Ping-pong studies of signaling pathways in hematopoietic stem cells
In under discussion | 2 comments | permalink
Stem cell research right now is everything about assays and models. Recent published studies of a role of hedgehog signaling pathway in hematopoietic stem cell (HSC) self-renewal is one of the best examples. After promising study, published in Nature earlier this year, new data suggest that targeting of hedgehog signaling pathway can not be beneficial for treatment of leukemia.
It reminds me of a game of ping-pong (aka table tennis) when different labs are studying the same pathway but using different assays getting opposite results (gene is dispensable - indispensable).
I’d like to point out some “ping-pong studies” of signaling pathways regulating maintenance of HSC below (references indicate scientific group and link to the paper):
| signaling pathway |
indispensable |
dispensable |
| notch |
David Scadden Blood 2002 | Freddy Radtke Blood 2005 and Warren Pear Cell Stem Cell 2008 |
| N-cadherin |
Linheng Li Nature 2003 | Sean Morrison Cell Stem Cell 2008 |
| beta-catenin |
Tannishtha Reya Cancer Cell 2007 David Scadden Cell Stem Cell 2008 |
Freddy Radtke JEM 2004; Werner Held Blood 2008 |
| hedgehog | Tannishtha Reya Nature 2009 | Iannis Aifantis Cell Stem Cell 2009; Gary Gilliland Cell Stem Cell 2009 |
Can you imagine - working hard day and night (let’s say 10-14 h/per day) and spending the best years of your life in the lab (20s -30s) in doing experiments, suddenly you hear the news that another lab got completely different from yours results. Where is the truth? That’s exactly what happened with hedgehog (HH) signaling studies in HSC. One lab used a vav-Cre conditional knockout mouse as a model and showed that HH signaling is essential for HSC self-renewal, other labs - Mx1-Cre model and showed that HH is dispensable for HSC.
I am really puzzled by “ping-pong” studies. How will you know that your model is correct before you invest 2-3 years of your time in the project? Tomorrow scientists will create another model and disprove your results. If we have so much trouble to find the truth using mouse models, how will we translate our results in the clinic? Do you think we need more science to do or is it too much?
email this | tag this | digg this | trackback | comment RSS feed
vav-Cre more specifically for hematopoietic cells
Mx1-Cre a little bit generally in bone marrow tissue, like MSC.
It might be related with the contrast results.
We had a similar situation with Wnt in MSCs. One group (DeBoer) used Li ions to inhibit GSK3beta, saw increasd proliferation of cells and determined that, as in many cancers, Wnt promoted cellular proliferation.
Our group came along and showed that Dkk1, a Wnt antagonist, promoted proliferation and also inhibited differentiation. However, we also found that Li ions inhibited differentiation. We couldn’t resolve this until we found a more specific inhibitor of GSK3beta, which supported our results. Now you can find papers about the “biphasic” effect of lithium, depending on dose.
Certainly, pong-pong results happen all the time. This is just an indicator of possible outcomes and illustrates how you should set up your experiments to consider both of the ping-pong possibilities, not just the evidence for one side, regardless of how strong it is, on its own.