Promotion of reprogramming to ground state pluripotency by signal inhibition

Abstract

Induced pluripotent stem (iPS) cells are generated from somatic cells by genetic manipulation. Reprogramming entails multiple transgene integrations and occurs apparently stochastically in rare cells over many days. Tissue stem cells may be subject to less-stringent epigenetic restrictions than other cells and might therefore be more amenable to deprogramming. We report that brain-derived neural stem (NS) cells acquire undifferentiated morphology rapidly and at high frequency after a single round of transduction with reprogramming factors. However, critical attributes of true pluripotency--including stable expression of endogenous Oct4 and Nanog, epigenetic erasure of X chromosome silencing in female cells, and ability to colonise chimaeras--were not attained. We therefore applied molecularly defined conditions for the derivation and propagation of authentic pluripotent stem cells from embryos. We combined dual inhibition (2i) of mitogen-activated protein kinase signalling and glycogen synthase kinase-3 (GSK3) with the self-renewal cytokine leukaemia inhibitory factor (LIF). The 2i/LIF condition induced stable up-regulation of Oct4 and Nanog, reactivation of the X chromosome, transgene silencing, and competence for somatic and germline chimaerism. Using 2i /LIF, NS cell reprogramming required only 1-2 integrations of each transgene. Furthermore, transduction with Sox2 and c-Myc is dispensable, and Oct4 and Klf4 are sufficient to convert NS cells into chimaera-forming iPS cells. These findings demonstrate that somatic cell state influences requirements for reprogramming and delineate two phases in the process. The ability to capture pre-pluripotent cells that can advance to ground state pluripotency simply and with high efficiency opens a door to molecular dissection of this remarkable phenomenon.

Figure 1. NS cells Undergo Rapid but Incomplete Conversion Towards a Pluripotent Phenotype

(A) Phase contrast images of NS cells in standard culture (upper) and 5 d after infection with the four factors (lower). (B) RT-PCR analyses for Nanog, endogenous (endo) Oct4, Fgf4, Rex1, and Blbp in infected foetal (fNS) and adult (aNS) cells 3 and 5 d d after infection. MEF infections under both MEF and NS cell culture conditions were analysed in parallel. Only at later time points, after 10 d, did a few colonies of ES cell-like morphology emerge from infected MEFs. (C) Flow cytometry analysis of Oct4-GFP activation in NS cells 5 d after infection. (D) Phase and fluorescence images of FACS purified GFP-expressing cells cultured on feeders. (E and F) Conventional (E) and quantitative (F) RT-PCR analyses of Oct4GFP expressing cells expanded on feeders. (G) Immunofluorescence staining for Oct4 and Nanog. (H) Staining for me3K27. Arrowheads indicate the nuclear body diagnostic of the inactive X chromosome.

Figure 2. MEK and GSK3 Inhibitors (2i) Promote Reprogramming to Full Pluripotency

(A) Example of 2i-iPS cell colonies generated from aNS cells replated at day 5 after infection in 2i medium. (B) Plates containing control and infected aNS cells and MEFs were cultured in 2i media from day 3 or day 5 after infection and stained for AP 10 d later. Numbers of AP-stained colonies are indicated. (C) Established 2i-iPS cell line. (D and E) Conventional (D) and quantitative (E) RT-PCR analyses of 2i-iPS cells generated from aNS and fNS Oct4GFP NS cells, and from fNS cells of non-transgenic C57BL/6 strain background. (F) Immunofluorescence staining for Oct4 and Nanog in 2i-iPS cell colony after first passage. (G) Immunostaining forme3K27. A cluster of residual Oct4-negative cells retains the nuclear body of an inactive X chromosome, providing an internal control. (H) Chimera generated from injection of 2i-iPS (aNS) cells into C57BL/6 host blastocyst.

Figure 3. 2i/LIF Induces Transition to Ground State Pluripotency

(A and B) MEF and NS cell–derived pre-iPS cell clones were plated at clonal density in replica wells and cultured on a feeder layer in complete medium. After 7 d, colonies were scored and medium replenished (control) or switched to 2i. (A) Representative images of colonies 13 d after plating. Oct4 reporter (GFP) expression indicates pluripotent status. (B) Number of colonies exhibiting Oct4 reporter activation (GFP) in the indicated culture conditions. (C) RT-PCR analysis for Nanog and Rex1 in MEF and NS clones in control (well 3) and 2i conditions (wells 1 and 2). (D) Activation of Oct4GFP reporter after transfer of a pre-iPS cell colony to serum-free 2i/LIF. Duration of culture in 2i is indicated on the left. (E) Emergence of GFP-expressing colonies (arrowheads) after the addition of MEK inhibitor (i) or both MEKi and GSKi (ii) to pre-iPS cells in the presence of serum and LIF. Serum-free 2i/LIF treated cells are shown for comparison.

Figure 4. Sox2 Reduces Efficiency of NS Cell Conversion to Pluripotency

(A) Genomic PCR analysis for retroviral transgenes in 2i-iPS clones generated from NS cells that were infected with the four factors. (B) Southern blot analysis for Sox2, Oct4, Klf4, and Myc retroviral integrations in fNS 2 cells. Arrowheads indicate bands corresponding to retroviral integrations and red dots those of the endogenous genes. Multiple bands in the Sox2 and Oct4 lanes are attributable to cross-hybridisation and pseudogenes, respectively. (C) Chimera generated from injection of fNS 2 2i-iPS cells into C57BL/6 host blastocyst with C57BL/6 mate and pups. Offspring coat colour demonstrates transmission of an fNS2-derived haploid genome. (D) Genomic PCR analysis for retroviral segregation in the offspring. Oct4 reporter (GFP), which is homozygous in fNS 2 cells, is used as a positive control. Absence of Oct4 and Klf4 amplification product in some offspring confirms the low transgene copy numbers. (E–G) Comparison of the reprogramming efficiency of aNS cells infected with either four factors or three factors. 8 × 10⁵ aNS cells were plated and medium switched to 2i five days after infection. (E) Examples of emerging colonies. Oct4 reporter (GFP) expression indicates pluripotent status. (F) Colony counts performed at day 8 after 2i switch. (G) Plates stained for AP 11 d after 2i switch. Puromycin selection was applied 3 d prior to staining. Reprogramming efficiency was calculated taking into account total number of GFP expressing colonies at day 11, 371, and 1137 for four- and three-factors plates, respectively, the number of plated cells, and assuming a 42% probability for a given cell being infected simultaneously with Oct4, Klf4, and c-Myc. (H) RT-PCR analysis for pluripotency markers in 2i-iPS cells (–Sox2) derived from aNS and fNS of Oct4-GFP mixed background and C57BL/6 inbred backgrounds, respectively. (I and J) Immunofluorescence staining for Oct4 and me3K27 (I) or Nanog (J) in 2i-iPS cells generated without exogenous Sox2.

Figure 5. Klf4 and Oct4 Are Sufficient to Induce NS Cell Conversion to a Pre-iPS Cell State That Can Be Further Converted by 2i/LIF

(A) Pre-iPS cell colonies 3 wk after infection of aNS cells with Oct4 and Klf4. (B) Appearance of clusters of cells expressing Oct4-GFP within established pre-iPS colonies, after switch into 2i/LIF. (C) Established 2i-iPS cell line generated by two-factor infection. (D) RT-PCR analysis for pluripotency markers in two-factor 2i-iPS cells. (E) Genomic PCR confirms retroviral integration of Oct4 and Klf4 only. (F and G) Immunofluorescence staining for Oct4 and me3K27 (F) or Nanog (G) in 2i-iPS cells generated without exogenous Sox2 and cMyc. (H) Adult chimera generated from injection of adult NS cell–derived two-factor 2i-iPS cells into C57BL/6 host blastocyst. (I) Schematic comparison of the reprogramming dynamics of NS cells infected with either four, three (–Sox2), or two (–Sox2 and –cMyc) factors.

Figure 6. Transfection-Induced Reprogramming Commonly Stalls at a Pre-Pluripotent Stage and Can Be Advanced to Completion by Mek/Erk Inhibition

Transduction of somatic cells with Oct4, Klf4, c-Myc, and Sox2 transgenes gives rise at early stages to cells that have undifferentiated morphology and express some markers of ES cells. These first appearing ES cell–like cells do not convert to full pluripotency, even if continuously propagated. In the case of NS cells, partially reprogrammed cells are evident just 3 d after transduction The block to full pluripotency can be released by applying small molecule inhibitors (2i) of the Mek/Erk pathway and of glycogen synthase kinase-3 in the presence of LIF. Transition to authentic iPS cell status occurs rapidly and at high efficiency. Alternatively, 2i/LIF may be applied shortly after transduction before emergence of pre-iPS cell colonies. In this case fully reprogrammed iPS cells are isolated directly, reflecting rapid transit through, or possibly bypass of, the pre-iPS cell stage. In either scenario use of 2i enables isolation of authentic iPS cells without recourse to reporter genes or unreliable morphological assessment. In addition, reprogramming of neural stem cells to ground state pluripotency in the presence of 2i/LIF is not enhanced by exogenous Sox2 and may proceed with only 1–2 integrations of Oct4, Klf4, and, optionally, c-Myc. Although, germline competent iPS cells can be obtained in standard ES cell culture conditions without use of 2i. This occurs at low frequency over 3 wk or longer. We surmise that this low-efficiency “stochastic” pathway avoids passing through or becoming trapped in the pre-iPS cell stage.