What does it take to turn a stem cell into a cure?

Blogging on Peer-Reviewed Research

Last week, I reported on this new breakthrough in stem cell research, in which scientists have discovered how to trigger the stem cell state in adult somatic cells, like skin cells, producing an induced stem cell, a pluripotent cell that can then be lead down the path to any of a multitude of useful tissue types. I tried to get across the message that this is not the end of embryonic stem cell (ESC) research: the work required ESCs to be developed, the technique being used is unsuitable for therapeutic stem cell work, and there’s a long, long road to follow before we actually have stem cell “cures” in hand. A review on LiveScience emphasized similar reservations. Seizing on this one result as an excuse to end research on ESCs would be a great mistake.

So let’s consider what it takes to turn a stem cell into a medically useful tool. One “simple” (we’ll quickly see that it is anything but) example is finding a cure for type 1 diabetes. We understand that problem very well: people with this disease have lost one specific cell type, the β cells of the pancreas, which manufacture insulin. That’s all we have to do: grow up a dish full of just one cell type, the β cells, and plant them back in the patient’s gut, and presto, no more diabetes (setting aside the chronic difficulty of removing whatever destroyed the patient’s original set of β cells, that is). Sounds easy. It’s not.

Turning one cell type into another cell type is a routine procedure in normal development: a good part of development is simply that, the process of induction, or using external signals to activate a particular pattern of gene regulation in a target cell. What we want to do is induce a generic stem cell to express all the genes and proteins found in a β cell — we want to know what the switch used by the embryo might be. If we know how the embryo normally switches on the β cell genes, we’ll just flip that same switch in our culture of stem cells.

Here’s our first problem: in normal development, we don’t go straight from pluripotent stem cell direct to pancreatic β cell. There is a simple molecular switch, all right, but it’s a switch to turn a restricted cell type, a pancreatic endocrine progenitor cell, into a β cell. Only the pancreatic endocrine progenitor cell has a receptor for the molecular signal (it’s in a state developmental biologists call competent). You could throw buckets of the signal, whatever it is, at the stem cell, but if the stem cell is not competent, it is blind and will not respond. So, clearly, what we need to find is the molecular signal to switch the stem cell into a pancreatic endocrine receptor cell, hit it with that, then expose it to the β cell inducing signal.

But wait! The embryo’s cells do not normally switch from stem cell direct to pancreatic endocrine receptor cell — it isn’t competent to receive the pancreatic endocrine receptor cell signal. We need to convert our stem cells into pancreatic endoderm cells first, so that they’ll be able to read that signal.

And so it goes, back and back. Development is hierarchical and sequential, and you have to carefully shepherd your cells through a whole series of steps in order to get to that final desired state; there don’t seem to be any shortcuts, or at least not any shortcuts that don’t create additional dangers. In order to understand how to turn on the β cell fate, we need to understand the whole developmental pathway that leads to that cell, and we have to recreate it in the dish. Here’s the normal developmental hierarchy:

The lineage of the developing pancreas in vivo and in embryonic stem (ES) cultures. The
left panels are mouse embryos at different stages of development (embryonic day 3.5 through 13.5
[E3.5-E13.5]), and the black bars on the right indicate the equivalent stages in human development.
Embryos are oriented with anterior to the left and posterior to the right. The E13.5 stage shows the
dissected stomach, pancreas, and duodenum with the stomach (s) and dorsal pancreas (dp) to the
left and the ventral pancreas (vp) and duodenum (d) to the right. The LacZ staining in the E9.5 and
13.5 embryos shows expression of Pdx1. The lower left panel is a pancreatic islet
showing glucagon-expressing α-cells (green) and insulin-expressing β-cells (red). The curved
arrows highlight several signaling pathways involved pancreas development that have been used
to direct ESCs into the pancreatic lineage. The middle panel indicates the lineage of the developing
endocrine pancreas. The arrowheads in the cell lineage diagram indicate two separate roles of
Nodal/Activin signaling: to initiate gastrulation and to promote endoderm vs. mesoderm fate in a
dose-dependent manner.

To make a β cell, you have to first convert ES cells into mesendoderm, then into endoderm, then into Anterior Definitive Endoderm (ADE), then into midgut endoderm, then into general pancreatic tissue, then into pancreatic endocrine cells, and finally, you can apply a signal to switch them into the β cell state. This is a non-trivial problem in a dish. In the embryo, the series of switches are applied to the cells as different tissues come into contact with them and as the internal environment of the embryo changes over the course of normal development — but all those signals are absent in a dish that only contains stem cells, so the researcher has to apply them sequentially and in the proper dosage manually. As the diagram illustrates, many of the signals are known—molecules like nodal and FGF and Notch.

Here’s the good news: researchers can do this! The yields aren’t great, and it takes a meticulously applied protocol, but human embryonic stem cells have been grown in culture and carefully steered down the developmental pathway to produce β cells that secrete insulin. An example of the steps and various molecular markers assayed is shown below.

(click for larger image)

Efficient differentiation of human embryonic stem cells (hESCs) into definitive endoderm
(DE). A: The ESCs are directed to become pancreatic endocrine cells over the course of 2-3 weeks.
By adding various factors at different stages, hESCs are first
differentiated into a mesendodermal progenitor and then into DE. DE then transitions through
stages reminiscent of primitive gut tube, foregut endoderm, pancreatic endoderm, and finally into
the endocrine cell lineage. By analyzing different markers (shown under each cell stage), hESC
differentiation is correlated with pancreas development in vivo.. B: hESC differentiation into DE. The hESC cell line H9 was cultured using standard
techniques. To induce differentiation, ESCs were cultured in RPMI+Activin (100 ng/ml) and low
serum for 3 days. Before differentiation, hESCs express Oct4
(image shows one ESC colony). After differentiation, >80% of cells are endoderm cells that
coexpress Sox17 and FoxA2 (these cells are a monolayer). A small number of HESCs differentiate
into mesoderm and express brachyury (TBra).

Notice that in order to do all of this work we need detailed knowledge of the normal course of pancreas development, and many labs and many experiments are required to work out all of the steps. β cells derived from human ESCs have been made in culture, though, and have even been transplanted into mice with type 1 diabetes, where they have rescued the mice from hyperglycemia. So the cure is around the corner, right?

Not quite. Efficiency and expense are a huge issue; none of these protocols are ready to deliver volumes of transplant-ready β cells. Another issue is that the transplanted cells aren’t quite right—sure, they express insulin, but they have other properties that are odd for β cells, and they might not be a stable, self-regenerating population. Another very serious concern brought up in my previous description is that scientists are working with pluripotent stem cells, and one common assay is to inject these cells into a host mice and observe the development of teratomas, cancerous tumors that contain many different tissue types. Some of the experiments that transplanted pancreatic cells induced in culture from stem cells had the side effect of riddling the host pancreas with teratomas … an ugly fate for a mouse, and one you definitely don’t want to risk in a human patient.

Being able to generate induced stem cells from human skin cells is an important step, but it’s only one step, and it’s also a step at the very beginning of the process; the political triumphalism that is being used to declare the ethical dilemmas resolved and the biggest problem in stem cell research completely solved are premature and wrong. We have a long way to go yet before medicine can deliver useful therapies from what is currently very basic research, and we also need ever more detailed understanding of the intricacies of development in the human embryo.

And note that what I described above is the path to solve a relatively simple problem, the induction of a single cell type. The reconstruction of whole organs and the repair of complex tissues is orders of magnitude more complex. It’s good to get excited about the science being done at every step, but we also have to be realistic about the immensity of the problem being addressed. Let’s not get so satisfied with the small but crucial answers being found that we lose sight of the grander goals, and shut down research in the mistaken idea that the big problems have been solved.

Spence JR, Wells JM (2007) Translational embryology: using embryonic principles to generate pancreatic endocrine cells from embryonic stem cells. Dev Dyn 236:3218-3227.


  1. Great White Wonder says

    Perhaps the best solution is to simply take cells from the person with diabetes, induce embryogenesis but ablate the chromosome regions that encode for a functioning brain, and implant into the embryo into a willing donor (who will be reimbursed, of course, unless it’s a volunteer or the diseased person’s own mother). At some point, this viable but brainless “pancreas pod” can be birthed and grown to whatever size is needed until the cells are harvested, at which time other organs of the pod can be used, after which the pod can be composted.

  2. Dustin says

    or the diseased person’s own mother

    Really? I’d think most of the conversations would go like this: “I already gave birth to you once, you brainless little shit, and I sure as hell won’t do it again unless you pay up. And why don’t you ever call? And when are you going to marry that girl you’re shacking up with? And why are you shacking up with her, why don’t you go out and find a nice girl? And…”

    I think just popping an insert your own embryo kit and a check in the mail to mom would be a great timesaver for a great many people.

  3. andrew says

    I think Great White Wonder fails to detect the reason there’s so much hubbub about this sort of thing in the first place. –;

  4. says

    Andrew– No doubt! I was waiting for the next line… “and the remaining embryo parts could be seasoned, dried, and used for jerky! Everybody WINS!”


  5. Tulse says

    Exactly, Great White Wonder — everyone should have their own bevy of headless clones ready for transplant donation. (Perhaps they can be warehoused hanging from the ceiling like that cool scene in Coma.)

  6. George says

    The true debate needs to change now. While these cells are not yet fully equivalent of ES cells much less fully equivalent embryos, we can see that the path may be there to convert an existing adult cell into a fully capable ES cell equivalent. Such a cell, could give rise to clone. Yes, this is some distance technically.

    But it begs the question as to whether such a cell is in religious terms the same as an embryo. Technically it could be. Then some or maybe eventually many whole adult cells could be made into an embryo, thus it must have the same status as an embryo at all times. Thus, biopsies must be banned. The logic is inescapable and shows the absolute folly of the issue with ES cells in the first place.

  7. bluthetan says

    damn you PZ, I read your blog to keep informed of the latest balderdash and flapdoodle that’s vomited up by right wingnut delusional xtians and other freakish cults. Much to my surprise I have managed to learn a tiny bit of information concerning the induction process, stem cells being competent or not, and about non trival problems in a petri dish. A most informative post. Keep up the excellent work. Highest Regards blu

  8. Schmeer says

    Hmmm… Great White Wonder, I’m afraid we’re going to need to have a look under your bed, in the closet, basement and garage. Please don’t make any travel plans and do let us know where we can reach you.

  9. says

    “Then some or maybe eventually many whole adult cells could be made into an embryo.”


    Is this satire?

    Do you honestly believe you can ‘assemble’ a multicellular, multiorgan organism from ‘parts’? I don’t know where you live, but I live in a place I call Reality. I think it is a very different place from Science-Fiction Land.

    PS- can someone give me a link that will teach me how to do HTML tags? I’d like to jazz up my posts with some italics and strikethroughs and such and such.

  10. says

    “…everyone should have their own bevy of headless clones ready for transplant donation.”

    Please no. I have enough trouble figuring out how to store everything I own now. I have no idea where I’d put a headless clone — maybe that spot in the basement? No that’s too close to the cat litter.

  11. Spaulding says

    Exactly right, George. You’d have to not only ban biopsies, but also donating blood, scratching, spitting, and menstruating. Heck, the act of popping a zit probably destroys thousands of potential human lives if you take reproductive cloning into account.

    The idea that human worth is determined by a little piece of invisible, immortal magic ( which we each carry around from conception ) is not a valid stance in a bioethics debate.

    At best, “soul” is a poetic, metaphoric reference to subjectively experienced, gestalt neuron activity. At worst, it’s a construct of wishful thinking in response to worldly injustice and fear of death. Either way, it’s not precise or real enough for a debate about specific medical techniques or legal status.

  12. Christian says

    But Katherine Kersten proclaimed that this solved the problems with stem cell research and that we have president bush and his “brave moral stand” to thank for it. Are you saying that conservative bloggers can’t be trusted?

  13. Jason F. says


    Anytime I take things like this and relay them to whatever conservative Christians I encounter, they almost always respond with, “But embryonic stem cells have totally failed to produce any treatments for anything, whereas adult stems cells have”.

    Is that at all true? Are they likely just referencing that list the conservative Congressman pasted together a couple of years back?

  14. says

    Christian, you can also add my new favorite “MD” shill for the Right, Beverly Nuckols to that list. If we’re really lucky, maybe Ms. Nuckols will grace us with some of her stellar insight into biomedical research on this thread!

  15. AtheistAcolyte says

    It’s funny that Great White Wonder brings (trolls?) that up, we essentially do the same thing with animals (pig-valve transplants), who have significantly larger and more-functional brains than the proposed “pancreas pod” (indeed, perhaps even a fair number of now-living humans). In addition, we can eat the animal corpses, rather than simply compost them. Waste not, want not.

    Actually, I’m feeling a new tingle down my Evilutionist spine: Why don’t we induce identical twins in each pregnancy, then chain up one of the twins (at random) and use them for slave labor and eventual organ harvesting? Ooooh, how deliciously eeeevil.


  16. Natasha Yar-Routh says

    Ok revealing myself for the hopeless geek I really am, I one version of the history of Superman’s home world of Krypton it was devastated by a war over the rights of clones who were created to be organ banks for the rest of the people. That war being the reason for the whole sterile crystal motif in the movies and later books.

    OK, go back to talking about real things again.

  17. Darb y says

    I don’t understand why Type 1 diabetes is often trotted out as an example of stem cell application, when it has probably the worst potential – even stem cells developed from the individual’s own cells will, once differentiated and active, express the markers that produce the autoimmune reaction that caused the condition in the first place!

    And that sort of rejection is the main longterm issue with most embryonic applications as well – undifferentiated cells don’t tend to set off a response, but they have to differentiate to do their jobs, and then they’ll be foreign cells.

  18. CJO says

    See, those damn things are piping up already!

    Next it’ll be LOLPancreas Pods, and the jig is up.

    I CAN HAS LO11y?

  19. DLC says

    Thanks for the interesting article on a topic many people consider “too complex” for mere mortals. (end of backhanded snark at the D.I. )

  20. Sili says

    Darb y beat me to it.

    It would seem to me (as a complete outsider) that we have a better hope of curing diabetes by discovering what/how/why those β-cells in the first place. Then mayhaps kids could be screened at birth and ‘fixed’. In that case we’d only need to grow new pancreases (pancreares? D. Marjanović?) for the interregnum generation.

    So until we solve the problem of that sorta autoimmune diseases growing organs would seem the only use of stemcells.

  21. LisaS says

    PZ- Thanks for the nice post. Very interesting. Like bluthetan (#9), I come to Pharyngula for reasons other than science, but have been learning some fascinating information in the process. Today I certainly didn’t expect to be learning about induction when I woke up!

  22. says

    Eh, that’s what embryoid bodies are for, you don’t need whole embryos, or embryos without brains people. All you have to do is aggregate ESC into embryoid bodies and they’ll make most anything you want. Then you just purify it.

    These guys for example, got a pretty good beta cell purification regime going on.

    The other great therapy that would be killer would be hematopoietic stem cells. We already have the method in place to generate them, they can be purified by fluorescence activated cell sorting, and you only need a handful of cells – like 10-20 – to repopulate the entire marrow.

    You ask me, that’s the most promising by far. The pancreas is a tough, stupid, difficult to access organ. It doesn’t have a resident stem cell population. Further, diabetes represents a continuing immune attack, so implants would require immune modulation in concert with the beta cells. Possible, but dangerous.

    The blood on the other hand is very amenable to transplant, and since you’re replacing the immune system, a great deal of the previous programming gets reset.

  23. Rjaye says

    Yes, yes, yes. This was the type of info I was waiting for when I first read the article in the paper, and later saw on the evening news, about stem cells developed from skin cells.

    Now I wonder if a person’s own skin cells, if used, are developed into the cells needed for that person’s transplant, would it still cause an immune system response? Or is that still an unknown? Does it still cause teratomas in mice?

    While there are still many questions and research to be done in this area, I was so glad to finally see a story on real research in this area, instead of maybes inferred from more basic studies. While transplantable tissues are way off, this has been a fascinating story to follow.

  24. Peter Ashby says

    One thing PZ didn’t mention in relation to the figure showing how to turn human ES cells into ß-cells is that many of the factors shown under each step are not signaling molecules that can be applied to cells they are transcription factor genes, the ones like Sox-2 and Pax-4 are proteins that sit in the nucleus and turn other genes on and off. In the work shown extra copies in dna constructs would have been inserted into the cells. This is not really the way you want to go for a therapy. So a usable, safe therapy is even further away as we need to find chemical/hormonal ways of doing the same thing. More work for the high throughput screening factories I think.

    My guess at what is going wrong in not locking these cells into being ß-cells is their chromatin isn’t right. What transcription factors that set cells on developmental pathway do is modify the dense packaged bits of dna that will need to be opened so that it is and will remain open. The bits of the genome that will not be needed also get packaged up and shut down. One or both of those processes doesn’t seem to be getting done in these cells so they are not what we call terminally differentiated, at the end of a one way street.

    My further guess is that it will take as yet non appreciated signals from the all important cellular environments during development that will be key to this. It is fairly easy to notice the signs that a cell is proceeding down a developmental pathway. We are not good at telling how firmly it is committed to it.

  25. Peter Ashby says

    Khan do a google for: growth factors cartilage regrowth

    and you will get a whole slew of results, patents, trials in veterinary medicine as well as human. Things like FGF4 and IGF-1 are being used now to achieve this. It would certainly be an alternative solution to what I had done in my hands, joint fusion.

  26. alchemist says

    To PZ (or anyone else who can answer this).

    I a chem pHD who finds this all fascinating, but biology is out of my league. I’m just looking for some answers that have come up in debates…

    Does anyone know the functional difference between the use of embryonic stem cells, umbilical stem cells and adult stem cells?

    My guess would be that ESC would have greater plasticity, and better conform to the new tissue. However umbilical might have create fewer transcription errors than stem cells from adult tissue. Anyone know if there is research on this subject?

    An argument I’ve been getting lately is that Embryonic stem cells cause cancer but Adult stem cells do not.

    Obviously this post rejects that argument. However, I’m curious if the cancers are similar, or if ASC and ESC have entirely different problems that exist when implanting them into tissue.

  27. Peter Ashby says

    Alchemist your questions are good ones, I shall attempt an answer.

    The differences lie primarily in their capacity to differentiate into different cell types. ASC have very limited types of cell they can become usually limited to at least one tissue and often only one cell type. The definition of an ESC is that it is Totipotent, meaning it can make ANY cell type in the body of the animal from which it comes, given the right environments and cues.

    This also relates to their cancer risks. ESC make embryoid bodies and teratomas because they can, no ASC can make the range of tissue types you see in these tumours. As you move down the developmental/differentiation pathway from an ESC you begin to restrict the range of cells and tissues you can make. The first choice that PZ highlighted in the first figure is between ectoderm and mesendoderm. That is a bigger change than you might think. All of our nervous systems are made from ectoderm as well as things like all the melanocytes in your skin. So that first decision cuts out an awful lot and you subset the possibilities at each branch point.

    USC are what is termed pluripotent (perhaps), in that they can make many cell types. But you will not get all out of them.

    As for stem cells not ‘causing’ cancer that depends on where you put them. The cellular environment is a big constraint on what cells can do. Take an ASC and put it somewhere it shouldn’t be, or not take care that it can only be where it should be, and the brakes to proliferation and growth can be removed resulting in tumours.

    This is also why many cancers only become dangerous when they metastasise, iow they escape their origin and lodge colonies in new places in the body, where those brakes may not operate.

    And we are also realising that cancers have their own stem cells and that it comes back after radio or chemotherapy because those treatments only kill the non stem cancer cells. So we need to find ways to kill cancer stem cells, preferably ones that don’t also kill the ASC population they resemble, a tough ask.

    Let me know if you don’t understand any of this.

  28. alchemist says

    I got enough. Thanks. I’m bridging into academia now, so I’m going to have to go back and take more bio courses on the side.

  29. says

    BaldApe: Thank you kindly!

    MarkH: EBs… Eh… Yes, I fully agree with you that withdrawing LIF from ES cells and letting them differentiate will robustly generate many different types of somatic cells: in my own hands, using R1, E14Tg2a, and AINV (see Daly & Kyba) ES cells I have generated primitive erythroid, endothelial, cardiomyocyte, and myeloid cells. The problem lies in purifying them . EBs are incredibly heterogenous and FACS isn’t always reliable: you can find CD31/Pecam on endothelial cells… as well as undifferentiated ES cells and monocytes. Similarly, what surface markers are on Cardiomyocytes? Yes, you can purify these cells with a MHC- or cTNT- or Nkx2.5- promoter driving GFP, but that would require insertion of that transgene or a knock in.

    Don’t get me wrong: I think the ES-EB system is incredible. I used it for Aims I and II of my thesis as well as my Development paper. We’re not at the point where we have defined the ‘cocktail’ that reproducibly gives high yield of a number of cell types and there seems to be disagreement in the field for a number of these factors, i.e., DKK in generation of the cardiac lineage.

    Best of luck on your defense! I haven’t scheduled mine yet but I am getting close; probably March or April.

  30. says

    This is an interesting read, especially for someone like me with Type 1 diabetes.

    I won’t be holding my breath. I think the work that Dr. Denise Faustman is doing holds more hope (maybe not much more) at this stage.