Taboo topics about iPS cells: the elephant in the lab series

elephant in the room

Wait, what’s that taboo elephant doing in the IPS cell lab?

You don’t see it?

I’m starting a new series called “The Elephant in the Lab” where I discuss controversial laboratory topics that people are usually too afraid to publicly discuss. We are starting with iPS cells. (for a description and definition of iPS cells go here).

iPS cells have only been around a half dozen years, but already there are semi-taboo topics about iPS cells that collectively the field is disinclined to talk about publicly. These issues are complicated and not always pleasant, but ignoring them won’t make them go away. In fact, quite the contrary talking about them will in reality help resolve them and possibly in the future make some of these issues in effect “go away”. Despite these issues discussed below, I still am a huge fan of iPS cells and believe in their great potential. However, the more openly we discuss iPS cells the better we can advance the field.

Here are my top 10 taboo topics (In other words, what everyone is thinking but doesn’t dare to say).

1. The vast majority of labs still use the retroviral transduction method to make iPS cells despite (1) its arguably compromised translational relevance and (2) dozens of different papers described non-retroviral methods that have garnered thousands of citations. We as a field now have a whole host of methods available to us to make iPS cells, but how useful are these? Apparently not particularly, despite the high impact nature of many publications describing new iPS cell methods. Some of these methods arrived with a big splash, but almost do not appear to work at all in the real world of a variety of labs trying them.

2. Every iPS cell line, even if genetically identical to others, is epigenetically and functionally unique, representing a serious roadblock to genetic studies and disease modeling. There is an inherent stochastic element to making iPS cells and to cellular reprogramming that makes every iPS cell line unique, even if they were derived from a shared, genetically identical parental cell line. This reality makes phenotyping iPS cells a challenge and complicates their potential as tools for disease modeling and for genetics studies. For example, if you want to map genetic elements linked to certain disease phenotypes using iPS cells, the inherent functional variability (probably epigenetic in its basis) of even genetically identical iPS cells could make your quest impossible. In addition, this inherent variability amongst iPS cell lines is a major roadblock to their use as disease models.

3. Real human iPS cells are not that easy to make. I hear all the time people asking what the secret is to making iPS cells, particularly human iPS cells. More recently people have been asking…..Do we need copper incubators? Do we need some special additive to the media? Do we need low oxygen?

Lowering the risk of fungal contamination, using epigenetic modifying or kinase inhibitory drugs, and employing more physiological levels of oxygen can be helpful, but these are not the critical factors. The most important thing in my opinion is the level of reprogramming factors relative to each other and the scientist who is doing the work being extremely fastidious and rigorous.

4. Most iPS cell publications are not really important. There are hundreds and hundreds of iPS cell papers out there, but in reality most of these papers do not address fundamental issues. Period. This applies even to some of the most highly cited iPS cell papers, most of which did nothing substantial to advance the field. Of course there are also quite a few gems out there in the form of amazing papers, but they are immersed in a lot of noise of all those other papers. Heck, I suppose you could say more generally that most papers in science as a whole are not really important?

5. iPS cells take too long to make. As a result  of this timing issue, patient-specific iPS cells generally will not be able to be used for patient-specific therapies for acute injuries (at least not for acute treatment) such as stroke, heart attack, acute brain injury, etc. iPS cells have great potential for autologous therapies, but they will take months to make and validate per patient no matter how much the technology advances. What this means is that patient-specific iPS cells just won’t work for acute injuries.

6. iPS cells are expensive. Yes, it is sad but true that the cost of making and validating iPS cells is too high for them to ever be widely used as patient-specific medicines. This is a huge issue, but one almost totally ignored. iPS cells are super expensive to make and validate even in pre-clinical studies. Such studies are far less expensive than the type of iPS cells that would approach clinical grade. Just how expensive are iPS cells? Roughly $20K per line just to get to first base. Want the home run of a clinical grade iPS cell line that the FDA would stamp “OK” on? Easily more than $100K per line. The idea of making an international iPS cell bank with so many common haplotypes that most people could take advantage of them for therapy is a noble idea that is in the works, but I don’t think anywhere close to reality.

7. iPS cells are unstable. Yes, iPS cells can literally change overnight in culture and they will change over time in culture. The longer you grow your iPS cells the more stable they become in some ways, but only because cellular evolution is occurring in the dish and you are selecting for more homogeneous populations that grow better. The frustrating reality of iPS cell instability and drift in culture is that unlike other cell lines, it may not be legitimate to compare data obtained from the same iPS cell line if the experiments were done too many passages apart. One could make the same argument about any cells, but iPS cells change more over time.

8. The iPS cell field is fairly cut throat. When we go to meetings such as ISSCR or others, iPS cell researchers sure seem chummy and cooperative with each other, don’t they? However, the reality is there has been and continues to be fierce competition amongst iPS cell researchers, many of whom seem to cluster into cliques. Again, this is probably not particularly unique to iPS cells, but could be applied to other fields as the Myc field that I work in or other areas.

9. A surprisingly high fraction of all published iPS cell methodological protocols are difficult to recapitulate. Again, this is not a popular thing to say, but it is unfortunately accurate.

10. iPS cells do not and likely will not ever replace ES cells. Do I need to say any more?

You might also be interested in checking out these 10 commandments of working with hESC that raise some related issues from a somewhat humorous perspective by another author.

7 thoughts on “Taboo topics about iPS cells: the elephant in the lab series”

  1. Could you provide a contrast to the cost of producing iPS cells to that of producing clinical grade hES cells. How much would these cost?

    1. Great question. Clinical grade hESC-based therapies have the advantage of being batch-prepared products that can be given to dozens of patients, greatly reducing relative cost. I’m not in biotech, but I would guess such a therapy would be in the $10K-$20K range. However, hIPSC, if used in a truly patient specific manner, would be far more expensive. They must be derived independently and validated for every patient. In theory such a product might have to be safety tested for every patient as well since one person’s iPS cells can be quite different than another’s. Such a clinical grade iPS cell product, for the company to even break even, would be in the $100K and up range I would estimate. Some folks are talking about cheaper allogeneic iPS cell-based therapies + immunosuppression, but that kinda defeats the purpose of iPS cells in my opinion. Others are talking about iPS cell banks that would be immunocompatible with most people in the world, but that seems, at least at present, highly impractical and we still don’t know if the current iPS cell production methods are optimal for clinical use. You could make a whole huge bank of iPS cells supposedly for clinical use only to realize after a few years and millions of dollars that your method for making those iPS cells is now out of date and not really appropriate for clinical use.

  2. Nobody needs a cooper lined incubator or copper shelves. UV in the incubator is similarly useless. Just proper sterile technique will do the trick. hES cells do grow better in 10%CO2 5%O2

  3. For those of us interested in medical research, but not medical specialists or researchers, it would be a courtesy to include in reports of acronym-identified entities a straightforward definition. What does “iPS” stand for?

  4. Although I think that most of what you say is correct, the hES cell field still has it’s issues as well. I’ve been working with iPSC technology for over 3 years and I’m very comfortable with the some of the new non-integrating (footprint fee) reprogramming methods. There are some “tricks” that I’ve learned to help produce iPSC lines, most of which I had to learn on my own. I think that iPSC will become more relevant than hES cell for disease modeling but there’s still a lot of work that needs to be done in order to ensure what you are modeling is correct.

Comments are closed.