ヒトを含む哺乳類では、受精卵が分裂して血液や筋肉など多様な体細胞に変わり、その種類ごとに個性づけ（分化）されます。体細胞は分化を完了するとその細胞の種類の記憶は固定され、分化を逆転させて受精卵に近い状態に逆戻りする「初期化」は、起きないとされています。

初期化を引き起こすには、未受精卵への核移植である「クローン技術」や未分化性を促進する転写因子というタンパク質を作る遺伝子を細胞に導入する「iPS細胞技術」など細胞核の人為的操作が必要です。

もし「特別な環境下では、動物細胞でも“自発的な初期化”が起きうる」といったら、ほとんどの生命科学の専門家が「それは常識に反する」と異議を唱えることでしょう。

しかし、理研発生・再生科学総合研究センターの小保方研究ユニットリーダーを中心とする共同研究グループは、この「ありえない、起きない」という“通説”を覆す“仮説”を立て、それを実証すべく果敢に挑戦しました。

共同研究グループは、まず、マウスのリンパ球を用い、さまざまな化学物質の刺激や物理的な刺激を加え、細胞外の環境を変えることによる細胞の初期化への影響を検討しました。

その過程で、酸性溶液で細胞を刺激することが初期化に効果的だと分かりました。

実験では多能性細胞に特有の遺伝子「Oct4」が発現するかどうかで初期化の判断をします。

詳しい解析の結果、酸性溶液処理によってリンパ球のT細胞に出現したOct4陽性細胞は、T細胞にいったん分化した細胞が初期化された結果、生じたものであることを突き止めました。

また、このOct4陽性細胞は生殖細胞を含む多様な体細胞へ分化する能力をもつことが分かりました。

さらに、ES細胞やiPS細胞などの多能性幹細胞などではほとんど分化しないとされる胎盤など胚外組織に分化することも発見しました。

一方で、酸性溶液処理以外にもガラス管の中に細胞を多数回通すなどの物理的な刺激や、細胞膜に穴をあける化学的刺激でも初期化を引き起こすことが分かりました。

小保方研究ユニットリーダーは、こうした細胞外刺激による体細胞からの多能性細胞への初期化現象をSTAP（刺激惹起性多能性獲得）、生じた多能性細胞をSTAP細胞と名付けました。

また、STAP現象がリンパ球だけで起きるのではなく、脳、皮膚、骨格筋、肺、肝臓、心筋など他の組織の細胞でも起きることを実験で確認しました。

細胞外刺激による細胞ストレスが、分化状態にある体細胞の記憶を消去して初期化する－という今回の成果は、これまでの細胞分化や動物発生に関する常識を覆し、細胞の分化制御に関する新しい原理の存在を明らかにしたものです。

細胞の分化状態の記憶を自由に消去したり、書き換えたりできる次世代の細胞操作技術となる可能性が高く、再生医学以外にも老化や免疫など幅広い研究に新しい方法論を提供します。

今後、ヒト細胞への適用を検討するとともに、さらに初期化メカニズムの原理解明を進めていきます。

◇　◇ 

 Drs. Obokataの業績を讃える論文が見つかったので、掲載しておこう。

“It’s amazing. I would have never thought external stress could have this effect,” says Yoshiki Sasai, a stem-cell researcher at the RIKEN Center for Developmental Biology in Kobe, Japan, and a co-author of the latest studies. It took Haruko Obokata, a young stem-cell biologist at the same centre, five years to develop the method and persuade Sasai and others that it works. “Everyone said it was an artefact — there were some really hard days,” says Obokata.

Obokata says that the idea that stressing cells might make them pluripotent came to her when she was culturing cells and noticed that some, after being squeezed through a capillary tube, would shrink to a size similar to that of stem cells. She decided to try applying different kinds of stress, including heat, starvation and a high-calcium environment. Three stressors — a bacterial toxin that perforates the cell membrane, exposure to low pH and physical squeezing — were each able to coax the cells to show markers of pluripotency.

But to earn the name pluripotent, the cells had to show that they could turn into all cell types — demonstrated by injecting fluorescently tagged cells into a mouse embryo. If the introduced cells are pluripotent, the glowing cells show up in every tissue of the resultant mouse. This test proved tricky and required a change in strategy. Hundreds of mice made with help from mouse-cloning pioneer Teruhiko Wakayama at the University of Yamanashi, Japan, were only faintly fluorescent. Wakayama, who had initially thought that the project would probably be a “huge effort in vain”, suggested stressing fully differentiated cells from newborn mice instead of those from adult mice. This worked to produce a fully green mouse embryo.

Still, the whole idea was radical, and Obokata’s hope that glowing mice would be enough to win acceptance was optimistic. Her manuscript was rejected multiple times, she says.

To convince sceptics, Obokata had to prove that the pluripotent cells were converted mature cells and not pre-existing pluripotent cells. So she made pluripotent cells by stressing T cells, a type of white blood cell whose maturity is clear from a rearrangement that its genes undergo during development. She also caught the conversion of T cells to pluripotent cells on video. Obokata called the phenomenon stimulus-triggered acquisition of pluripotency (STAP).

The results could fuel a long-running debate. For years, various groups of scientists have reported finding pluripotent cells in the mammalian body, such as the multipotent adult progenitor cells described4 by Catherine Verfaillie, a molecular biologist then at the University of Minnesota in Minneapolis. But others have had difficulty reproducing such findings. Obokata started the current project in the laboratory of tissue engineer Charles Vacanti at Harvard University in Cambridge, Massachusetts, by looking at cells that Vacanti’s group thought to be pluripotent cells isolated from the body5. But her results suggested a different explanation: that pluripotent cells are created when the body’s cells endure physical stress. “The generation of these cells is essentially Mother Nature’s way of responding to injury,” says Vacanti, a co-author of the latest papers2, 3.

One of the most surprising findings is that the STAP cells can also form placental tissue, something that neither iPS cells nor embryonic stem cells can do. That could make cloning dramatically easier, says Wakayama. Currently, cloning requires extraction of unfertilized eggs, transfer of a donor nucleus into the egg, in vitro cultivation of an embryo and then transfer of the embryo to a surrogate. If STAP cells can create their own placenta, they could be transferred directly to the surrogate. Wakayama is cautious, however, saying that the idea is currently at “dream stage”.

Obokata has already reprogrammed a dozen cell types, including those from the brain, skin, lung and liver, hinting that the method will work with most, if not all, cell types. On average, she says, 25% of the cells survive the stress and 30% of those convert to pluripotent cells — already a higher proportion than the roughly 1% conversion rate of iPS cells, which take several weeks to become pluripotent. She now wants to use these results to examine how reprogramming in the body is related to the activity of stem cells. Obokata is also trying to make the method work with cells from adult mice and humans.

“The findings are important to understand nuclear reprogramming,” says Shinya Yamanaka, who pioneered iPS cell research. “From a practical point of view toward clinical applications, I see this as a new approach to generate iPS-like cells.”

Nature 505, 596 (30 January 2014) doi:10.1038/505596a Read the related News & Views article.

Stimulus-triggered fate conversion of somatic cells into pluripotency

Haruko Obokata, Teruhiko Wakayama, Yoshiki Sasai, Koji Kojima, Martin P.Vacanti, Hitoshi Niwa, Masayuki Yamato & Charles A. Vacanti

Here we report a unique cellular reprogramming phenomenon, called stimulus-triggered acquisition of pluripotency (STAP), which requires neither nuclear transfer nor the introduction of transcription factors. In STAP, strong external stimuli such as a transient low-pH stressor reprogrammed mammalian somatic cells, resulting in the generation of pluripotent cells. Through real-time imaging of STAP cells derived from purified lymphocytes, as well as gene rearrangement analysis, we found that committed somatic cells give rise to STAP cells by reprogramming rather than selection. STAP cells showed a substantial decrease in DNA methylation in the regulatory regions of pluripotency marker genes. Blastocyst injection showed that STAP cells efficiently contribute to chimaeric embryos and to offspring via germline transmission. We also demonstrate the derivation of robustly expandable pluripotent cell lines from STAP cells. Thus, our findings indicate that epigenetic fate determination of mammalian cells can be markedly converted in a context-dependent manner by strong environmental cues.

Bidirectional developmental potential in reprogrammed cells with acquired pluripotency

Haruko Obokata, Yoshiki Sasai, Hitoshi Niwa, Mitsutaka Kadota,Munazah Andrabi, Nozomu Takata, Mikiko Tokoro, Yukari Terashita,Shigenobu Yonemura, Charles A. Vacanti & Teruhiko Wakayama

We recently discovered an unexpected phenomenon of somatic cell reprogramming into pluripotent cells by exposure to sublethal stimuli, which we call stimulus-triggered acquisition of pluripotency (STAP)1. This reprogramming does not require nuclear transfer2, 3 or genetic manipulation4. Here we report that reprogrammed STAP cells, unlike embryonic stem (ES) cells, can contribute to both embryonic and placental tissues, as seen in a blastocyst injection assay. Mouse STAP cells lose the ability to contribute to the placenta as well as trophoblast marker expression on converting into ES-like stem cells by treatment with adrenocorticotropic hormone (ACTH) and leukaemia inhibitory factor (LIF). In contrast, when cultured with Fgf4, STAP cells give rise to proliferative stem cells with enhanced trophoblastic characteristics. Notably, unlike conventional trophoblast stem cells, the Fgf4-induced stem cells from STAP cells contribute to both embryonic and placental tissues in vivo and transform into ES-like cells when cultured with LIF-containing medium. Taken together, the developmental potential of STAP cells, shown by chimaera formation and in vitro cell conversion, indicates that they represent a unique state of pluripotency.