Cell Stem Cell. 2015 Jan 8;16(1):39-50. doi: 10.1016/j.stem.2014.10.019. Epub 2014 Nov 13.

Inhibition of pluripotency networks by the Rb tumor suppressor restricts reprogramming and tumorigenesis.

Kareta MS¹, Gorges LL², Hafeez S³, Benayoun BA⁴, Marro S³, Zmoos AF², Cecchini MJ⁵, Spacek D², Batista LF⁶, O'Brien M², Ng YH³, Ang CE³, Vaka D², Artandi SE⁶, Dick FA⁵, Brunet A⁴, Sage J⁷, Wernig M⁸.

Author information

1: Department of Pediatrics, Stanford University, Stanford, CA 94305, USA; Department of Genetics, Stanford University, Stanford, CA 94305, USA; Department of Pathology, Stanford University, Stanford, CA 94305, USA; Stanford Institute for Stem Cell Biology and Regenerative Medicine, Stanford University, Stanford, CA 94305, USA.
2: Department of Pediatrics, Stanford University, Stanford, CA 94305, USA; Department of Genetics, Stanford University, Stanford, CA 94305, USA.
3: Department of Pathology, Stanford University, Stanford, CA 94305, USA; Stanford Institute for Stem Cell Biology and Regenerative Medicine, Stanford University, Stanford, CA 94305, USA.
4: Department of Genetics, Stanford University, Stanford, CA 94305, USA; Paul F. Glenn Laboratories for the Biology of Aging, Stanford University, Stanford, CA 94305, USA.
5: London Regional Cancer Program, Children's Research Institute, Western University, London, ON N6A 4L6, Canada.
6: Department of Medicine, Stanford University, Stanford, CA 94305, USA; Department of Biochemistry, Stanford University, Stanford, CA 94305, USA; Stanford Institute for Stem Cell Biology and Regenerative Medicine, Stanford University, Stanford, CA 94305, USA.
7: Department of Pediatrics, Stanford University, Stanford, CA 94305, USA; Department of Genetics, Stanford University, Stanford, CA 94305, USA. Electronic address: julsage@stanford.edu.
8: Department of Pathology, Stanford University, Stanford, CA 94305, USA; Stanford Institute for Stem Cell Biology and Regenerative Medicine, Stanford University, Stanford, CA 94305, USA. Electronic address: wernig@stanford.edu.

Abstract

Mutations in the retinoblastoma tumor suppressor gene Rb are involved in many forms of human cancer. In this study, we investigated the early consequences of inactivating Rb in the context of cellular reprogramming. We found that Rb inactivation promotes the reprogramming of differentiated cells to a pluripotent state. Unexpectedly, this effect is cell cycle independent, and instead reflects direct binding of Rb to pluripotency genes, including Sox2 and Oct4, which leads to a repressed chromatin state. More broadly, this regulation of pluripotency networks and Sox2 in particular is critical for the initiation of tumors upon loss of Rb in mice. These studies therefore identify Rb as a global transcriptional repressor of pluripotency networks, providing a molecular basis for previous reports about its involvement in cell fate pliability, and implicate misregulation of pluripotency factors such as Sox2 in tumorigenesis related to loss of Rb function.

PMID:: 25467916
PMCID:: PMC4389904
DOI:: 10.1016/j.stem.2014.10.019

[Indexed for MEDLINE]

Free PMC Article

Images from this publication.See all images (6)Free text

Figure 1

Rb Loss Removes a Block to the Cellular Reprogramming of iPSCs

(A) Reprogramming of early-passage mouse and human fibroblasts. Rb was depleted by either a germline mutation (mouse)or shRNA knockdown (human). Cells were infected with 4F and stained for AP activity after 14 or 30 days in culture, respectively. Representative images are shown.
(B) Wells with AP⁺ cells and proper morphology from Oct4-GFP MEFs with lentiviral constructs expressing hairpins to Rb, p53, GFP, Luciferase, or an empty vector (ev), subjected to the 4F (see Experimental Procedures). Significance was assessed by a t test to the 4F-infected sample (n ≥ 3).
(C) Efficiency of iPSC formation was tested using Oct4-Neo^R MEFs followed by treatment with G418 on day 15 and AP staining after 5 days of selection. Significance was assessed by a t test to the 4F sample.
(D) Efficiency of iPSC formation was tested using cKO MEFs stained for AP activity and Nanog expression to identify iPSC colonies. Significance was assessed by a t test (n = 3).
(E) Efficiency of iPSC formation was tested with Rb overexpression by infection with Ad-Rb-GFP or Ad-GFP. Significance was assessed by a t test (n = 2).
(F) Efficiency was tested for Rb knockdown MEFs and knockout MEFs for each of the pocket proteins. Significance was assessed by a t test to the 4F sample (n ≥ 3).
(G) Kinetics of reprogramming was determined by infecting shRb-infected Oct4-GFP MEFs with 4F. Plates were scored by the earliest GFP⁺ colony appearance up to day 20. Significance was assessed by a t test comparing the controls against the three shRb hairpins (n ≥ 3).
(H) Analysis of early reprogramming events by SSEA1 expression in shRb (M) or control (C) infected MEFs 4 and 6 days after 4F (n = 3).
(I) Dox-dependence assay where shRb-infected Oct4-GFP MEFs were infected with 4F on day 0 and treated with Dox on day 1. The cells were switched to Dox-free media at regular intervals and GFP⁺ colonies were screened on day 15 (n ≥ 3).
All plots, unless noted, display the mean ± SD where *p < 0.05, **p < 0.01, and ***p < 0.001. See also .

Inhibition of Pluripotency Networks by the Rb Tumor Suppressor Restricts Reprogramming and Tumorigenesis

Cell Stem Cell. ;16(1):39-50.

Figure 3

Rb Represses the Expression of Pluripotency Factors

(A) RNA sequencing of Rb^lox/lox MEFs after Rb knockout by Cre expression (C) or GFP (G) followed by infection with an empty puromycin-selectable lentivirus (P) or the 4F. k-means clustering () was first performed to divide the genetic expression levels that go up upon 4F expression (Cluster I), down (Cluster III), or remain unchanged (Cluster II) (G4F versus GP). Each of these clusters was then further divided by whether they change after loss of Rb (G4F versus C4F). Genes of interest are annotated on the left and are marked red, blue, or black depending on whether their expression levels go up, down, or stay unchanged, respectively, in the C4F set (upon loss of Rb). See also .
(B) GSEA profiles comparing the G4F and C4F set to expression profiles of matched MEF and iPSCs (). The green graph shows the running enrichment score for the gene set as it is calculated running down the genes from the RNA sequencing () that are ranked by their enrichment in either the G4F or C4F sets (red/ blue heatmap). Normalized enrichment scores (NESs) and false discovery rates (FDRs) for the gene sets are reported.
(C) Gene expression in control (Con, Rb^+/+; p130^+/lox; p107^+/+), cKO (Rb^lox/lox; p130^+/+; p107^+/+), and cTKO (Rb^lox/lox; p130^lox/lox; p107^−/−) MEFs as assessed by RT-qPCR. Expression is graphed as the relative expression of Ad-Cre-infected MEFs compared to that of Ad-GFP normalized to Arppo. Statistical significance was determined by a paired t test of the Ad-Cre and Ad-GFP ΔCt values. nd, no data.
(D) Expression measured by RT-qPCR for either Rb cKO MEFs infected with Ad-GFP or Ad-Cre (n = 4) or mESCs (n = 3), shown relative to the Ad-GFP cells. Statistical significance was determined by a paired t test of the Ad-Cre MEFs and an unpaired t test for the mESCs.
(E) Expression measured by RT-qPCR is graphed as the relative expression of shRB-in-fected human fibroblasts compared to those infected with an empty vector (ev) and normalized to GAPDH. Statistical significance was determined by a paired t test of the knockdown and control ΔCt values; n = 2.
(F) FACS analysis of GFP in Sox2-GFP knockin MEFs after shRb infection compared to controls. Dashed line demarcates GFP⁺ gate.
All plots display the mean ± SD where *p < 0.05, **p < 0.01, and ***p < 0.001. See also .

Inhibition of Pluripotency Networks by the Rb Tumor Suppressor Restricts Reprogramming and Tumorigenesis

Cell Stem Cell. ;16(1):39-50.

Figure 4

Rb Directly Binds Regulatory Elements in the Oct4 and Sox2 Loci

(A) qPCR of Rb ChIP in Ad-GFP-infected cKO MEFs compared to Ad-Cre-infected MEFs and the mESCs as negative control. Primers are to the proximal promoter (). Significance was assessed by an unpaired t test to the average of the Ad-Cre-infected MEFs, including the mESC sample (n = 2).
(B) Rb (blue) binding at Sox2, Oct4, and Mcm7. Enrichment is shown as the log₂ value of the fold change between normalized ChIP and the input samples (n = 2). The transcription start sites (TSS) are noted.
(C) Gene expression by microarray of WT and TKO mESCs after their differentiation into embryoid bodies (n = 3). Pluripotency genes () and a selection of differentiation markers from endoderm, mesoderm, ectoderm, and trophectoderm are shown. These are also scored for their status as direct Rb targets.
(D) The average expression of the pluripotency genes shown in (C). Plots show the mean (horizontal line), the 25^th to the 75^th percentile (box), and the extent of the data (bars). Significance was tested by a paired t test.
(E) Rb ChIP in mESCs (d0) and after their subsequent differentiation into EBs (d3, d6, d12, and d18). Binding is shown (gray bars, left axis) as percent input normalized to a control locus to further account for cell number. Expression from the locus (either Oct4 or Sox2) is shown relative to the levels in the mESCs (blue line, right axis). Significance was determined by an unpaired t test to the respective d0 values for the ChIP and qPCR sets.
(F) Schematic of Rb interaction surfaces including the general E2F binding surface (G) and the LxCxE-binding domain (L), which interacts with other silencing factors.
(G) Reprogramming efficiency of the ΔG or ΔL mutant MEFs compared to those derived from a wild-type littermate. Efficiency was screened by AP and Nanog staining. Significance was determined by a paired t test.
(H) Gene expression in the ΔG and ΔL MEFs compared to wild-type MEFs as assessed by RT-qPCR. Expression is graphed as the relative expression of the mutant to the wild-type. Statistical significance was determined by a paired t test of the ΔCt values.
All plots, unless noted, display the mean ± SD where *p < 0.05, **p < 0.01, and ***p < 0.001. See also .

Inhibition of Pluripotency Networks by the Rb Tumor Suppressor Restricts Reprogramming and Tumorigenesis

Cell Stem Cell. ;16(1):39-50.

Figure 5

Rb Loss Perturbs the Chromatin State of Pluripotency Genes by Modulating Histone Modifier Recruitment

(A) Histone modification profiles for H3Ac (light green), H3K4me3 (dark green), and H3K27me3 (red) displayed as a heatmap. The bottom track indicates regions of significant changes for each mark by their respective color.
(B) GSEA results showing enrichment of iPSC and Rb gene sets upon gain of H3 acetylation and H3K4me3. Data shown are the rank of the gene set relative to all identified gene sets, the normalized enrichment score (NES), and the false discovery rate (FDR).
(C) Examples of genes that show an increase (blue) or decrease (gray) in H3K4me3 domain breadth upon loss of Rb. Black bars represent the extent of the domains.
(D) Significance for enrichment of transcription factor targets at genes that gained substantial H3K4me3 breadth in MEFs upon loss of Rb against the expected genome-wide value from 1,000 random samplings expressed as — log10 (p value) in one-sided Wilcoxon tests.
(E) Examples of a gene in which the H3K4me3 peak breadth in wild-type MEFs (gray) is broadened upon loss of Rb (blue), compared to mESCs (green). Horizontal black bars represent the extent of the peak.
(F) Fold-enrichment for overlap with top 5% broadest H3K4me3 domains (“buffer domains”) for genes that lost or gained substantial H3K4me3 breadth in MEFs upon loss of Rb. Enrichment and significance (one-sided Wilcoxon tests) were calculated against expected genome-wide values from 1,000 random samplings.
(G) ChIP-qPCR of HDAC and EZH2 in Ad-GFP- and Ad-Cre-infected MEFs plotted relative to the mESCs as a negative control to account for relative primer strength. qPCR was performed with primers to the DE, PE, proximal promoter (PP), and first exon (E1) of Oct4. Significance was determined by a t test.
(H) ChIP-qPCR of HDAC binding shown as in , shown for the SRR1 and SRR2 enhancers, the PP, and the E1 of Sox2.
All plots, unless noted, display the mean ± SD where *p < 0.05 and ***p < 0.001. See also .

Inhibition of Pluripotency Networks by the Rb Tumor Suppressor Restricts Reprogramming and Tumorigenesis

Cell Stem Cell. ;16(1):39-50.