I got in this morning and the place is buzzing with yesterday's news. So the big question is whether this technology will be useful for stem cell therapy?
One professor was willing to wager that these new findings will not lead to therapeutic stem cell therapy. In the lab, we were actually split down the middle on this issue. You see the problem is that cancer and stem cells share a lot in common. This technology involves introducing four transgenes (two of which are proto-oncogenes) into cells. This protocol may increase the likelihood that these induced stem cells turn into teratomas after they are implanted.
A major factor is how long do the transgenes have to be active. If those genes need to be on for only a short time, we could simply inject the protein and/or the mRNA into the cells. If the genes need to be on for weeks (it takes 2-3 weeks for the conversion) then there may be complications.
MarkH at denialism has a nice post on all these issues plus the potential impact of this work on the culture wars.
So what is the feeling out there?
- Log in to post comments
It's a big finding, a peek into how cells determin how to dedifferentiate. Whether it translates into some medical application, I think humans tend to overestimate our ability ... but then again we did send a man to the moon.
Keep in mind that these studies were done in mice, where there is almost 25 years of experience with ES cells and how to use them. We have less than 10 (severely restricted) years of experience with human ES cells, and Rudy Jaenisch will be the first to tell you that human cells are 'an order or two in magnitude more complex than mouse cells.'
It bothers me that so many commentators are conflating findings in the mouse to the human, where this or any other de-differentiation protocol has worked.
I can only guess this will throw a wrench into plans on Capitol Hill to overturn Bush's veto of the stem cell bill.
Well it looks like turning on the deprogramming algorithm may be slightly harder (it sounds like introducing the four factors don't really work and that other factors may be required) however the underlying program must be very similar, so I'm guess that it won't be that hard.
As for complexity, you must have missed the flurry of posts a couple of weeks ago, see these:
http://scienceblogs.com/transcript/2007/05/a_little_note_on_complexity_…
http://scienceblogs.com/pharyngula/2007/05/step_away_from_that_ladder.p…
and this compilation of links by coturnix:
http://scienceblogs.com/clock/2007/05/complexity.php
Big step, but really this is just an expansion of the results from paper from last year. Nice confirmation really.
I'd say it will take about 1-2 years to figure out how to do this with inducible expression of the 4 genes - especially myc, and who knows how long to decode the programming for human cells. That's a total crapshoot. But once that occurs, the ability to make ES cells from patients with specific diseases will be a godsend, with or without the oncogene problem present. I'm hoping these things can be accomplished within a few years, but it may be much longer. After all, it took about 18 years to get human stem cells after they were first isolated in the mouse. While our level of knowledge then isn't comparable to what it is now, the human systems always seem to find a way to throw a wrench in things.
In terms of therapeutic potential I think it ultimately will be possible. If it is just 4 genes you could conceivably make a single plasmid delivering them in a tetracycline or tamoxifen-controlled polycistronic cassette. Then using a phi c31 integrase system, plug them into a discrete genomic location that doesn't have a known transcribed gene without using a retrovirus.
Select for transformed cells and then voila - you eliminate the oncogene problem, the retroviral insertion problem, and hopefully still end up with cells.
It might be that we could learn to move the cells to a chosen determined state in vitro before implantation for therapy and that that could reduce their potential tumorigenicity.
I see where the complexity conversation could address my concerns. My point is that a main contributor to this field, arguably the leading expert, is so cautious about claims that this can be translated to humans, that I think it is inappropriate for the science media juggernaut (of which we are all a part!) to be presenting this as the next big thing for human tissue regeneration.
These sorts of claims (and the manner in which they are hailed) establishes too much of a battlefield of scientific competition.
For a more verbose argument, see a recent post of mine.
The findings are just terrific, and life is going to be good for a while for the authors of those papers.
But I'll put 10 bucks down and say that this won't be useful for therapy. I'll also stick my neck out and state that the 4 transgenes will need to stick around (at varying levels, probably cycling up and down in some kind of co-ordinated cycle) for about 2 weeks.
And I hope someone proves me completely wrong.
One question that I am thinking about since I've been working in dermatology:
How differentiated is a fibroblast compared to other cell types? Would the same work with kerationocytes or neurons? As far as I remember the work of one of my lab mates cultured fibroblasts display some degree of de-differentiation in the direction of their progenitor myofibrolast cell type. Maybe for fibroblast it is just a few steps back, while terminally differentiated keratinocytes would need many more steps.
Sparc,
Those experiments are being done right now - in fact we will soon know whether any mouse cell can be transformed into an iPS cell.