Now some are questioning whether we even need iPS cells if transdifferentiation is so powerful.

These folks ask “why not eliminate the middle man (i.e. iPS cells) if we can directly make the differentiated cells we want from cells like fibroblasts?”

I’m hearing this question more in the last week what with a new, very neat paper in PNAS from Marius Wernig on transdifferentiating fibroblasts into neural progenitors.

Previously Wernig’s team had made neurons from fibroblasts. What makes this new paper quite interesting is that neural progenitors are likely to be far more useful for regenerative medicine than actual neurons because neurons made in a dish are unlikely to transplant and engraft well.

My colleague Amy Adams from CIRM has a nice piece on Wernig’s work on the CIRM Research Blog.

Transdifferentiation is without a doubt a very powerful approach to exploit in the future to make clinically relevant stem cell-based drugs for regenerative medicine. However, it is very new (making even iPS cells seem like a mature technology) and there are very few publications on it.

My own lab toyed with it a few years back with some interesting, but unpublished results.

The reason why I think transdifferentiation has not been published more is that it is a very tricky, inefficient technology right now. It is very possible that this will change, but it is still early days. For this reason alone, I think iPS cells are here to stay for a long time.

In addition, we also do not know how cells made from iPS cells and transdifferentiated cells really stack up head-to-head in a clinically relevant setting. iPS cell derivatives could prove superior. The jury is out.

So for now, we need hESC and iPS cells as we continue to learn more about transdifferentiation, a very exciting new technology. It is also important to point out that transdifferentiation was only something people tried after being inspired and instructed by Yamanaka’s iPS experiments!


  1. Interesting piece. Your point that the jury is still out on iPSC vs transdifferentiation makes sense. Regarding efficiency, there is data by Ieda et al. (Cell, Volume 142, Issue 3, 6 August 2010, Pages 375-386) showing that the efficiency of direct programming of fibroblasts to cardiomyocytes using Gata4, Mef2c, & Tbx5 is 20%, compared to <0.1% efficiency using the Yamanaka factors to derive iPS cells from fibroblasts. As stated in the preview to this article, so-called iCMs (inducible cardiomyocytes) may provide an advantage over iPS-derived cardiomyocytes since they can be produced in greater yields with faster kinetics, and do not have residual pluripotent cells that could produce teratomas ( Even if iCMs prove to win out over iPS for clinical applications, iPS derived from patient fibroblasts will still be useful for creating models of congenital diseases in order to understand the effects of gene mutations on the various stages of lineage-specific development.

    • Chris, this is a good point. I would say, however, that that 20% efficiency reported seems almost too good to be true in general in the field and anecdotally I’ve heard that the efficiency of transdifferentiation is usually <0.1%. However, perhaps in some contexts the efficiency of transdifferentiation in certain experiments can be very high.

      If so, I'm then wondering why are so few papers out there on transdifferentiation?

      Whatever the current efficiency, I predict that transdifferentiation is here to stay, it will continue to improve, and ultimately be extraordinarily useful for clinical applications including applications where iPS cells and ESC may not be applicable.

  2. Great stuff as usual, and thank you for this blog. Looks like Yamanako plans to enter first ever human clinical trials (in Japan?) sometime next year targeting AMD with an iPS cell transplant. IMO, reviewing efficacy data side-by-side with that of Advanced Cell Technology’s hESC derived cells, currently in a Phase I/II FDA trial for the same disease, should provide for some compelling debate and hopefully future treatment options. And yes, the jury is still out.

Comments are closed.