Tet1 and Tet2 regulate 5-hydroxymethylcytosine production and cell lineage specification in mouse embryonic stem cells

Abstract

TET family enzymes convert 5-methylcytosine (5mC) to 5-hydroxymethylcytosine (5hmC) in DNA. Here, we show that Tet1 and Tet2 are Oct4-regulated enzymes that together sustain 5hmC in mouse embryonic stem cells (ESCs) and are induced concomitantly with 5hmC during reprogramming of fibroblasts to induced pluripotent stem cells. ESCs depleted of Tet1 by RNAi show diminished expression of the Nodal antagonist Lefty1 and display hyperactive Nodal signaling and skewed differentiation into the endoderm-mesoderm lineage in embryoid bodies in vitro. In Fgf4- and heparin-supplemented culture conditions, Tet1-depleted ESCs activate the trophoblast stem cell lineage determinant Elf5 and can colonize the placenta in midgestation embryo chimeras. Consistent with these findings, Tet1-depleted ESCs form aggressive hemorrhagic teratomas with increased endoderm, reduced neuroectoderm, and ectopic appearance of trophoblastic giant cells. Thus, 5hmC is an epigenetic modification associated with the pluripotent state, and Tet1 functions to regulate the lineage differentiation potential of ESCs.

Figure 1. Tet1 and Tet2 regulate 5hmC levels in mouse ES cells and are associated with the pluripotent state

(A) Normalized transcript copy numbers of Tet1, Tet2, Tet3 and Oct4 in v6.5 mouse ES cells, determined using absolute standard curves of plasmid templates. Gapdh transcript levels were used to estimate RNA content. (B) Densitometric measurement of 5’-hydroxymethyl-dCMP (5hm-dCMP) spot intensities detected on thin layer chromatography (TLC) analysis of enriched CpG sites in the genome of ES cells after 5 days of siRNA transfection to deplete Tet1 and/or Tet2. Values are depicted as percentages of total dCMP species comprising 5-methyl-dCMP (5m-dCMP), dCMP and hm-dCMP. Error bars indicate SD of 7 replicates from 3 independent experiments. siRNA treatments are denoted: C, control; T1, Tet1 SMARTpool; T2, Tet1 SMARTpool; T1+T2, combined Tet1+Tet2 SMARTpool. A representative TLC autoradiogram is shown in Figure S1B. For ES cell morphology and knockdown efficiency of Tet1 and Tet2, see Figure S1. (C) Measurement of 5hmC levels in genomic DNA from transfected ES cells and HEK293T based on dot blot analysis. Values are expressed relative to levels in control-transfected ES cells. HEK 293T cells were transfected with either TET1-catalytic domain (CD) or mock-transfected (−). Error bars indicate SD of 5 replicates from 2 experiments. In (B) and (C), P values derived from ANOVA with Bonferroni’s multiple comparison test are denoted: *, P < 0.05; ***, P < 0.001. (D) Quantitative RT-PCR measurement of Oct4, Tet1, Tet2 and Tet3 transcript levels in mouse ES cells cultured for 1–4 days on gelatinized (feeder-free) plates in the presence of LIF, or differentiated by LIF withdrawal, or LIF withdrawal plus addition of 1 µM all-trans retinoic acid (RA). Data are represented as mean ± SEM from 3 independent experiments. (E) Measurement of hmC levels by TLC (left) and dot blot (right) in genomic DNA extracted from ES cells after 4 days of LIF withdrawal and treatment with RA. Error bars indicate SD of 3–5 replicates. (F) Quantitative RT-PCR of Tet1, Tet2 and Tet3 transcript levels expressed relative to levels in ES cells (marked as dotted lines), during reprogramming of MEFs by viral transduction of Oct4, Sox2, Klf4 and c-Myc into iPS cell clones Error bars indicate SD (n=3). (G)TLC (left) and hmC dot blot (right) analyses showing 5hmC detected in iPS cells but not in fibroblasts. hmC levels from dot blots are expressed relative to levels in ES cells. Error bars indicate SD of 3–5 replicates. P values in (E) and (G) are derived from t-tests.

Figure 2. Tet mRNA levels are regulated by Oct4

(A) Quantitative RT-PCR of Tet1, Tet2 and Tet3 transcript levels in ES cells transfected for 5 days with Oct4, Sox2 and Nanog SMARTpool siRNA. Normalized transcript levels are expressed relative to levels at the start of transfection. Data are mean ± SEM from 3 experiments. P values derived from ANOVA with Bonferroni’s multiple comparison test are denoted: *, P < 0.05; **, P < 0.01; ***, P < 0.001. For ES cell morphology and knockdown efficiency of Oct4, Sox2 and Nanog, see Figure S2. (B) TLC analysis of genomic DNA purified from mouse ES cells differentiated by Oct4 RNAi. Values of 5hmC levels are mean ± SD of 3 replicates from 2 independent experiments. (C) Vista plot of sequence conservation between the human and mouse Tet1 gene loci upstream of the first coding exon, depicting CNS regions in the vicinity of the Tet1 transcription start site and coding exon 1. Regions numbered 1 to 6 spaced at 1 kb intervals were probed by ChIP PCR. For a diagram of all the CNS regions in the Tet1 locus, see Figure S2C. (D) Oct4 binding was detected as amplification from Oct4 biotin-mediated ChIP samples from ES cells at the target sites depicted in (C). Control ChIP was performed in cells expressing only the biotin ligase (BirA). For a complete analysis of Oct4 binding to all CNS regions in the Tet1 locus, see Figure S2D. (E) The mouse-human sequence alignment at the Tet1 site 5 showing a conserved Oct4-Sox2 consensus element (boxed). The Oct4 consensus sequence is highlight in red. (F) Sequence conservation between the human and mouse Tet2 gene loci upstream of the first coding exon. Regions numbered 1 to 5 spaced at 1 kb intervals were probed. (G) Oct4 binding was detected as amplification from Oct4 biotin-mediated ChIP samples from ES cells at the target sites depicted in (F). (H) The mouse-human sequence alignment at the Tet2 site 4 showing a conserved Oct4-Sox2 consensus element (boxed).

Figure 3. Tet depletion selectively affects cell lineage markers and skews ES cell differentiation

(A) Effect of Tet1 (T1), Tet2 (T2), Tet3 (T3) or combined Tet1+Tet2 (T1+T2) SMARTpool siRNAs on the expression of pluripotency and selected lineage marker genes, assessed after 5 days of transfection. C, non-targeting control siRNA. Data are represented as mean ± SEM, n=3–4. *, P < 0.05; **, P < 0.01; P < 0.001 by ANOVA with a post-hoc test for comparison to control. See also Fig S3A. (B) Enhanced growth and extensive hemorrhagy in Tet1-kd tumors compared to GFP-kd controls. Refer to Fig S3 for knockdown efficiency in stable Tet-kd ES cell clones. (C) Gross histology of teratomas formed after injection of control clones, revealed through low power images of hematoxylin/eosin (H&E)-stained sections (left), showing high contribution of differentiated neuronal tissue (pink). Higher power image (right) shows mature epithelium with squamous differentiation (arrowhead), terminally differentiated neuronal tissue (block arrow) and immature glandular tissue (white arrow). (D) Low power image of a Tet1-kd teratoma (from Tet1-kd/shRNA#2.1), showing predominance of immature glandular tissue (purple). Higher power image (right) shows necrotic tissue with blood and scattered giant cells (black arrow) and glandular tissue (white arrow). (E) Tet1-kd teratoma showing highly proliferative glandular tissues. Inset: a cell undergoing mitosis is marked with asterisk. (F) Tet1-kd teratoma showing a cluster of trophoblastic giant cells (arrows). (G) Periodic acid-Schiff staining of a serial section from the Tet1-kd tumor shown in (G), showing glycogen-rich granules (purple stain) in giant cells (arrows). (H) Hemorrhagy in Tet2-kd tumors compared to scrambled-kd controls. (I) Low power image of H&E stained Tet2-kd teratoma (from Tet2-kd/shRNA#3), showing areas of fully differentiated neuronal tissue (pink) and glandular tissue (purple). Inset: neuronal tissue. Histology of Tet1-kd shown is representative of tumors from Tet1-kd/shRNA#2 (each clone #2.1 and #2.2 injected into 2 mice per experiment) from 3 independent experiments performed with 2 different strains of immunodeficient mice. Each Tet2-kd clone was injected in 2 mice per experiment in 2 independent experiments.

Figure 4. Tet1 depletion in ES cells predisposes cells to differentiate along a trophoblastic lineage

(A) Time-course of expression of Elf5 in control and Tet1-kd clones grown in TS conditions for 1–4 weeks. Data represent mean ± SEM of 4 independent experiments. (B) Expression of trophectoderm markers Cdx2, Eomes and Elf5 in the Tet1-kd/shRNA#2 line and subclones cultured in TS cell conditions for 4 weeks compared to parental cells in TS or ES cell conditions. Normalized transcript levels in (B–C) are expressed relative to levels in TS cells (set as 100). (C) Mid-gestation embryo chimerism of GFP-labeled control (scrambled shRNA) or Tet1-knockdown (Tet1-kd/shRNA#2) cells injected into blastocysts after culture in ES or TS conditions. Brightfield (left) and wholemount GFP fluorescence (right) images were taken at E10.5. Abbreviations: e, embryo; ys, yolk sac and p, placenta. (D) Scoring of GFP presence in embryos and placentas at midgestation. (E) GFP antibody staining of placental section showing chimerism of a Tet1-kd subclone (sc3) from TS cell culture.

Figure 5. RNAi depletion of Tet1 in ES cells skews differentiation towards the endoderm-mesoderm lineages in vitro

(A) Expression of CD4 and GFP in CD4-Foxa2/GFP-Bry ES cells transfected with Tet1 siRNA and differentiated in serum-free media for 4 days to form EB. EB cultures were reaggregated and treated with Activin A (Act) at the indicated concentrations at Day 2. Numbers in quadrants denote percentage of gated cell populations. (B) Percentages of CD4-high and GFP-high cell populations in Day 4 EB after siRNA treatment. Three independent experiments are shown comparing control and Tet1 siRNA#2 treatment. (C) Quantitative RT-PCR analysis of Foxa2 and Brachyury in Day 4 EB differentiated from Tet1-kd/shRNA#2 stable clones. Data are mean ± SEM of 3 clones in each group, representative of 2 independent experiments. (D) Western blot analysis of phosphor-Smad2, total Smad2/3, Eomes and Lefty in whole-cell lysates of Day 4 EB.

Figure 6. Tet1 depletion has different effects on DNA methylation at target gene promoters

(A) Bisulphite sequencing analysis of CpG methylation status at the Lefty1 promoter in ES cells transfected with control or Tet1 SMARTpool siRNA. The average % methylation at each CpG site is derived from sequencing of 20–24 clones. (B) Bisulphite sequencing analysis of the Elf5 promoter in ES cells, TS cells and various Tet1-kd ES clones in TS culture condition. The % methylation at each CpG site is derived from sequencing of 10–12 clones. Three Tet1-kd/shRNA#2.1 subclones were analyzed and errors bars are mean ± SEM.