From Paul Knoepfler's blog:
http://www.ipscell.com/2014/01/review-of-obokata-stress-reprogramming-nature-papers/

"...something akin to hitting the cells over the head with a sledgehammer of a pH 5.7 (physiological pH is more typically thought of as around 7.4), they report the blood cells of 1-week old mice turned on expression of an Oct-GFP reporter as they floated around in clusters in the media."

"...the team provided pretty good evidence that the STAP cells arose from the differentiated blood cells themselves rather than potentially from rare pre-existing primitive stem cells in the cell populations."

"...After a relatively quick read, no particular red flags jump out at me from the STAP cell paper. It just seems too good and too simple of a method to be true, but the data would suggest so far at least that this team is onto something really important."


But key open questions remain before anyone can really say just how important this is.

1. Will it be reproducible by other labs?
2. Will it work in human cells?
3. Will it work in adult cells?
4. What are the molecular mechanisms?
5. Do these cells possess significant rates of mutations or epi-mutations, the latter being abnormalities in the epigenome?
6. Are these cells tumorigenic (besides forming teratoma)?

In particular, if the answer to one or more of the first 3 questions is no, then the impact could be significantly muted.

Perhaps I don't know enough, but I am struggling to understand the results presented in figure 1i and how they can come about - any clarification would be great, because internet searches have failed to resolve this.

Both the fibroblast and the ES line shown in that gel have an unrecombined TCRB band, and the lymphocyte sample is representative of a population with multiple recombined variants but no unrecombined one, as expected. However, both OCt4-GFP+ samples, which are derived from CD45+ lymphocytes as per the text, have bands of both recombined and unrecombined lenghts - are there known cellular mechanisms that will undo this VDJ recombination, somehow?

In my view the main concern is that is all done with a gfp-oct4 reporter transgenic line. if for whatever reason that reporter became responsive to pH you will have Oct4 expression , and this may be the event leading ti reprograming rather than the indicator of an ongoing reprograming state? What about an independent gfp-oct4 reporter line.? what about a non tarnsgenic liene? are these cells also reprogarmed?
In any case I think this is a fascinating paper both in its simplicity and econmic implicances on regenerative therapy. You cannot patent HCl tretments!

Figure 1i lane 3...
At higher magnification the background of that lane 3 is darker than the rest of the gel.
Also vertical straight change background on each side.

In Fig 5d, antibody staining for OCT4 and NANOG - signal is cytoplasmic and should be NUCLEAR!! As it is the case for KLF4 and Esrrbeta, in the same figure. This result does not "compromise" any of the main claims of the paper, but reveals careless experiments, missed also by reviewers… signs of times.

Knoepfler keeps on blogging, although there's maybe not that much substance at this stage (he mentions not having read the sister paper yet). Readers may find an interview with Vacanti of interest

http://www.ipscell.com/2014/02/interview-with-charles-vacanti-on-stap-cells-link-to-spore-stem-cells-more/

Would have been interesting to ask him to clear up the issue of Fig. 1i.

Knoepfler turns skeptic:

http://www.ipscell.com/2014/02/in-a-pickle-over-stap-stem-cells-top-5-reasons-for-skepticism/

Vacanti has made revolutionary claims in the past that have not been easy for most labs to reproduce. These have concerned "spore-like" cells (or VSELs - Very Small Embryonic Like stem-cells), which were claimed to be virtually indestructable. In the interview linked above, Vacanti makes the claim that STAP cells are probably VSELs. Anybody wishing to comment on that work could make a start here:

https://pubpeer.com/publications/11135375

There is also an orphan wikipedia page:

http://en.wikipedia.org/wiki/Spore-like_cells

FWIW Knoepfler is not a fan and links to further critical discussion:

http://www.ipscell.com/2013/07/are-vsels-the-sasquatch-of-the-stem-cell-field/

Hi,
I've set up a new page for people to crowdsource their results from attempts to replicate the STAP stem cell method here. I blogged about it here: http://www.ipscell.com/2014/02/invitation-to-crowdsource-stap-stem-cell-follow-up-data-here/

The page is here: http://www.ipscell.com/stap-new-data/

Hopefully some people will make their data publicly available and for discussion this way so we can all learn more and talk about it...faster.

Paul

I'm doing a new poll 2nd week into the STAP situation (you can vote again in this one) to judge the dynamic ranging opinions on the STAP paper and methods. http://www.ipscell.com/2014/02/do-you-believe-in-stap-stem-cells-new-poll-week-two/
Please voice your opinion via this poll. I'm curious if people are becoming more convinced or less or staying about the same as this story evolves.
Paul

New article features some evidence of endogenous dedifferentiation in response to low pH et al offered up by developmental biologists including Gurdon. Also features interview with Vacanti. http://www.biosciencetechnology.com/articles/2014/02/new-stem-cells#.UvpRHUbv6HI

STAP stem cell crowdsourcing flowing along nicely!
http://wp.me/P1xWpk-4pP
3 reported attempts, all didn't work. Still early days, but nice to see people into openscience like this.
Paul

They conclude that their data indicate that strong stress reprograms mature lymphocytes into multipotent stem cells. As discussed on this site I also think their data are still not very solid for this conclusion.

It seems to me that the data might instead demonstrate stem cells residing in mouse tissue such as "Muse cells" (Proc. Natl. Acad. Sci. USA. 108(24):9875-9880 2011) and they are the only survivors of the treatment. Usage of T cell clones, instead of crude T cells, would be helpful to exclude this possibility.

This comment on a previous paper from the group seems relevant to add some context to the discussion here:
https://pubpeer.com/publications/D02B9A8C8A6D5C99B95E4FB3E1E13C#fb6333

A compilation of some of the image issues

http://stapcell.blogspot.com/

The Riken research institute probes 'irregularities' in images associated with STAP cell discovery.

http://mainichi.jp/english/english/newsselect/news/20140215p2a00m0na009000c.html

"At this point in time, our judgment is that the research findings are solid," a Riken representative commented.

Latest news from Knoepfler's crowd-sourced replication/extension effort is that 0/8 attempts have led anywhere so far.

http://www.ipscell.com/stap-new-data/

Here's an English translation of a popular Japanese blog post: http://horikawad.hatenadiary.com/

update: 2014.2.14

"Regarding the previous entry that talked about a portion of the image that was suggested to be unnatural, we have received numerous inputs from the readers that it may be an artifact that emerged during the compression of the jpeg image. Therefore, we have now eliminated the said image and the accompanying part of the blog. We apologize to the readers and everyone involved for any inconvenience caused.

Also, because the said blog entry has created unforeseen amounts of attention, there is a risk that the contents of this passage could be reported by the mass media who misinterprets and sensationalizes the content, and because some of the comments left on my blog were verging on threats, the said blog article will be temporarily held from publication. It is one of our duties as researchers to discuss published work from various angles, however, it seems that our deeds can be easily misinterpreted by those outside of science.

As I have talked about it on multiple occasions in this blog, I am greatly amazed by the STAP research by Dr. Obokata and her colleagues and look forward to seeing its future progress. That is why we dared to talk about the feature of the image that was suggested to be unnatural. We thought it would be unfair to just talk about how amazing the study is without raising the issues of possible caveats. We'd like to stress that it was never aimed to personally dishonor the researchers involved."

update: 2014.2.15

It appears that Riken has launched its own investigations on the part of the image suggested to be unnatural.

(link to the news on how Riken started the investigation)

We will wait to hear the results of this investigation. We've seen other people suggest that our blog was key that led to this investigation, but that is not true. Our blog article went online on Feb 14 while Riken had already contacted the outside consultants as of Feb 13. It is presumable that it was the following English website that triggered the investigation.

(link to PubPeer)

Also, the following site shares the results from some of the attempts to independently test the STAP protocol. It seems that it has yet to be completely reproduced, however, there are some results that seem slightly positive.

(Link to Knoepfler Lab)

There's something I find slightly odd about Extended Data Fig. 2g. Most of the rightmost (4th) lane seems to be darker than the immediately adjacent background. The issue is just discernible with the naked eye, but I've adjusted levels and contrast to highlight it in this image:

http://imgur.com/cSwFTJn

Intuitively, I would have expected any signal simply to add to the background and therefore to increase the overall intensity. There are other variations in background of similar magnitude across the image (darker towards the bottom left), but the change at the edge of the 4th lane is rather sharp and well aligned with the lane.

Is a scale bar in Fig. 2a correct? In my view, size of DAPI-positive nuclei is completely different among panells in Fig. 2a. However, they used only one scale bar for them. What do you think about this?

Week 2 polling on STAP stem cells is now up on my blog and reflects more uncertainty: http://www.ipscell.com/2014/02/stap-stem-cells-week-2-poll-results-more-uncertainty/

It will be nice to see replications in other labs but this seems to have been a case of images being mixed up when assembling the paper. Surely these results will be reproduced soon?

Here is an English translation of an article in the Japanese Newspaper Asahi Shimbun (http://www.asahi.com/articles/ASG2L53TJG2LPLBJ002.html):

======================
On February 18th, in response to the inquiry regarding a suggestion that two placenta images looked identical, Dr.Teruhiko Wakayama, a professor at Yamanashi University and a co-author on the STAP article led by the unit leader Obokata, has responded to Asahi News paper and explained, “It is a simple mistake in which 2 different images of the same mouse taken from different angles were used when one of them was supposed to be removed”.

Prof. Wakayama used STAP cells to produce the mice and imaged them. He took multiple images of the same embryo from different angles. As a result, Obokata misunderstood the identity of those images and used them redundantly. Being overextended with making figures while performing revision experiments all by herself may have been a cause for such misunderstanding, explains Prof.Wakayama.

He adds, “After multiple revisions, the final manuscript does not refer to the two images, but she forgot to remove them”.

Waseda University to begin its own investigations

On February 18th, Waseda University from which Obokata has obtained her PhD has also launched its own investigations. Her dissertation refers to the paper she published in 2011 while studying at Harvard University, and some blogs have raised skepticism that there may have been another duplication of some of the images in this publication.

The figures in question describe the presence of certain genes, however, there were several places in which what appear to be identical images were labeled with different gene names. Prof. Charles Vacanti has explained to the journal, Nature, that it appears to be a simple mistake without any ill intent and has requested to removes these images from the article last week. Spokesperson at Waseda University states, “Even if the said images were removed from the article, we consider that this will have no impact on the validity of her doctorate”.

Nature News have a interesting piece on the affair

http://www.nature.com/news/acid-bath-stem-cell-study-under-investigation-1.14738

Of note is their own survey of stem-cell experts about their success with the protocol. Nobody reported success from their preliminary tests.

If you increase the brightness in Fig. 3B you will see that the control image on left is all red without any green and the acid-treated panel on the right is all green with no red. Different filters cubes used on the scope for control vs. experimental = bad news.

There is skeptic data shown in Obokata et al. 2011, Nature Protocols too.
http://www.nature.com/nprot/journal/v6/n7/fig_tab/nprot.2011.356_F5.html

Corresponding mean number of B cells (Fig. 5a) and that of neutrophils (Fig. 5b) are unnaturally similar to each other, while S.Ds are different between the corresponding values.

It is almost impossible that two independent data with different cell types and conditions extremely looks alike.

Extended Data Fig. 2g, lane 2 and lane 3 are too similar.

http://imgur.com/uqvuu8H

This can happen? Any idea?

Extended Figure 2a
Endogenous Oct4 staining does not match with Oct4-GFP expression.
It is acceptable that Oct4 staining increases from day2 to day7, but a complementary pattern is quite odd. A fundamental question is: Does Oct4-GFP faithfully reflect expression of endo?

Two regions in Extended Data Fig. 2g seem to display very little noise. This can be seen in this processed version, in which a 'sharpen' operation slightly increases the noise and equalises the background, followed by a shadowing operation. The absence of noise gives a flat grey appearance.

http://imgur.com/03p5VDb

The sentence for "Karyotype analysis" in methods section of this article is exactly similar to "Chromosome preparation" methods in other researchers' article, "In Vitro Cell Dev Biol Anim. 2005 Sep-Oct;41(8-9):278-83.". Nevertheless, authors of the Nature article did not cite this source.

http://stapcells.blogspot.com/2014/02/blog-post_7834.html

Please check out the following HP address:
http://stapcells.blogspot.jp/2014/02/nature-article.html

Also, check out the following photo in which Nature 505 pp641-647 Fig.2e and Obokata's doctoral dissertation Fig.14 are compared.
http://2.bp.blogspot.com/-oh3E_LK3egM/UxwOpCvs1MI/AAAAAAAABDk/r2kN6CQfVEM/s1600/NatureandD.jpg

The former is for STAP cells-derived teratoma, and the latter is for bone marrow sphere-derived teratoma.
However, immunohistochemistry images in Nature's Fig.2e and D.thesis's Fig.14 are similar to each other.

The ensuing investigation revealed that Obokata plagiarized doctoral thesis.
33 pages of Obokata’s doctoral dissertation introductory chapter, it was almost all copy and paste.

Original
http://stemcells.nih.gov/info/basics/pages/basics1.aspx

Obokata’s doctoral thesis
http://stapcells.up.seesaa.net/image/Background.pdf

The interim report of the Riken investigation is here:

http://www.riken.jp/en/pr/press/2014/20140314_1/

The video image of STAP cell generation displayed in RIKEN homepage suggests to me that two GFP+ cells are dead because of the following reasons.
1. These cells form bubbles inflating and disappearing for an instant just before green fluoresence generation (about 11 seconds after the start of the vido).
2. These cells do not show any movements at all after the fluoresence fomation.
3. The sizes of these cells are very small like apoptotic cells.

If I see Videos 1 and 2 supplemented in this paper without any assumptions, I will believe that these video images display phagocytosis of green particles by macrophages. Macrophages reside in spleen and express CD45. Because macrophages are known to ingest dead cells, it can be supposed that several dead cells generating green fluorescence can be phagocytosed by a macrophage and thereby look like cell clustering. Macrophages move very actively in culture even after phagocytosis, and therefore green fluorescent “cell clusters” look like mobile with cell processes.

When I read the paper for the first time one of the things that struck me was the incredible perfect differentiation of pancreas and endoderm in their teratomas. I even thought they were sections of real pancreas and intestine not a teratoma. With all the unethical usage of images in the paper and in Obokata thesis now I think that it's possible that those pictures indeed are sections of pancreas and intestine. I hope that someone in Riken is willing to analise this issue.

Fig1c. At upper right panel at day 7, should see two populations because still a large number of cells w/o gfp.

Dr. Obokata offered to withdraw her doctor thesis against Waseda university.
(Japanese article)
http://sankei.jp.msn.com/science/topics/science-16569-t1.htm

It is worth linking to a detailed discussion of the FACS experimental design posted by Jun Seita on Paul Knoepfler's blog:

http://www.ipscell.com/2014/03/guest-post-by-dr-jun-seita-a-simple-proposal-to-make-stap-protocols-clearer/

Before a press conference on the final report of investigation committee about STAP papers, RIKEN replaced the interim report of the investigation released on 15 Mar with a new version which deleted large parts of gel images on 31 Mar WITHOUT ANY NOTIFICATION, just a day before the press conference.

See pp.9-10: large parts of Gel 1 / Gel 2 images were deleted.

Original version on 15 Mar

http://web.archive.org/web/20140314222554/www.riken.jp/~/media/riken/pr/topics/2014/20130314_1/document-5.pdf

New version on 31 Mar
http://www.riken.jp/~/media/riken/pr/topics/2014/20130314_1/document-5.pdf

In the press conference on 1 Apr, the committee explained Obokata requested to delete the large parts of Gel 1 / Gel 2 as mentioned above because they were "unpublished data".

But some experts argue that Gel 1 seems to be an original version of a gel image in their application to international patent, Fig. 20.

http://patentscope.wipo.int/search/en/detail.jsf;jsessionid=D69EB75A5D6EF3620894BE3F2E2716BB.wapp2nA?docId=WO2013163296&recNum=1&tab=Drawings&maxRec=&office=&prevFilter=&sortOption=&queryString=

Madhusudana Girija Sanal | Mar 10 2014 18:33 EST
It is a common observation that stressed cells, dying and dead cells start to become fluorescent. So early observation of florescence as a pointer to pluripotency (Oct4-GFP) needs to be substantiated by other experiments. The authors need to provide more evidence (molecular biological and morphological) and details (protocol) regarding STAP during the early hours to days post-stress, because this period seems to be the most critical in reprogramming (compared to characterization of the final product-STAPs). The authors underscore that the production of STAP is not the result of any type of “selection” during culture. However in a subsequent protocol paper (doi:10.1038/protex.2014.008) they see a ‘possibility of negative cell-type-dependent bias’. In the new protocol paper they also didn’t observe any T-Cell receptor rearrangement. So what type of PTPRC (CD45+) cells contributed to STAP? It is also difficult to understand or define the difference between STAP cells and STAP stem cells.


http://www.ncbi.nlm.nih.gov/pubmed/24476887#cm24476887_3331

The Science piece about the retraction is quite interesting. The gel splice that was commented early on had been spotted by a referee at Science (who then rejected the MS).

http://www.sciencemag.org/content/344/6189/1215.full (sub required.)