Tissue collection and low-pH treatment
To isolate CD45+ haematopoietic cells, spleens were excised from 1-week-old Oct4-gfp mice (unless specified otherwise), minced by scissors, and mechanically dissociated using a Pasteur pipette.
(i) Adherent cells should be dissociated into single cells, either mechanically or enzymatically (by trypsin or collagenase). For the tissues described in Fig. 3a (Obokata et al. Nature, 2014a), muscle, adipose tissue and fibroblasts were enzymatically dissociated, whereas others were mechanically dissociated.
(ii) Primary cells should be used. We have found that it is difficult to reprogram mouse embryonic fibroblasts (MEF) that have been expanded in vitro, while fresh MEF are competent.
(iii) For the experiments reported, we used a Oct-3/4-EGFP transgenic mouse line (Ohbo et al, Dev Biol, 2003; Yoshimizu et al, Dev Growth Differ, 1999), which is maintained by the RIKEN Bioresource Center as GOF18-GFP line11 transgenic mouse (B6;B6D2-Tg(GOF18/EGFP)11/Rbrc). Homozygotes of the transgene were used for the live imaging to obtain the enhanced signal.
(iv) Cells from mice older than one week showed very poor reprogramming efficiency under the current protocol. Cells from male animals showed higher efficiency than those from female.
Dissociated spleen cells were suspended with PBS and strained through a cell strainer (BD Biosciences 352340).
After centrifuging at 1,000 rpm for 5 min, collected cells were re-suspended in DMEM medium and added to the same volume of lympholyte (Cedarlane), and then centrifuged at 1,000 g for 20 min.
(i) The purity of the starting cells is important for achieving STAP conversion. For lymphocytes, contamination with red blood cells may inhibit the reprogramming event. When using adherent cells, the presence of extracellular matrix may interfere with reprogramming.
(ii) Alternatively, red blood cells may be removed by suspension of the cell pellet in 1.8 ml of H2O (Sigma W3500). After 30 seconds, add 0.2 ml of 10× PBS (Gibco 70011-044), followed by 3 ml of 1× PBS (Gibco 10010-023), and strain the cell suspension through a cell strainer.
The lymphocyte layer was isolated and stained with CD45 antibody (Abcam ab25603). CD45+ cells were sorted by FACS Aria (BD Biosciences).
(i) FACS sorting can be an important step for the confirmation of cell purity, but can affect both cell viability and reprogramming efficiency. Skipping this step may increase reprograming efficiency, although this may result in a reduction in confidence in cell identity.
After cell sorting, 1 × 106 CD45+ cells were treated with 500 µl of low-pH HBSS solution (titrated to pH 5.7 by HCl) for 25 min at 37°C, and then centrifuged at 1,000 rpm at room temperature for 5 min.
(i) The buffering action of HBSS is weak, so carry-over of the solution may affect pH. Please adjust pH to 5.7 in cell suspension by the following method. First, suspend the cell pellet with 494 µl of HBSS pre-chilled at 4°C, then add 6 µl of diluted HCl (10 µl of 35% HCl in 590 µl of HBSS) to adjust to a final pH of 5.7. Please confirm the final pH in a pilot experiment, and optimize the volume of HCl added, as necessary. Alternatively, suspend the cell pellet in HBSS-pH 5.4 pre-chilled at 4°C.
(ii) The HBSS we used is Ca2+/Mg2+ free (Gibco 14170-112).
(iii) Incubate the cells suspended in HBSS in a CO2 incubator.
(iv) Cell viability is a critical parameter in this step. Under optimal conditions, massive cell death is observed at two days after plating, as shown in Figure 1d (Obokata et al. Nature, 2014a).
(v) If you find massive cell death at one day after plating, it may be ameliorated by shortening the incubation period with low-pH HBSS solution to 15 min.
After the supernatant (low-pH solution) was removed, precipitated cells were re-suspended and plated onto non-adhesive culture plates (typically, 1×105 cells/ml) in DMEM/F12 medium supplemented with 1,000 U LIF (Sigma) and 2% B27 (Invitrogen).
(i) The use of non-adhesive culture plates is recommended, as the formation of cell clusters is an important step for reprogramming, and the adhesive surface may inhibit cell movement needed to form clusters.
(ii) Cell density is critical, and depends on cell viability. Density should be maintained at 1×105~1×106 cells per cm2 of culture surface.
(iii) B27 (Invitrogen 17504-044) may show variation between batches. Please check the quality by N2B27-2iLIF culture of ES cells.
Cell cluster formation was more sensitive to plating cell density than to the percentage of Oct3/4-GFP+ cells. The number of surviving cells was sensitive to the age of donor mice, and was low under the treatment conditions described above when adult spleens were used.
(i) The donor mouse should be 1-week old or younger. Reprogramming efficiency is dramatically reduced using cells from older animals.
(ii) STAP cells are derived from clusters of multiple cells; they are not monoclonal.
The addition of LIF during days 2–7 was essential for generating Oct3/4-GFP+ STAP cell clusters on day 7, as shown in Extended Data Fig. 1f (Obokata et al. Nature, 2014a). Even in the absence of LIF, Oct3/4-GFP+ cells (most of which showed dim signals) appeared transiently in low-pH-treated CD45+ cells during days 2–5 of culture, but subsequently disappeared, suggesting that there is a LIF-independent early phase, whereas the subsequent phase is LIF-dependent.
(i) Since LIF is essential for the late step of reprogramming, the reprogramming event may depend on the genetic background. We mainly used 129, C57BL6, or their F1 strains, as all of these genetic backgrounds are associated with high responsiveness to LIF.
(ii) The GFP signal is weaker than that of ES cells carrying the same reporter because of the smaller cell volume of STAP cell than ES cell as found in Figure 1g (Obokata et al. Nature, 2014a).
STAP stem-cell conversion culture
To establish STAP stem-cell lines, STAP cell clusters were transferred to ACTH-containing medium on MEF feeder cells (several clusters, up to a dozen clusters, per well of 96-well plates).
(i) ACTH (1-24) is available from American Peptide and other companies. We used ACTH synthesized by Kurabo on consignment. The composition of this medium is GMEM, 15% Knockout Serum ReplacementTM (KSR, Invitrogen), 1 × non-essential amino acids (NEAA), 1 × Sodium Pyruvate, 10-4M 2-mercaptoethanol, 1000 U/ml LIF, and 10 µM ACTH (Ogawa et al, Genes Cells, 2004). The STAP cell cluster was isolated, dissected into small pieces as in the case of injection into blastocysts as shown in Figure 4a (Obokata et al. Nature, 2014a), and seeded on mouse embryonic fibroblast feeder cells in the ACTH medium.
(ii) ACTH-containing medium was purchased from DS Pharma Biomedical (Osaka, Japan) as [email protected]
After 4–7 days of culture, the cells were subjected to a first passage using a conventional trypsin method, and the suspended cells were plated in ESC maintenance medium containing 20% FBS.
(i) ESC maintenance medium consists of KnockoutTM DMEM (Life Technologies), 20% FBS, 1 × NEAA, 1 × Glutamine, 1 × Nucleosides, 10-4M 2-mercaptoethanol, and 1000 U/ml LIF.
(ii) FBS lots should be confirmed for suitability for use in the culture of mouse ES cells.
(iii) We have established multiple STAP stem cell lines from STAP cells derived from CD45+ haematopoietic cells. Of eight clones examined, none contained the rearranged TCR allele, suggesting the possibility of negative cell-type-dependent bias (including maturation of the cell of origin) for STAP cells to give rise to STAP stem cells in the conversion process. This may be relevant to the fact that STAP cell conversion was less efficient when non-neonatal cells were used as somatic cells of origin in the current protocol.
Subsequent passaging was performed at a split ratio of 1:10 every second day until reaching subconfluency. We tested the following three different genetic backgrounds of mice for STAP stem-cell establishment from STAP cell clusters, and observed reproducible establishment: C57BL/6 carrying Oct4-gfp (29 of 29), 129/Sv carrying Rosa26-gfp (2 of 2), and 129/Sv × C57BL/6 carrying cag-gfp (12 of 16). STAP stem cells with all these genetic backgrounds showed chimaera-forming activity.
FI stem cell conversion culture
STAP cell clusters were transferred to Fgf4-containing trophoblast stem-cell medium (Tanaka et al, Science, 1998) on MEF feeder cells in 96-well plates (Obokata, Nature, 2014b).
(i) TS medium consists of RPMI 1640 with 20% FBS, 1 mM Sodium Pyruvate, 100 µM 2-mercaptoethanol, 2 mM L-glutamine, 25 ng/ml of recombinant FGF4, and 1 µg/ml of heparin.
(ii) Different lots of FBS may results in significant differences in the behavior of cultured cells.
In most cases (40 of 50 experiments), colonies grew in 10–50% of wells in 96-well plates. In a minority of cases (10 of 50 experiments), no colony growth was observed and/or only fibroblast-like cells appeared.
(i) The cells in proliferative colonies also appear similar to fibroblasts, but gradually change morphology, coming to resemble epithelial cells.
The cells were subjected to the first passage during days 7–10 using a conventional trypsin method. Subsequent passages were performed at a split ratio of 1:4 every third day before they reached subconfluency.
(i) The cells must not be dissociated completely. Partial dissociation is optimal to maintain viability and self-renewal, as seen in the case of embryo-derived trophoblast stem cells.