The following is a guest post returning to the topic of STAP cells by Keith Jacobs, a PhD student at Washington University in St. Louis studying radiosensitivity in stem cells. You can read more of Keith’s thoughts on science at The Blabbering Biologist, or follow him on Twitter and Google+.
Haruko Obokata, the scientist who first reported the STAP cell methodology, defends her discovery.
Stem cell biology has stimulated intense scientific and clinical interest since the first successful isolation and culture of human embryonic stem cells in 1998. Unfortunately, as a result of legal restrictions on the use of embryonic stem cells put into place under the Bush administration, research during the mid-2000s was forced to focus on developing novel methods for obtaining pluripotent stem cells without the use of embryonic tissue. The first major breakthrough occurred in 2006 when scientists collaborated on Nobel Prize-winning research that successfully created pluripotent stem cells from differentiated fibroblasts, labeled as induced pluripotent stem cells (iPSCs). While there have been numerous recent improvements upon the original discovery of iPSCs, including transdifferentiation which can bypass the stem cell state entirely when converting one cell type to another, all of these techniques remain relatively inefficient.
In January of 2014, scientists from Harvard University and the
Centre for Developmental Biology in Japan published a pair of papers in Nature describing a technique for obtaining pluripotent stem cells from any other cell type, simply by incubating cells in an acidic buffer for 30 minutes, known as stimulus-triggered acquisition of pluripotency (STAP). In addition to its simplicity and the ability to work with cells from any origin, this technique also demonstrated a greatly increased efficiency over previous methods. These findings were considered to be ground-breaking and expected to revolutionize the stem cell field, resulting in international media coverage. Due to the simplicity of the assay, it was surprising to many that the method had not been previously discovered. Additionally, many labs struggled to reproduce the results, albeit often with somewhat different methods. While inability to reproduce a technique does not on its own suggest any major flaws in a publication, this raised suspicions about the validity of the data.
Within weeks however, a myriad of allegations threatened the integrity of the publication. First, online commenters noticed apparent duplications of figures from previous publications. The Wall Street Journal then posted evidence that the lead author had plagiarized part of her PhD thesis. While all authors continued to support the overall claims of the paper despite admitting some “mistakes”, the increasing questions about the study soon caused one of the co-authors to request retraction of the paper. The Riken Centre began an internal investigation into several questionable figures and on April 1st officially declared that research misconduct occurred due to falsification of data. They reported two specific instances of misconduct, in addition to multiple other inconsistencies. Currently the two STAP papers have not been retracted and the lead author continues to stand by her findings. While the story is even more complicated by the recent resignation of the chairman of the investigative panel over accusations of fabricated images in his own publications, it appears that these publications have all but met their ultimate demise.
Personally, I believe that the overall phenomenon is likely true, however in order to sell the paper to a top journal, the authors needed a large amount of confirmatory evidence and controls that were too difficult to obtain. When this paper was first published, both myself and other members of my department thought how beautifully the phenomenon of stress-induced de-differentiation would explain many previously unexplained experimental observations. However, even when you strongly believe that the results of a novel discovery must be true based on all the numerous little observations and associated phenomena you have noticed, it can be very difficult to obtain enough clean, believable confirmatory evidence in order to appease reviewers. It is therefore easy to understand the temptation to fabricate some controls, and this explains why the author in question adamantly stands by her work despite the declaration of fraud. The existence of clear and evident research misconduct does not completely discount the discovery, however it does promote a strong lack of credibility, and it will need to be successfully repeated by multiple other labs before it will be accepted by the scientific community ever again.