STAP NEW DATA

This is a crowdsource page for people who want to post their findings on their attempts to validate the STAP stem cell ( STAP細胞) method.  I’m going to color successful or even moderately encouraging reports green and failures/discouraging results in red. *For those with longer attention spans, see some additional important considerations for this crowdsourcing effort at the bottom of this page.

Please if possible email me ([email protected]) actual data (e.g. photomicrographs, etc) and I’ll include it. I’m encouraging people to include their full names, but not requiring it.

Update on 3/24/14.

Elevent reports so far…

Felix wrote on 3/24/14

We have been repeating the experiment with MEFs from OG2 mice. We indeed observed a significant increase of GFP positive population 4 days after acid (pH5.7) treatment, however almost all GFP positive cells are also showing strong red/blue fluorescence under the microscope, indicating the presence of intense cellular autofluorescence after treatment.

Ray and Sandy wrote on 2/19/14:

We used MEF and treated in HBSS at normal pH or titrated to pH5.7 with HCl as described in the manuscript in suspension culture (~5×10^5 cells in 2mL  DMEM/F12 media + 2% B27 + LIF). On day 1, some of the MEF formed small clumps in suspension, but this happened with and without low pH treatment. On day 7, the cells were plated down onto slides using cytospin, fixed/permeabilized and stained for Oct4. Cells in all conditions appeared Oct4 negative (results attached).  Please note that the culture medium was not changed during the 7 day incubation period as it was not mentioned in the manuscript.

Ray and Sandy 1 STAP stem

 

 

Ray&Sandy 2 stap stem

Dr. P wrote on 2/17/14 about a mixed bag of results:

STAP Experiments we have tried:

1. Rat Spleens from 6 month rats, LSM separated lymphocytes, treated exactly as published, cultured in DMEM/F12,2%B27,Lif. Oct4 detected after 7 days by western Blot.

2. Repeat of the above is underway.

3. Spleen from 35week Oct4-IRESGFP mouse ( GFP knocked in to endogenous Pou locus, on day 4 as it stands.

IF you don´t sort the Oct4-GFP spleenocytes, you will get GFP positive cells, but very few at the beginning of the culture. If after 7 days I see a significant increase in GFP, I will sort them, and take picture.

Side by side with VSELS I have in the incubator, you can´t tell the different visually.

I will post pictures when I have a complete data set.

We also tried MEFs and my own lymphocytes, but these experiments we used regulare iPS/Lif media while waiting for B27—–these experiments produced no oct4.

Update from Yoshiyuki on 2/13/14. Suspension culture (100,000 cells/ml) pH5.7, incubation time 25 min. See images for various conditions below. Note from Paul–Yoshiyuki now reports this green signal is determined to be autofluorescence. Bummer.

STAP 1st Try V2

Andres wrote on 2/13/14 regarding a STAP experiment in progress at 8 days out using human fetal fibroblasts. See images below. Spheres generated (left), but seem fibroblastic in nature as do cells migrating out of the spheres once plated down either on geltrex (middle) or on feeders (right).

PH STAP test

Yoshiyuki Seki on 2/13/14 wrote:

We used mouse embryonic fibroblast derived from Nanog-EGFP Tg.
Now we are performing resuspended culture in B27 + LIF or serum + LIF for 4 days. In B27 + LIF medium, we can’t detect GFP-positive, while we can detect weak-GFP positive cells in serum + LIF. However, we observe many dead cells in GFP-positive clump. Therefore it might be difficult to expand clump of GFP positive cells in adherence culture.

Nanog GFP STAP

Hong wrote on 2/12/14

I have try the pH=5.62 DMEM Media to treat Mouse ES cell (2nd passage) for 25min, centrifuge 1000rpm/5min, raise them in B27 and Lif Media for 6 days to detect the endogenous mOct4(RT-PCR) but nothing no matter with/without treatment.

Sasha wrote on 2/12/14

We have tried to generate STAP cells from mouse embryonic fibroblasts, mouse adult neural stem cells and mouse embryonic neural stem cells. We did not observe Oct4-GFP reactivation from either cell type after the exposure to pH5.7, pH 4.7 and pH3.2 at different cell densities (including 100,000 cells/ml described in the paper). We have also tried gelatin- and laminin-coated tissue culture plates as well as low-attachment dishes without a substrate. All with no Oct4-GFP reactivation after 7 days of culture (assessed by flow cytometry). B27 and LIF were used.STAP stem cell attempt flow

SK wrote on 2/12/14

Try (1) Neonatal whole blood = failure
Try (2) Adult whole blood from three animals = failure
Try (3) Adult spleen cells from three animals = in progress. at day 2 most cells dead, some cells aggregate.biggest problem i keep running into is to keep the pH the same over 30 minutes. The more cells die in the low pH conditions, the more rapidly does the pH change. B27 and LIF were used. SK

Elliott Schwartz wrote on 2/11/14:

A post-doc in my lab tried the acid treatment with human fibroblasts plated in mTeSR. 2 abnormal-looking groups of cells were picked, but nothing seemed to happen to them afterwards. I’d say try #1 was a failure

Ruben Rodriguez wrote on 2/11/14:

The pictures shown in the NewScientist article resemble mouse es cells… and human es cells in its primed form don’t look anything like that; maybe they use naive conditions, but somehow I doubt it.
Also, I’m one week in with my own pilot test with human neonatal fibroblast, and the combination of acidic shock and serum free media makes the cells to form aggregate, but a live staining indicates that they are not Tra-1-60 positive… and other than that; the culture looks pretty boring, no MET or anything interesting going on. It’s still possible that they need more time, but based what on what I’ve seen so far I’ll put a big red flag on this one. Picture at left is control pH 7.0 and picture at right is pH 5.6.  Ruben Alvarez Rodriguez, PhD, The Salk.stap try pH 5.6STAP try pH 7.0

Ethan said on 2/10/14:

It has been more than a week now, have yourself or someone you know of been able to reproduce the mouse STAP cells. We tried, didn’t work for the first time.

 

* Note (from Paul) that the results posted here are not necessarily endorsed by me or reflective of my own views. I would suggest that blog readers not take any one result too seriously and rather look for patterns. Also, I wonder…could there be a bit of a bias in this crowdsourcing toward negative results because those who get positive results, assuming there are positive results in the mix, may be more inclined to try to publish it in a journal versus in this domain? I don’t know.

134 thoughts on “STAP NEW DATA


  1. Question to you and others- what is required to prove it’s not true? People still defend older sketchy stem cell papers that could not be reproduced, as exemplified from some commenters on this blog. What if the big labs say they don’t see it, what do we do?


    • This is a great question, but a tough one to answer. In theory in science it is difficult or nearly impossible to absolutely prove a negative thing like that STAP cells are not real. However, at a practical level if say 90% of labs who try the STAP method cannot get it to work and about 10% of the others get iffy results, then I’d say it’s unlikely to be factual. Again back to your point though it is possible that some folks may nonetheless still believe in STAP even if most people do not. This is the case with VSELs and MUSE right now. On the other hand, if STAP works for most who try to recapitulate the results I think we can safely say it was proven to work.


      • I am not a scientists, but surprised see the comment coming from supposedly a scientist.
        I understand that Dr. Obokata found the process taking place when stress was applied to the cells, which led a series of trial of different stress with varying rates of success. It finally ended up with that level of pH.
        It indicates that, though the mechanism stayed throughout the trial, it is a delicate and complicated process. Even if only 10 % or even 1 % succeeded in “reproducible”, the next step to take is to carefully study common factors for the successful experiments and further testing incorporating them. The process continues, but no one will be forced to participate. Those who are sincerely concerned with the possibility of the authenticity of the findings will pursue.
        The research by Dr. Obokata is at that stage and shall welcome all the comments that help their follow-up.
        Please do not be offended with the comment by non-scientist, who is only excited about the finding.


        • Dear Katsuaki,

          It is hard to say but Ms. Obokata’s finding may false-confirmed. Univ. Yamanashi failed to reproduce although the person was in the site. What is difference ? Same person, same method could not show its success. Something wrong in Riken, or their reagents (maybe contaminated, and in this case, fuge new finding could be arise) used. We commonly see decrescence of pH upon cultivation (originated from CO2 and organic acids), but had no reports on this kind of findings.


  2. A post-doc in my lab tried the acid treatment with human fibroblasts plated in mTeSR. 2 abnormal-looking groups of cells were picked, but nothing seemed to happen to them afterwards. I’d say try #1 was a failure


  3. It would be great if you encourage people to actually post data; as it stands the comments stating “yes it worked” or “no it didn’t work” seems hardly rigorous.


  4. Re: the previous two comments of not being able to replicate the findings.

    What those people did was not replicate the experiments, it was try slightly different experiments with the assumption that the protocol would still work.

    Replicate= actually following the same protocols as published in the paper!!


  5. Hi all – not sure if it is just me, but I’m struggling to get the RNAseq/ChiPseq data from the second STAP paper. When I go to Biosample on NCBI (http://www.ncbi.nlm.nih.gov/biosample/), and query one of the accession numbers (eg. SAMN02393426), all I seem to get is the sample meta-information, but not the data. Any thoughts would be a great help.


  6. Try (1) Neonatal whole blood = failure
    Try (2) Adult whole blood from three animals = failure
    Try (3) Adult spleen cells from three animals = in progress. at day 2 most cells dead, some cells aggregate.

    biggest problem i keep running into is to keep the pH the same over 30 minutes. The more cells die in the low pH conditions, the more rapidly does the pH change.


  7. Has anyone contacted the lab that produced the data and asked for a visit? I know a number of techniques that need to be learned in the lab where they were developed.


    • Are you saying the you know the lab was deficient jn certain techniques?

      Which ones?

      Or are you saying the lab knows techniques other labs do not?


  8. We have tried to generate STAP cells from mouse embryonic fibroblasts, mouse adult neural stem cells and mouse embryonic neural stem cells. We did not observe Oct4-GFP reactivation from either cell type after the exposure to pH5.7, pH 4.7 and pH3.2 at different cell densities (including 100,000 cells/ml described in the paper). We have also tried gelatin- and laminin-coated tissue culture plates as well as low-attachment dishes without a substrate. All with no Oct4-GFP reactivation after 7 days of culture (assessed by flow cytometry).


  9. I have try the pH=5.62 DMEM Media to treat Mouse ES cell (2nd passage) for 25min, centrifuge 1000rpm/5min, raise them in B27 and Lif Media for 6 days to detect the endogenous mOct4(RT-PCR) but nothing no matter with/without treatment.


  10. I like the idea of this crowdsourcing blog regarding preliminary results of STAPs. However, it would tend to bias towards posts of negative results so early on as those that get any positive results in this gold rush will want to keep to themselves for the paper rush. I am of course also trying this ‘Stem Cell Ceviche’ method in various forms , but know not to report anything publicly, especially positive, until personally convinced.


    • Thanks for the comment, Steve.
      Yes, I agree with you on the potential bias and I mentioned as much at the very bottom of this page. One problem comes, however, if say 95% of people cannot get it to work. Probably few to none of these folks will be able to publish their negative data in a journal. So in fact the bias can and likely will go the other way for journal articles on STAP–they will be quite biased toward positive results for STAP…that is assuming there are any.


      • You are correct that negative results tend not to be publishable. Also, SPOREs and VSELs, which these are being compared to, have had a very controversial past. However, we labs who reprogram all the time and have the right tools, expertise and regents to verify should in the short run (weeks) either be able to or not be able to reproduce, even if ‘reproduce’ only means a few Tra 1-60+ positive human cells to show proof of principle.


        • Steve: we wouldn’t withhold positive results thinking that we’d publish them- it’s just as tough to publish “replicate” studies as it is to publish negative results. That said, wouldn’t it be great to have this crowd- this unofficial consortium – publish on their vast array of results?


          • Steve,
            I agree with Jeanne. I think it will be difficult to publish even “positive” STAP data in journals. As to negative results? Forget about it. Who’s going to publish that? I am not saying journals should avoid publishing negative results by any means, but editors don’t get excited by negative results and especially in the context of replication experiments.

            Also journals would likely require teratoma assays, which can take months, possibly lengthy mouse germline contribution assays, perhaps extensive RNA-Seq, genomics, and other not necessarily so quick to obtain data…even for an “expert” lab. You may be able to get a good picture of whether you think it is working or not within weeks, but that doesn’t mean you can publish that in a journal.

            As to the wisdom or lack thereof of putting data out there in the public domain before and/or instead of going to a journal, related to your comment “…know not to report anything publicly, especially positive, until personally convinced.”, I can see two sides to that. Making big unpublished claims directly to the media seems risky for many reasons, but on the other hand sharing preliminary data in the context of crowdsourcing with the understanding that it is just preliminary data, seems like a positive thing to do.


              • True, but I am pretty sure anything Irv Weissman submits will be published due to his high credibility. It’s not as easy for the rest of us!


              • Hi Pat,
                You are right. Publishing negative data on a replication of an already published study can happen (e.g. Weissman’s paper on VSELs), but there is definitely more challenges to that. Also in the STAP case what can people publish really if they believe they’ve found that STAP isn’t real? Some pictures of cells NOT glowing green with a Oct4-GFP reporter? Pictures of qPCR of those cells not expressing pluripotency genes? Data of the cells not forming teratomas? I can already feel the journal editors yawning in boredom. On the other hand, maybe some journals will agree to publish such negative results because the papers might be end up highly cited? It’s going to be really interesting to watch it all play out.


            • What I meant is that I would not want to publically put out half-baked data on the internet just to invite all the obvious “did you do this?, did you do that?, you should have done this, etc”. Posting on the internet for the whole world to see is not like showing data at your lab meeting or to your colleagues…


  11. As the stem cell data manager in Lifemap Discovery database http://discovery.lifemapsc.com/
    we are debating whether or not to to add the STAP stem cell protocol to our huge collection of stem cell differentiation protocols.
    This page is very helpful to evaluate this novel technique and whether we are on the dawn of a new era, the STAP stem cells era, or not…
    Thank you Paul for this important initiative! Hope to hear your comments about our stem cell database, http://discovery.lifemapsc.com/.


  12. We used mouse embryonic fibroblast derived from Nanog-EGFP Tg.
    Now we are performing resuspended culture in B27 + LIF or serum + LIF for 4 days. In B27 + LIF medium, we can’t detect GFP-positive, while we can detect weak-GFP positive cells in serum + LIF. However, we observe many dead cells in GFP-positive clump. Therefore it might be difficult to expand clump of GFP positive cells in adherence culture. If we will gets final result, we will feedback to this web site.


    • Hi Yoshiyuki,
      do you see same fluorescence signal in wt cells, i.e. without a GFP reporter?
      The point is whether this is auto-fluorescence of apoptotic or otherwise affected cells.
      Thanks
      Leonid


      • We don’t perform FACS and use wt MEF. It is possible that GFP signal is auto-flourescence. Therefore, we are trying to transfer these clumps to adhesive plate in ACTH + LIF medium.
        Thanks.
        Yoshiyuki


        • Yoshiyuki,

          What fragment of ACTH did you use in your medium? And at what concentration?

          Additionally, what type of HBSS did you use for the low pH treatment? With calcium and magnesium or without?

          Thanks in advance,
          Sarah


        • Jeanne: Thank you for your advise.
          Unfortunately, because I can detect Red channel, this signal is autofluorescence. I will retry. Thanks


  13. It would be very interesting to see data on the surface marker expression profile as these cells are undergoing the first stages of reprogramming. Its promising that some investigators are seeing is Nanog-GFP and live Tra-1-60 expression but are the fibroblasts losing CD13 and gaining SSEA4 and Tra-1-60 double positive staining at all?


  14. Steve:
    I think it’s great that people are posting their one-off experiments. I actually see this as a big international lab meeting- I wish we could do this whenever there’s a new technology that doesn’t work as advertised. I don’t know how others feel, but I think that sometimes we need to share information for the good of the scientific community. Think of all the wasted time and money if we were all to test this secretly…
    Jeanne


  15. Hi Paul, well this blog is going great, so glad to see the lively discussion! I agree with Jeanne, I think we are pulling of a labmeeting-without-borders to some extent.

    Since you are collecting so many questions, opinions and experiences from researchers trying STAP, do you think you could reach Vacanti again for an interview to address the progress on this story?

    Thanks for the good work!


    • Thanks for the positive feedback.
      I have been continuing to email back and forth with Dr. Vacanti. Our communications have been cordial and he has answered most of my questions. I imagine he is a very busy man right now. When I have been planning to do a new blog post, I’ve often let him know in advance and/or asked him questions to try to clarify things. I appreciate his openness. I may ask him again for another interview as the STAP story continues to progress. Note that I’ve also invited Dr. Obokata to do a brief interview, but I never heard back from her.


  16. A scale bar in Fig 2A in Obokata’s Nature Article is very interesting. Nuclear size is VERY different among figures in Fig 2A, but there is only one scale bar.


  17. One thing always confused me is that in the STAP nature paper, the author never told us how many mice she used to get stap cells. The other thing is that she also did not specify the sex of the sacrificed mice. And then She just injected the stap cells into the blastocyst and all mice grew up normally. My confusion is that if she injected a male mouse stap cell into a female blastocyst, how can she expect the mice grow up perfectly normal?


    • In the traditional gene targeting method, male ES cells are injected into unsexed blastocysts. As long as the male ES cells contribute to the germline, the resulting chimera will usually develop as a normal fertile male mouse.


  18. STAP Experiments we have tried:

    1. Rat Spleens from 6 month rats, LSM separated lymphocytes, treated exactly as published, cultured in DMEM/F12,2%B27,Lif. Oct4 detected after 7 days by western Blot.

    2. Repeat of the above is underway.

    3. Spleen from 35week Oct4-IRESGFP mouse ( GFP knocked in to endogenous Pou locus, on day 4 as it stands.

    IF you don´t sort the Oct4-GFP spleenocytes, you will get GFP positive cells, but very few at the beginning of the culture. If after 7 days I see a significant increase in GFP, I will sort them, and take picture.

    Side by side with VSELS I have in the incubator, you can´t tell the different visually.

    I will post pictures when I have a complete data set.

    We also tried MEFs and my own lymphocytes, but these experiments we used regulare iPS/Lif media while waiting for B27—–these experiments produced no oct4.


    • Dear Dr. P,
      I am slightly concerned about your positive control: VSEL cells. The very ones, whose pluripotency is just as controversial as that of STAP cells? Would it be possible for you to use common ES cells instead?
      Best regards,
      Leonid Schneider


      • Hi Leonid,
        The VSELs were only a visual size control looking through the scope. I make no claims as to any functional similarities. IF-I can make STAP cells, THEN I will compare the two with our collaborators. According to the STAP paper, STAP cells only proliferate in ES STEM CELL media, VSELS do not, so there is your first difference:)

        Cheers,

        Dr. P


    • What happened to your experiments? You seem to be the only one so far to have had some ‘success’…I know it may take some time to analyze or compile results, I am just curious.


      • I have repeated the the westerns three times now and I get Oct4 positivity from the rat spleen cells treated as described. I sent the blog administrator one of the blots and I requested that it not be published yet. I am repeating the procedure again on the rat spleenocytes. If I get the same result, I will go public here with it.

        Additionally, I have also run the protocol on spleenocytes from 35 week old Oct4-GFP mice. I did not sort prior to low ph treatment (i do not have a sorter on hand, so I had to order magnetic separation columns) Both untreated and treated cultures show GFP expression from d0 until d9.
        On Monday, I hope to be able to sort out the CD45+ and then analyze the CD45- fraction. I will report back.


          • what do you mean by “activate the spleen cells?” All I did was mince and enzymatically dissociate the spleen, passed through a cell strainer (0.44uM), treated with HBSS Ph 5.7-30″, spin 5minutes, plate in the media as published. I used B27 without vit A.


            • Hi – I queried activation state of your cells because splenic T lymphocytes in adult mice remain in Go “resting” state of the cell cycle unless activated through the TCR. Activation (e.g. by conA) induces synchronised entry into G1 and release of LIF.


        • Dear dr. Debs,

          I’m very glad to hear of your results. It is also very helpful that you are sharing your success while we wait on details from the authors.
          Without giving away too much, could you respond to some of the gaps in the paper based on your experience?
          - When you see a GFP signal in the first week post treatment, have you confirmed by DAPI or just a red channel that these cells are alive and not autofluorescent?
          - You mentioned seeing some GFP in the control too right? Do you see any background staining in the control when you do the WB?
          - once you have STAP-stem cells, are you planning on analyzing TCR rearrangements in those proliferating lines or teratomas/mice derived from them?
          - Have you had any luck with other cell types? The field would be very happy to do this with MEF and avoid having to harvest many newborn spleens every time just to study STAP formation.


          • Hi Anne,
            I am not hiding anything, I just don´t want to put up a blot and get slaughtered because it does´t meet the requirements of many other scientist reading this blog.

            The only success I have had with STAP so far is Oct4 expression from unsorted adult rat spleenocytes at 7 days. These cells do not contain any GFP transgene.
            This is being repeated and if it works I will share it here.

            As pertains to your GFP questions: I have no idea if the STAP worked on the spleenocytes from the Oct4GFP mouse.
            I will take the cells out of culture on monday, analyse them and report back.

            I have no intentions of making teratomas or studying TCR rearrangements.

            I did try the STAP protocol on MEFs, but I wash´t´happy with the experiment and I am repeating it.

            cheers,


            • I really appriciate your effort and sharing your data with us; thanks for that!

              But I have concerns about the readout with Oct4 WB only from spleen tissue. Oct4 expression is reported from many mesenchymal stem cells studies and, assumed they were true, could compromise the findings regarding Oct4 positive cells from spleen.
              I would therefore highly recommend to include relatively homogenous fibroblast as STAP resource.

              Keep up your good work!


  19. hello guys,

    there is a interesting paper done in india 10 years that was able to cause transdifferentiation of lymphocyte to granulocyte using sodium nitroprusside. using nitroprusside may make the difference

    google: in vitro lymphocyte-to-granulocyte transdifferentiation induced by chemicals


  20. It’s worth knowing they also got the same results by pipette manipulation squeezing of the cells.

    Perhaps the results attributed to acid bath exposure are the results of the tiny diameter handmade pipettes they are using which squeeze and stress the cells.


  21. As a potential stem cell patient waiting for maturity in the field, let me point out that the frequent controversies on high profile publications (first Mitalipov paper and now this) are extremely trust-eroding.

    Every new announcement evokes more skepticism than even an year ago, and the field overall sometimes appear headed towards gene-therapy’s mistrusted death to obscurity.

    It is frightening to consider that if such high profile claims are so unverifiable, how many lesser claims are complete bogus as well ?

    Is it possible to have a crowd-verified journal where every claim starts off as just that, a claim and only independent external lab verification elevates a claim to that of a publication. I know it isn’t practical for all claims, but the public could come to demand such rigor for a few important claims.


    • Subra:
      I am embarrassed that there appear to be so many mistakes made in publications in this field. It is the responsibility of the senior authors to make sure that the figures are accurate and the images not duplicated, and some of us have been less than diligent. I’m not excusing the authors of these error-filled papers, but I do want to point out that this kind of thing happens far more often in other research fields than in our own. You should look sometime at Retraction Watch, an interesting blog that reports cases of publications being withdrawn for various reasons: retractionwatch.com


    • Dear Subra,

      I’m glad to know that many people are following the progress in the stem cell field. The main thing I would say, is that publications simply cannot be completely vetted beforehand; You should simply not accept a paper as a new truth. A collection of papers from unrelated authors, reviewed by an independent expert, that is starting to get somewhere. So please realize that the field needs time (in this case only a couple weeks for the first answers!!) and specific expertise (there are many subfields regarding stem cells, a lot of stem cell scientists have not noticed holes in this particular paper) before calling a discovery real. The larger public and specifically the media needs to take responsibility and point out that a paper is just a first step for a real discovery!


      • Dr. Loring and Anne,

        Media sensationalism has a lot to do with the controversies, agreed, but this spotlight is all the more reason for the science to be beyond suspicion.

        It would be especially tragic if potential stem cell students consider it a suspicious choice in the future.

        As an interested bystander, I sense a widening gap between promise and reality and wanted to bring it to the attention of insiders such as yourselves.

        As such, my original comment was hoping to initiate dialogue on whether it is time to consider alternative publication mechanisms (such as this brilliant crowd-sourcing verification idea) that provide rigorous certainty to scientists.

        Perhaps a world wide group of labs each of which commits to verify 1 paper of its choice per year as part of peer review.


    • Actually, it’s quite known that the papers in top journals like Nature and Science are full of mistakes and wrong data. On the other hand, in lower-IF journals, the data are usually much more verified. That’s simply because if you have something that could be published in Nature or Science, you’ll rush the publication before somebody else will steal it and take the credit ;)


  22. Well, I personally believe this discovery. Everyone seems to be shocked about this simple methods, they immediately jump into the experiments without developing any skills there.
    Although methods looks very simple and straightforward, you still need time to establish system. Let’s us wait, do not draw any conclusion too early.


    • I’m sorry, but as is clear from the response on this specific blog within two weeks of publication, there are likely 200 labs worldwide trying this. If those people, including the most skilled researchers ever to have worked with pluripotent stem cells, and including Wakayama, cannot reproduce the technique, it’s not a problem of good hands. We need a protocol with missing details, or maybe some stardust?


      • I agree to Steve. As for sure I belive some important steps of the recipie are missing. This because of its ongoing patent progress.

        They may not be sure of the process. Maybe it can be more ways and they want to secure their rights as good as it is possible. That I belive is also a great deal of work. When this is clear I am sure the complete recipie will be published.

        And the the sceptisism will be gone and all of you will dance in you labs! :)


      • I’m from Kobe, but not Riken. I just wondered if Obokata’s way is right, how would people suffering gastric ulcer could not fix it with STAP stem cells. As a tiny organic chemist, I really want the answer from medical view. People with gastric ulcer has STAp in their blood ?


        • Speaking of other instances of acid in the human body local to Obokata, those outside of Japan may be unaware, that drinking acid, in the form of acetic acid, vinegar, is a very popular Japanese folk remedy claimed to have a variet of medical benefits, including lowered blood pressure, reduced cholesterol and blood sugar leve, and reduced obesity. Many Japanese people drink acid as a sort of health promoting elyxir. Some Japanese medical professionals have also produced research supporting this folk belief. See
          Kondo, T., Kishi, M., Fushimi, T., Ugajin, S., & Kaga, T. (2009). Vinegar intake reduces body weight, body fat mass, and serum triglyceride levels in obese Japanese subjects. Biosci Biotechnol Biochem, 73(8), 1837-1843. Retrieved from, http://livar.net/UploadedFiles/Article/Vinegar%20Intake%20Reduces%20Body%20Weight.pdf


  23. I’ve noticed that the people posting are either not using the same age mice or the same tissue type they used in the paper. They have clearly stated in the methods that adult mice did not work, which is why they had to go with 1 week old ones.
    Before you post your negative results, make sure you are using the same conditions as they did since they chose those conditions for good reasons.
    I do believe in the authenticity of the results in the paper and it can be reproducible given the right conditions.
    Good luck to all!


      • I appreciate your blog but, and scientific controversy is good but, with all due respect, this latter comment strikes me as extremely sloppy science. It is irresponsible too considering its importance. You used what you had, instead of what you must? Sounds like the story of going to find your lost keys under the lamp post because there’s light there. You would be lucky to find anything there. Should have waited to get some young animals and do the experiment the right way!


        • In the country I am in, I would have to wait several months to get an animal experiment ethics approval just to remove the spleens.

          As it is published, the Nature paper leaves many technical details out, so almost anything you do to try and repeat the experiment will be sloppy.

          Should we use B27 with or without Vit A, for example.


    • Exactly. After the people make it start working, they can start playing with the conditions. But right now they should simply follow the protocol.


  24. I’m getting a bit impatient with the critiques of the results being posted.

    The motivation of this crowd is NOT to exactly repeat what was published.
    The objective of this sharing of data is NOT to crush a particular researcher or paper. The motivation is to find out whether this general method will be useful for reprogramming cells. If it only works on baby mice, then very few of us have any use for it.

    Let’s just all try it for the cells we care about, and tell each other what happened.


    • “If it only works on baby mice, then very few of us have any use for it.”

      Thank you,yes, count me. I will not do much of anything with STAP cells if it only works in young animals.


  25. Does anyone find the TCR rearrangement data of STAP stem cells or chimeric mice derived from T cell-derived STAP cells in the article paper? Fig1i and Extended Data Fig.2g may not to be sufficient. Without this information, I think authors can not conclude that pluripotency is induced by acid treatment in fully differentiated somatic cells.


    • I agree on that point. I do not think the data demonstrating these cells are not selected for is somewhat limited. CD45+ cells from a 1 week mouse may prove to be a lousy marker in this system. I am an immunologist and it is well known that splenocytes from early mice respond differently from adult mice in many many ways. I think the requirement for infant mice is likely the stumbling block here as admitted to in the protocols. The protocol even acknowledges conversion is lousy in adult mice, and the lack of published data from adult mice suggest is does not translate well to other models.


  26. Hi people, I am not a stem cells biologist, not even a biologist, and I don’t want to annoy you with naive questions, but still I have one.
    Wouldn’t it be possible the experiment really worked, but because something special in the lab itself? Teruhiko Wakayama allegedly reproduced successfully the results in the lab but couldn’t do it by himself outside. What if it’s a genuine “something else” put in the mixture by inattention?
    I am remembering an episode from Malcom in the middle…someone has been recently saved by “House”, who knows ;).


  27. Yoshiyuki Seki is working in the same institute(Kansaigakuin college) where the corresponding author Wakayama worked.
    XXXXX (note from admin: comment edited for content)


  28. Does anyone know if there has been any word of where the full protocol is to be published? Although the results are strange and need to be reproduced, I feel people are perhaps to quick to shout this down


  29. Here is a 2010 publication from Cell Cycle that may be of interest – identifies MARCH7 gene as linked to Nanog expression in adult mouse T lymphocytes.
    “A LIF/Nanog Axis is Revealed in T lymphocytes that lack MARCH-7, a RINGv E3 Ligase that Regulates the LIF-Receptor.” Cell Cycle 2010 9:20, 4213-4221: Thompson HL, Whiston RA, Rakhimov Y, Taccioli C, Liu C-G, Croce C and Metcalfe SM
    The data identifies a potential mechanism for regulation of LIF-responsive gene expression in T cells.


  30. Dear Subra,
    I think we are after the same thing- reproduction- and its true that too many stories fall flat when others try it. However, publication is necessary in competitive fields- you cannot wait for the reproduction in most cases. A typical student or postdoc will spend 4 years on a project, another year on getting it published, I think they need the credit at that point and move on with their career, let alone not get ‘scooped’ by three competitors. A more realistic approach is to apply extra caution when reviewing sensational stories or best- to take any publication with a grain of salt.


    • That’s just not the case. Read the post you’re commenting on and you’ll spot several replication attempts performed within the month, your average peer review takes several months


      • Hi Anon,

        I do not agree. For one, this is an exceptionally fast protocol, with ingredients many could get their hands on in a week and with enormous attention, which makes the first reproductions happen way faster then normal. Second, no one has proven it’s true or not- just that it doesn’t work so far. Now, labs will need a bunch of back-and-forth with the authors and the more detailed protocol to get this apparently tricky technique to work. If it’s true, they need to do all the QC up to tetraploid complementation to create that all-green mouse to prove so; it took Obokata two years to get the critical proof that the cells pass this QC, that is, with the help from several experts. If it’s not true, you’d still need to try it many times and do who knows what to really prove that point.
        So why don’t we just get back to adjusting our view of peer-reviewed articles, to one where we never believe it after the first paper but only once repeated by independent papers; this is common sense in science. If anything, there should be pressure on journals, let alone the top 3, to think about better reviewing practices.


  31. To Katsuaki Oki:
    Journals that publish research studies have a clear policy that is recognized by all of us as fundamental to our profession: the experiments must be described in enough detail and completeness that any other scientist who has the required equipment and reagents can reproduce the results. An example of this clarity is Dr. Yamanaka’s original reprogramming publication in 2006. His experiments were reproducible as they were written, and so hundreds of labs could get the same result following his description in the paper. Think of a cookbook- our papers are like the cookbooks that give precise measurements and directions that will guarantee a successful dish. I don’t like recipes that leave out the oven temperature or some of the ingredients. I don’t like scientific papers that aren’t reproducible.


    • I would like to add to that point: It would be very unethical to purposefully mislead all labs that are trying to copy the technique. Just think of how many man hours, how much precious money and reagents are spend worldwide on this right now- then picture killing 5 newborn pups at a time and dissecting them just for their spleen. For that reason alone, it would be an outrage if they are secretive about the details just to get the patent secured.


  32. T lymphocytes without activation are in cell cycle arrest – “G0″ – but when activated via TCR engagement they enter G1 of the cell cycle and proceed to cycle and expand in numbers leading to an antigen specific clone. Polyclonal activation can be induced by concanavalin A. We found that T cells lacking the E3 ligase, MARCH7 (required for degradation of the LIF-R) express Nanog when activated to enter the cell cycle. http://www.ncbi.nlm.nih.gov/pubmed/20962578
    The activation status of cells may well be linked to (i) ability to respond to LIF (required for STAP cell generation) and (ii) their plasticity: T cells are relatively plastic in terms of cell fate at time of activation: they then track towards different lineages according to the microenvironment. So – given that activation triggers the cell cycle and downstream phenotype, for T cells, this is a relevant consideration.


  33. I have found very strange thing in the gel image of Figure 1i of the first Nature paper. As you may know, splicing the lane 3 has ethics problem. But this is not the only problem. Methods says the authors used two primers Dbeta2 and Jbeta2.6 but Ensemble database shows
    these sequences are only 2.3kB distant from each other. Desipte of this closeness, there is no GL (ie., non-rearranged) TCR DNA band in the lane3. This is very odd given that other lanes show a band at GL and also given that in lane 3 several bands of hundreds bp can be seen.
    Some are arguing that the authors might have artifically modified manipulated the image but when they did so they were unfamiliar with
    the TCR gene dynamics. As you may know, TCR beta, unlike TCR alpha, allelic excusion works, and therefore one of the chromosomes has non-arranged TCR. So the lane 3 should show the GL band easily.


    • Please let me add one comment. Fig1b of the Khor and Sleckman 2005 Eur J Immunol 35:964 showed that nearly a half of T cells (hybridomas in the latter paper) retain the GL-type genomic DNA structure at the region containing TCR beta D and J locus, as indicated by an open arrow in the paper. The top panel of Fig2 of J Exp Med 188: 1669 is an example showing a germline-type band of TCR beta D2 and J2….


  34. Does anyone have the know-how to analyze the RNAseq independently? Would be keen to see wether the oct4 gfp pos cells indeed have a pluripotent profile, but are also different from potentially contaminatings escs..


  35. Why nobody goes to RIKEN to try it in their lab and see with their own eyes whether it works or not? Maybe it’s simply something you have to try first under supervision. I was also working on an enzyme that nobody believed exists, because they were not able to measure the activity. Yet I was able to and I know I didn’t falsify any results.

    I haven’t read the original paper, but from what I’ve read around the web, I understood that only small percentage of cells actually survived. So maybe all of you here around complaining about “two clumps of dead cells” just missed the living ones.


  36. Hi Anne,

    As an outsider with computer science background, I struggle to understand the variables involved that make replication *this* tough. Could you explain, may be with examples, what could go missing in the initial protocol that gets clarified later.

    As a related question, is the nature of these procedures essentially probabilistic ? i.e. Following the “same” procedure, is it possible that the desired outcome occurs only say 10% of the time ?


    • Hi Subra,

      That’s a very good question, and something scientists may fail to clarify often. What do you run into when trying to replicate something? I can tell you a bunch of assumptions we had to make in our lab about the protocol.
      For the CD45+ cells, the protocol is relatively complete because it is the main focus of the paper. We feel that possible deviations we made from the protocol because of lack of information, shouldn’t impact the result, only make it less efficient or slower. However, since it did not work at all, we have to assume that the technique is ‘tricky’; that very small details do make all the difference. Details include: how often to add or change culture media, how to transfer the cells to ACTH media to make STAP-stem cells, whether it matters to use DMEM vs GMEM basal media.. How they manage to use feeders in the ACTH media since they die rapidly in serum replacement (KSR) media.. people have also commented on the impossibility of keeping the pH stable during those 30 minutes, we have not even tried to measure that. We might also miss small details in handling while dissecting the spleens, getting rid of red blood cells, sorting for CD45 positive cells.. all steps with a lot of handiwork.
      It gets way more messy when we try to use other cell types, although supposedly they should work just as well with the acid treatment (Fig 3a). Since fibroblasts are much easier to get at high cell numbers, and much more practical to work with, they are a very valid system to use for reproduction and subsequent follow-up studies into why we can make STAP cells. But they grow in different conditions normally, so do we use the same STAP media initially? Same pH? Adherent or low-adherent culture plates? Do they need longer treatment since they are larger? Will fib-STAP still float when made from adherent cells? What passage number should the fibroblasts be? Do genetic background and sex matter? Do they use neonatal tailtip fibroblasts or embryonic fibroblasts?

      Finally, yes, any experiment can fail sometimes, even the best researchers don’t always know why- it’s biology. Sometimes you just have to repeat it, or optimize conditions to understand which variables are key. It is totally acceptable if even Obokata doesn’t succeed in every experiment she does; but it should work the second and the third time in her or Wakayama’s hands. This again depends on how well small details have been optimized, and how much tricky handiwork is needed along the way. iPSC reprogramming, especially with reprogrammable mouse systems, works 99% of the time, although not always at the same efficiency. STAP, with a supposed higher baseline efficiency should work more often.

      Hope this helps? cheers, a


      • Anne, that is very helpful. I don’t think everyone realizes what it is like to be a stem cell scientist and how it takes incredible perseverance and dedication despite experiments not working at times…and how little details like media choice, cell passage, etc. can have seemingly oversized effects on outcomes.


      • Hi Anne,

        Thanks for the fantastic reply. It is hard to appreciate the sheer number of variables and the fragility of new procedures from outside and this was very educative.

        Going back to a prior comment, you indicate that “paper”s no longer stand for “verified” facts and we should await public judgment before acceptance.

        My respectful 2 cents on this subject:

        If the nature of papers and incentives have deviated from tradition, let us adjust the processes rather than our expectations.

        I feel it is worth preserving the trust and expectation that all published papers are verified, not least to avoid the massive drain on scientists’ time.

        A concession of expectations probably belongs only in politics and of politicians, that too regrettably.

        If Intellectual property and obtaining immediate credit for work is the main concern against stringent publication norms, it may be worth considering an approach similar to Clinical Trials. Every published paper is tagged Phase I or II. Phase I papers are awaiting public judgment and Phase II papers are “verified”.

        This makes it crystal clear where each paper stands and perhaps helps scientists make better decisions with their time.


  37. Dear Dr. Knoepler

    I wish to respect you for setting this interesting site, STAP NEW DATA. But, all the date so far uploaded here are problematic as reconfirmation experiments for STAP. Before discrimination of the data as successful/encouraging or failure/discouraging, you should evaluate the quality of the experiments, as good, bad, or ugly.

    At this stage, I think, reconfirmation experiments must meet at least the two prerequisite conditions as follows:
    1. Researcher must use cell types skin to those used in Obokata’s experiments; that is, cells derived from new born mice, within 7 days after birth.
    2. Stressor subjected to the cells is passing-through a thin-pasteur pipette, and/or acidic treatment at pH 5.7 for 30 min.

    Sincerely,


    • Nekogu:
      I want to clarify that there are two reasons to try the STAP method:
      1. To determine whether the published results are correct- a replication of the experiments.
      or
      2. To determine whether the STAP method is as useful and simple as claimed by the publications.

      Almost all of us are interested in the STAP method only if it works in cells that we care about. For me, that is human or endangered species cells. We aren’t trying to prove or disprove the published results by seeing if we can exactly replicate them. There’s a big difference in motivation between 1 and 2. Do you understand?


    • Hi Nekogu,

      I agree that awe cannot Discard the story just because it doesnt work at first try, or in different settings. Have you done this trituration before? I dont think anyone here has. How long and fast do you do this? In what kind cell concentration? Do you think it should work with MEFs?


  38. You may probably already know, but the detailed protocol of the STAP cell production was disclosed from RIKEN.
    http://www.cdb.riken.jp/jp/04_news/articles/pdf/14/protocol_exchange_v1.pdf

    Despite the serious alleged plagiarism and doctored photos (e.g., same typos as those found in Guo et al.s document; devices used in the method section are no longer available and nobody know how Dr. Obokata got them or it’s unlikely that Dr. Obokata’s research group used the same apparatuses as these are not commonly used in Japan), RIKEN is yet to announce the results of the investigation. Instead, they disclosed the detailed protocol. Anyway, take a look.


    • By the way, in the website pointing out the plagiarism and doctors images I’m reading (this website is written in Japanese) , the blog author says whether or not Dr. Obokata’s team really did the experiment is questionable. According to the the author, several devices listed in the allegedly-copied method section are those rarely used in Japan or those no longer available due to corporate merger etc. If so, I would like Dr. Obokata and co-authors to prove they really did the experiment by showing to the world they do have these machines and the description in the paper is correct. Dr. Wakayama says he wants us to wait for one year for reproduction of the STAP cell. Ok, good. But I don’t see the point of Dr. Obokata’s team having us wait for close to one month for answers to our questions.


  39. I’m curious why there is so little note of using ACTH or any other modulation of the cortisol/inflammatory pathway in this. The literature involved it. I’m reminded of all of the failed mmRNA reprogramming attempts by people leaving out B18R. Even the J. Cooke lab paper on innate immune agonism has B18R in its materials and methods, but people ‘reproducing’ the experiments seem to leave out key components. From what I can tell, every single letter and word in these kinds of complicated experiments matters no matter how little a concept may be emphasized.


  40. I am Japanese molecular biologist and concerned about concealment of unfair action. So I show you whistle‐blowing from the inside on Japanese web.

    “I’m surprised with enormous responses to the last blog, beyond my expectation.
    As several people guessed, I would be ‘an insider’ from the view out of the institute, although I am not directly involved in the papers of matter.

    Inside the institute, I act with real name, and there’s no intention of hiding myself.
    My wish is solving the problem internally, otherwise I would provide all the information
    to the outside.

    The aim of my action is rapid withdrawal of the papers and pursuit of the truth as
    much as possible, with an anger to the damage to trust on science and the institute,
    and also a fear for possible restriction on my research activity that may be caused by
    keeping the matter obscure.

    I am confident that scientific facts on my side, but not confident at all with political strife.
    Thought of writing more evidences, but it will be later because the situation demands
    me to shoot second and third arrows without showing my strategy to the authors of the
    papers.

    The comparison data of ‘input’ described in the previous blog, can be viewed from the
    address below. As the blog cannot show a picture, use UCSC Genome Browser instead.
    Anyone can reproduce the same result using publicly open data.

    TCR-beta
    http://genome.ucsc.edu/cgi-bin/hgTracks?hgS_doOtherUser=submit&hgS_otherUserName=stopstap&hgS_otherUserSessionName=TCR%20beta%20rearrangement%20test

    TCR-alpha
    http://genome.ucsc.edu/cgi-bin/hgTracks?hgS_doOtherUser=submit&hgS_otherUserName=stopstap&hgS_otherUserSessionName=TCR%20alpha%20rearrangement%20test

    I throw in special one. In future, you understand this meaning at last.

    chrX
    http://genome.ucsc.edu/cgi-bin/hgTracks?hgS_doOtherUser=submit&hgS_otherUserName=stopstap&hgS_otherUserSessionName=Appendix


  41. We have been repeating the experiment with MEFs from OG2 mice. We indeed observed a significant increase of GFP positive population 4 days after acid (pH5.7) treatment, however almost all GFP positive cells are also showing strong red/blue fluorescence under the microscope, indicating the presence of intense cellular autofluorescence after treatment.


  42. 03/26/2014 – Japanese morning news ran a story alleging that the mice used for this experiment were switched or mixed up during the process, so the data is even more suspect…Not looking very positive, folks.


  43. *According to the article*, due to the fact that the type of mouse found to be used in the experiment was not type 129, as stated in the paper, but rather of types B6 and F1, which are both commonly used to create ES cells, “not a few” researchers suspect that STAP cells may in fact be ES cells.

Leave a Reply

Your email address will not be published. Required fields are marked *


× 8 = eight

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Anti-Spam Quiz: