New Message from Wakayama on STAP retraction & origin of STAP-SC

I have been corresponding now and then with Dr. Teru Wakayama about the ongoing STAP situation. He asked me to pass along the following message from him for clarification on the STAP Nature paper retraction and the origin of the STAP stem cells (STAP-SC).

I would like to take this opportunity to explain the reason for certain differences between the retraction statement in the published paper version of Nature Magazine and the online version of the STAP paper retraction, specifically related to reason No. (5) which was slightly different between the two.

Last month, I reported to the media about the apparent strain difference between mice used in our lab and the STAP cells. Our mouse line uniformly carries identical cag-gfp insertions in chromosome 18, however, STAP-SC appeared to have a different GFP insertion site in chromosome 15. After learning this, I asked for a further analysis to obtain more hints as to the original mouse strain corresponding to STAP-SC. My collaborator found that perhaps the GFP insertion site of STAP-SC was in fact not chromosome 15. However, importantly, the GFP insertion site is absolutely different between our mouse line and STAP-SC. We know this to be the case because we demonstrated that one primer (part of chromosome 18 and cag) only gave a PCR band in our mouse line, but not in the STAP-SC. Thus, the retraction reason of no. (5) is absolutely right. Meanwhile, we are now trying to find the true insertion site of GFP in STAP-SC. Unfortunately, for the paper version of Nature, I could not clarify this point because the deadline had passed. Only the online version could be corrected.

We apologize for any confusion, but in the best interests of science and complete transparency, we wish all of this information to be freely available.

Teru Wakayama

Notes from Paul: I did some minor editing of this text for clarity.

Below is the online retraction statement reason no. (5):

“(5) In the Article, one group of STAP stem cells (STAP-SCs) was reported as being derived from STAP cells induced from spleens of F1 hybrids from the cross of mouse lines carrying identical cag-gfp insertions in chromosome 18 in the background of 129/Sv and B6, respectively, and that they were maintained in the Wakayama laboratory. However, further analysis of the eight STAP-SC lines indicates that, while sharing the same 129×B6 F1 genetic background, they have a different GFP insertion site. Furthermore, while the mice used for STAP cell induction are homozygous for the GFP transgene, the STAP-SCs are heterozygous. The GFP transgene insertion site matches that of the mice and ES cells kept in the Wakayama laboratory. Thus, there are inexplicable discrepancies in genetic background and transgene insertion sites between the donor mice and the reported STAP-SCs.”

Interview with Nature on their editorial process in wake of STAP

NatureI asked Nature a half dozen questions about their editorial process. While they declined to answer any direct questions about the STAP cell paper situation, I thank them for answering these questions via a Nature spokesperson. The end result is an intriguing glimpse inside the editorial/review process at Nature.

1. Does Nature have any kind of automated (or human-based) system for checking submitted or accepted manuscripts for plagiarism? If so, when was this system instituted? If not, why not?

Nature uses plagiarism software (the CrossCheck service which uses the iThenticate software and the CrossRef database) to check all our published papers; however, the software did not detect plagiarism in this particular case. The manufacturers of the software are currently looking into this.

2. Does Nature have any kind of automated (or human-based) system for checking submitted or accepted manuscripts for image manipulation or duplication? If so, when was this system instituted? If not, why not?

Nature Publishing Group utilises tools to do randomised spot check analysis of images but we currently do not have the resources to undertake detailed image analysis on all our papers. For duplications, we do not have resources to check large numbers of figure panels against each other.   However, we are actively reviewing our policy on image checking and have decided to increase the number of checks that we undertake on Nature journal papers.  The exact number or proportion of papers that will be checked in the future is still being decided.

3. Does Nature vet potential reviewers for conflicts of interest (COI) before assigning them to manuscripts or does it rely on the reviewers to self-disclose COIs after they have been invited to be reviewers of specific manuscripts? On this page it suggests that Nature does not itself look for COIs (http://www.nature.com/authors/policies/competing.html), but rather relies on reviewers to self-disclose. What if a reviewer fails to disclose a significant COI?

Editors are well connected within the research communities that they serve and use their judgement not to request reviews from people in the same department or those who co-publish regularly.  However we cannot, of course, know if reviewers are on the same grant or who they may be consulting for, so have to rely on a system where a reviewer will make any potential COI known to us.

4. It’s been brought to my attention that while some Nature Publishing Group journals are signatories to the Committee on Publication Ethics (COPE), Nature Magazine is not. Why would Nature not be part of this group that works to make the publication process as ethical as possible? (see link here: http://publicationethics.org/category/publisher/nature-publishing-group).

Nature editors have engaged with COPE on occasion. However, they and the Chief Editors of the Nature journals have a long history of developing policies collectively for our journals – policies that are more attuned to basic research than is COPE, which has a strong clinical flavour.

5. Would a submitted patent application by an author for the technology described in a manuscript be required by Nature to be disclosed as a competing interest? If authors fail to do so, what is Nature’s policy on what happens as a result? This webpage would seem to clearly indicate that patent applications should be disclosed: http://www.nature.com/authors/policies/competing.html.

The Nature journals require authors to declare to the editors any competing financial interests in relation to the work described. As part of this policy, authors are encouraged to declare patents or patent applications whose value may be affected by publication. However, appreciating how much value can represent a competing interest is a subjective matter, and many authors have patent applications that do not amount to material value. In these cases, the editorial office may feel that the value is not substantive enough to justify declaration of competing interest.

6. Does Nature have a bioethics person on staff or on contract to consult with should ethical dilemmas arise during the review process?

A number of members of staff at Nature have responsibility for considering ethical dilemmas and providing advice during the review process.  We also solicit the advice of experts on subjects such as biowarfare, use of human oocytes, etc, when needed.

Week 4 poll: few hopeful about STAP stem cells anymore

I’ve been doing a weekly-ish poll on people’s sentiments about STAP stem cells. It’s not scientific but many people have told me they find it interesting.

STAP polling

Week 4’s poll is done with more than 500 votes and it is even more discouraging than week 3. See the purple line in the graph as that is the latest polling.

The question is “Do you believe in STAP stem cells?” and the answers range from 100% NO to 100% YES with a few in between answers. “CLOSE” means that you are close to convinced they are or are not real. Then we have leaners and on-the-fence.

Few people have much belief left in STAP stem cells at this point.

However, I am sure that if there is even one convincing replication result reported by a solid lab that this graph will flip again back to the positive as it was in Week 1. But that’s a big if.

Shortly I will start the week 5 poll or switch to a monthly format….this could be a long slog.

Interview with Dr. Teru Wakayama on STAP stem cells

wakayamaI asked Dr. Teruhiko Wakayama, who goes by Teru to those who know him, if he would be willing to do a Q&A on the STAP stem cell situation.

Teru was senior author on the STAP stem cell Nature letter on chimerism, but was not a senior (note this is a correction from an earlier version of this post) author on the other STAP paper, the Nature article on the production of STAP stem cells.

Teru kindly agreed to my invitation. He responded straightforwardly to what I thought were direct and sometimes tough questions. He also lays out a reasonable plan for STAP looking to the future. Thank you, Teru.

To me his answers reflect his great reputation as both a scientist and person. Teru is a true good citizen of the stem cell field. Teru is what I would call a “mensch”.  I fully support Teru’s proposal that we give STAP stem cells a year to be reproduced.

Here’s the interview.

1. How did the STAP stem cell collaboration begin between the Vacanti lab and your group? What made you decide to team up with them? Can you please tell us more about the beginnings of this research?

Teru: Dr. Kojima (Vacanti’s lab) contacted me by e-mail to help with chimera experiments. At that time, the project looks very much impossible. That’s why I accepted. I like such impossible experiments.

First time, Dr. Obokata brought strange cells, and there was no chimera after blastocyst injection. However, nearly 2 year later, Dr. Obokata found a very good method to generate STAP cell. Then, we could obtain good chimera.

2. Have you talked to Drs. Vacanti and Obokata recently? How did that talk go? If they are not talking, why is that so?

Teru: I have not talked to Dr. Vacanti.

I had a talk with Dr. Obokata, but in Japan, the main problem is, not reproducibility, it is the mistakes of picture or band. Now RIKEN and outside people investigate that problem. But she said that in her lab, she can create STAP cell.

3. What is your own level of confidence in STAP cells at this point? Are you getting more concerned?

Teru: Before I left RIKEN, I succeeded to make STAP cell from spleen. But only 1 time. At that time, Dr. Obokata taught me very well.

Now, some of my friends (not Japan) sent me e-mails, which, reported partial success (Oct expression only). Therefore I believe that within one year, someone will publish about STAP generation.

4. Like most people, I am convinced that the mouse studies on STAP are solid and convincing. You personally have a top-notch reputation as a scientist around the world. People are instead more specifically concerned about the STAP stem cells themselves. One of the most common questions I am getting asked at this point is this–could the STAP cells have been contaminated with either mESCs or mouse iPS cells? Is that possible? How might that have happened?

Teru: Thank you for your comment.

I established STAP-SC several times from STAP. It is unlikely that  contamination would always have happened. In addition, we established STAP-SC from 129B6GFP mice. At that time, we did not have this strain ES cell.

When I succeeded to establish STAP-SC, the original STAP cells expressed Oct4-GFP very much. In this condition, the establishment is much easier than ES cell establishment from blastocyst.

In addition, whole mRNA expression data suggest that STAP-SC are not ES cell.

5. You mentioned that STAP stem cell is not working in your new lab. Why do you think that might be in terms of the methods? What is different now? Also, you mentioned that STAP worked for you when you were at RIKEN. Can you please be a bit more specific? Did you do 100% of the steps in the STAP induction yourself? Again, could iPS cells or ES cells somehow have gotten into the culture?

Teru: I just learned one time from Dr. Obokata, then left RIKEN.

Do you know when we moved to a different lab in the past, how difficult it was then to reproduce even my own techniques? When I moved from Hawaii to Rockefeller, I spent half a year to reproduce cloned mice. This is my technique, but I still needed a lot of time. However, the method of STAP generation is not my technique, therefore, different lab and not my discovered techniques, so this is understandably more difficult to reproduce again.

I did 100% by myself, but each step was looked by Dr. Obokata. Much the same, one of my PhD student also succeeded to establish STAP-SC.

In those early stages of the experiment, we did not culture ES cell nor iPS cell at same time. After that, as control, we sometimes culture ES cell at same time.

6.  Many people are trying the STAP method around the world and getting frustrated because it isn’t working. I know there is a methods paper in the works, but given how important this is would you consider putting a detailed step-by-step protocol out to the scientific community right now? For example, I’d be happy to post it on my blog and it would have no effect on the publication of a methods paper in a journal. This might really help people to get it to work. Waiting a month or two for a methods paper to come out might be too late.

Teru: Yes, now RIKEN will soon publish the detailed method. I contributed chimera and to establishment. However, chimera and establishment is usual protocol, not specific, because it is the same or more easier than ES cell. Unfortunately, now all responsibility is in RIKEN, and I left RIKEN. Although I want to know but I do not know now.

Finally–Do you have any other questions or concerns that I did not ask but that you want to address? If so please add more.

Teru:  I do not escape. Because in my results everything is true. However, it is going to take time to reproduce the new technique. For example, the first cloned animal, Dolly, wasn’t reproduced for one and half years after it was published. Human cloned ES cell paper has still not yet been reproduced. Therefore, please wait at least one year. I believe that during this period someone or myself will success to reproduce it.

Note: At his request, I have made a scattering of small corrections to Teru’s answers just for a few typos/spelling/grammar issues since English is not his first language, but I did not change in any way the tone or significant content of his answers.