Continuous expression of reprogramming factors induces and maintains mouse pluripotency without specific growth factors and signaling inhibitors

The Tet‐On‐OSKM mice were described in previous studies.32, 33 They carry a doxycycline (DOX)‐inducible reverse tetracycline trans‐activator (M2rtTA) in the Rosa26 locus, and a single polycistronic OSKM transgene in the Col1a1 locus. Oct4‐GFP mice carried a GFP under control of the endogenous Oct4 distal promoter. Tet‐On‐OSKM mice and Oct4‐GFP mice were both obtained from the Jackson Laboratory. The SCID mice used for teratoma formation were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. All experiments involving animals were approved by the Institutional Animal Use Committee of the Institute of Zoology, Chinese Academy of Sciences in Beijing.

The Tet‐On‐OSKM/Oct4‐GFP mouse ESC line was derived from the blastocysts obtained from crossbreeding of the above DOX‐OSKM mice and Oct4‐GFP mice according to standard procedures. The cells were derived and further cultured in 2iL medium on the mitomycin‐c treated mouse embryonic fibroblast (MEF) cells (feeder cells). The detail components of slightly modified N2B27 medium14 were listed in Table S1. The 2iL medium14 contained N2B27 medium with the addition of PD0325901 (Stemgent, 04‐0006), CHIR99021 (Stemgent, 04‐0004), and LIF (Millipore, ESG1007). ESCs cultured in 2iL medium were switched into three different types of medium: N2B27 with 2i and LIF as the 2iL group, N2B27 with 2 μg/mL DOX (Sigma‐Aldrich) as the OSKM group, and N2B27 as the N2B27 group.

To generate induced pluripotent stem cells (iPSCs), Tet‐On‐OSKM MEFs were seeded onto the feeder cells at a density of 20000 cells per well in 6‐well‐plates. There were two types of induction media. The induction medium of the control group comprised 2iL with DOX. Once iPS‐like clones were picked up to culture in new dishes, DOX was withdrawn. The induction medium of OSKM group consisted of N2B27 with DOX, and the DOX was continued to be supplied throughout daily culture.

The PiggyBac (PB) transposon system was used. A PB transposase enzyme (PBase) vector and a PB‐GFP vector were constructed. The PBase vector contained the EF1a promoter and the coding sequence of the PBase. The CAG promoter, 3 × HA, GFP, and polyA sequences were cloned into the PiggyBac backbone to form the PB‐GFP vector (). These two vectors were transfected into OSKM‐iPSCs by using the Neon transfection system (Invitrogen, MPK5000). After 3 days, GFP‐positive cells were sorted via fluorescence‐activated cell sorting (FACS) and were seeded into 6‐well plates. Approximately 6‐8 days later, clones with all cells inside expressing GFP were collected. Figure S4A

To generate growth curves for ESCs and iPSCs, the Cell Counting Kit‐8 (CCK‐8, Sigma‐Aldrich, 96992) was used. After seeding 2500 cells/well in a 48‐well dish, a 1/10 volume of CCK‐8 solution was added to the medium for a two‐hour‐incubation at days 1, 2, 3, 4, and 5. The absorbance of each well at 450 nm was measured using a microplate reader. All the experiments were performed in quadruplicate.

Cells were incubated with 0.05 μg/mL Colcemid for 2‐3 hours. After trypsinization, cells were suspended in 0.075 M KCl at 37°C for 30 minutes. Then, the cells were fixed in solution consisting of methanol and acetic acid (3:1 in volume) for 30 minutes on ice and were dropped onto precooled slides. The cells were stained with Giemsa stain (Sigma‐Aldrich, GS500ML) under standard procedures.

Cells were fixed with 4% paraformaldehyde (PFA) for 10 minutes and were subsequently permeabilized and blocked with 0.5% Triton X‐100 (Sigma‐Aldrich) plus 2% BSA (Sigma‐Aldrich, A7906‐100G) for 1 hour. Then, cells were incubated in primary antibody solution overnight at 4℃ and secondary antibodies (donkey anti‐rabbit, Invitrogen, A21206) at room temperature for 1 hour. The primary antibodies were as follows: anti‐OCT4 (Abcam, ab19857), anti‐NANOG (Abcam, ab80892), anti‐SOX2 (Abcam, ab97959), anti‐SSEA1 (Abcam, ab16285), and anti‐TUJ1 (Biolegend, 802001). DNA was stained with Hoechst 33342 (Thermo Fisher Scientific) for 10 minutes. Images were captured using a two‐photon confocal laser scanning microscope (Leica, TCS Sp8). The BCIP/NBT Alkaline Phosphatase Color Development Kit (Beyotime, C3206) was applied to perform alkaline phosphatase staining according to the manufacturer's instructions.

KOD One™ PCR Master Mix ‐Blue (TOYOBO, KMM‐201) was used for PCR. The sequences of the primer pair were listed in Table. The PCR protocol was: 95℃/1 minutes (1 cycle), 94℃/30 seconds, 70℃/45 seconds (2 cycles), 94℃/30 seconds, 68℃/45 seconds (5 cycles), 94℃/20 seconds, 66℃/1 minutes (29 cycles), and 4℃ hold. S2

Total RNA was extracted using the TRIzol reagent (Thermo Fisher Scientific, 15596018). RNA was reverse transcribed using a ReverTra Ace® qPCR RT Master Mix with gDNA Remover Kit (TOYOBO, FSQ‐301). QuantStudio 6 Pro Real‐Time PCR System (Thermo Fisher Scientific) was used to perform the real‐time quantitative PCR analysis with THUNDERBIRD SYBR^® qPCR Mix (TOYOBO, QPS‐201) plus 50 × ROX reference dye (TOYOBO, QPS‐201). All these kits were used in accordance with the manufacturer's guidelines. Equal loading was achieved by amplifying GAPDH mRNA. The primers used were listed in Table S3. All reactions were conducted in triplicate.

Total RNA was extracted from cells with the TRIzol reagent (Thermo Fisher Scientific, 15596018). The Illumina platform was applied for RNA‐Seq. Reads were aligned to the mouse reference genome assembly (GRCm38/mm10) using STAR (version 2.7.1a)34 with default parameters, and a customized script was used to filter the uniquely mapped reads. The normalized gene expression level (Fragments Per Kilobase Million or FPKM) was obtained using Stringtie (version 2.0),35 and the analysis of differentially expressed genes (DEGs) was performed using Cuffdiff (version 2.2.1) with default parameters where two‐fold changes of gene FPKM and P‐value < .05 from the Cuffdiff application (http://cole‐trapnell‐lab.github.io/cufflinks/↗) were regarded as the cutoff values. All‐gene FPKMs were transformed by log2 and used to create scatterplots by R. Gene‐enrichment and functional annotation analysis was performed using the David tool.36 Heatmaps were generated using log‐transformed gene FPKM with the R “pheatmap” function (https://cran.r‐project.org/web/packages/pheatmap/index.html↗). After the log transformation, gene FPKM values (larger than 1 in at least one sample) were used in principal component analysis (PCA) with the R “prcomp” function and in the hierarchical clustering analysis with the R “hclust” function. Published RNA‐Seq data were downloaded from NCBI GEO GSE97954↗,37 and the R “sva” function was used to eliminate the batch effect.38

Cells were trypsinized and collected, and cell suspensions were injected subcutaneously into the flanks of SCID mice. Approximately 1 × 10⁷ cells were used in each injection. Teratomas were excised about one month later and were fixed and stained with hematoxylin and eosin (H&E) to perform histological analysis. For embryoid body (EB) formation, cells were harvested by trypsinization and seeded into bacterial culture dishes in the N2B27 medium.

Blastocyst injection was performed according to the procedures given in a previous report.39 Briefly, diploid blastocysts were collected from the uterus of 3.5 days post coitum (dpc) super‐ovulated female CD‐1 mice after mating with male CD‐1 mice. Cells were harvested by trypsinization and 12‐15 cells were microinjected into each blastocyst. After 1‐4 hours of culture, these processed embryos were transferred into the oviduct of pseudo‐pregnant CD‐1 mice at 0.5 dpc. Chimeras were identified by GFP expression or coat colors.

For statistical analysis, Students’ t test was used in GraphPad Prism 8 software. In all figures: *P value < .05; **P value < .01; ***P value < .001; ****P value < .0001.

To test whether continuous OSKM expression could maintain the cell identity of mouse ESCs, we used an ESC line stably expressing the pluripotency reporter Oct4‐DE‐GFP (GFP under control of the endogenous Oct4 distal promoter) and the single polycistronic OSKM transgene upon the addition of DOX (Figure 1A). This cell line was derived and cultured in N2B27 medium supplemented with the MEK inhibitor PD0325901, GSK3 inhibitor ChIR99021, and LIF (2iL medium). Next, the culture medium was switched from 2iL to N2B27 with DOX (OSKM medium) and the OSKM transgene was continuously activated (Figure 1A), whereas N2B27 medium alone was used as the negative control condition, and 2iL medium acted as the positive control condition. The ESCs in the 2iL medium (hereinafter referred to 2iL‐ESCs) maintained their typical ESC morphology and >90% of the cell population was Oct4‐GFP positive (Figure 1B‐D). However, ESCs cultured in N2B27 medium rapidly underwent cell differentiation and high rates of death, barely surviving after day 11 (Figure 1B,C). Nevertheless, ESCs cultured in OSKM medium (OSKM‐ESCs) survived and finally maintained typical ESC morphology for more than 77 days (38 passages) and 40%‐60% of cells were Oct4‐GFP positive (Figure 1B‐D). Different dosages of DOX exerted different effects on cell survival and generated different proportions of Oct4‐GFP, where 2 μg/mL was determined as the optimal concentration (Figure S1A). ESCs cultured in OSKM medium continued to proliferate for over 38 passages, although the growth rate was slower than that of 2iL‐ESCs (Figure 1B,E).

In summary, these results demonstrated that OSKM expression maintained self‐renewal of mouse ESCs after withdrawal of 2iL. The cell cycle analysis indicated that OSKM‐ESCs and 2iL‐ESCs shared similar distributions of G1, S, and G2/M phases (Figure 1F). Next, we tested whether OSKM‐ESCs could maintain pluripotency. The cells were alkaline phosphatase (AP) positive (Figure S1B) and expressed typical ESC markers OCT4, SOX2, NANOG, and SSEA1 by immunofluorescent staining (Figure 1G). Importantly, OSKM‐ESCs formed differentiated teratomas with structures of all three germ layers (Figure 1H). Thus, induced continuous expression of OSKM could maintain self‐renewal and pluripotency of mouse ESCs without the support of PSC‐specific culturing factors.

Next, we determined whether this system could reconstruct pluripotency de novo from MEF cells obtained from the above Tet‐On‐OSKM mice. MEF cells were reprogrammed by OSKM medium (N2B27 supplemented with DOX) and 2iL (N2B27 supplemented with 2iL and DOX, and DOX was withdrawn after reprogramming), respectively. After 14‐18 days of reprogramming (Figure 2A), AP‐positive clones were observed in both groups (Figure 2B,D and Figure S2A). Several clones were randomly selected from the OSKM group. The resulting cell lines could proliferate without visible differentiation beyond Passage 23 (Figure 2C). Cell lines carried normal karyotypes of 40 chromosomes (Figure 2E,F, and Figure S2B). ESC markers, such as OCT4, SOX2, NANOG, and SSEA1 were positive in OSKM‐iPSCs (Figure 2G). They successfully differentiated into structures of all three germ layers, including ectoderm (Figure 2H and Figure S2C,D), mesoderm (Figure 2H), and endoderm (Figure 2H).

We performed RNA‐Seq to further characterize the OSKM‐ESCs and OSKM‐iPSCs. Global gene expression profiles of OSKM‐iPSCs were quite similar to 2iL‐iPSCs (Figure 3A). And likewise, gene expression profiles of OSKM‐ESCs were similar to those of 2iL‐ESCs (Figure 3B). Thus, the global gene expression profiles of OSKM PSCs were similar to 2iL PSCs. Next, the DEGs were analyzed. In total, 644 upregulated genes overlapped in both the OSKM‐ESCs versus 2iL‐ESCs and OSKM‐iPSCs versus 2iL‐iPSCs comparisons (1,235 upregulated genes in OSKM‐ESCs and 1176 ones in OSKM‐iPSCs) (Figure 3C). These DEGs were enriched in pathways such as pathway of proteoglycans in cancer and pathway of focal adhesion (Figure 3D). Conversely, 308 DEGs were downregulated both in OSKM‐ESCs and OSKM‐iPSCs (Figure 3E), and they were enriched in signaling pathways regulating the pluripotency of stem cells (Figure 3F). Thus, we compared the expression levels of naïve pluripotency genes of OSKM‐ESCs and OSKM‐iPSCs to those of 2iL‐ESCs and 2iL‐iPSCs. As shown in the heatmap, naïve pluripotency genes, such as Klf2, Esrrb, Nanog, and Rex1 were consistently expressed to a lower degree in OSKM‐iPSCs compared to 2iL‐iPSCs, while expression levels were higher than those in somatic cells (Figure 3G). Real‐time quantitative PCR confirmed these results (Figure 3H), which were also observed in OSKM‐ESCs when compared to those of 2iL‐ESCs (Figure S3A). The RNA levels of Sox2 and Klf4 were comparable to those of 2iL‐iPSCs, although Oct4 expression was slightly lower than that in 2iL‐iPSCs; and endogenous Oct4, Sox2, and Klf4 were activated, albeit endogenous Klf4 to a lower degree (Figure S3B).

The PCA analysis and hierarchical clustering analysis (Figure 3I and Figure S3C) revealed that both OSKM‐ESCs and OSKM‐iPSCs were clearly distinguished from MEF cells. They were similar in profile in 2iL‐ESCs and 2iL‐iPSCs, and closer to serum/L‐ESCs (ESCs cultured under serum plus LIF). Furthermore, the OSKM‐ESCs collected at different culture timepoints (days 18, 25, and 38) exhibited similar gene expression profiles, indicating the system could maintain cell states stably throughout culture. In conclusion, OSKM sustained an alternative pluripotency state.

One major application of PSCs is to produce gene‐edited animal models for genome function research, disease modeling and drug screening. For this purpose, we developed a route to produce chimera mice with germline transmission using transgenic OSKM‐iPSCs (Figure 4A). We randomly selected two OSKM‐iPSC lines and inserted the GFP transgene into their genome using the PiggyBac transposon vector (Figure S4A) to examine their ability to produce chimeras. The resulting GFP‐transgenic subclones were collected (Figure 4B) and were further injected into mouse blastocysts to form chimeras. We dissected embryos at 12.5 days post coitum (dpc), and a chimera embryo with germline chimerism was observed (Figure 4C). Genotyping PCR results in further showed that OSKM‐iPSCs contributed to various organs and tissues (Figure S4B). Chimerism in the newborn and adult mouse was also observed, and genotyping PCR results confirmed the contribution of transgenic OSKM‐iPSCs (Figure 4D,E, and Figure S4C).

The accession number for the RNA‐Seq data reported in this paper is GEO: GSE173471↗.