PMC

Low Levels of Tbx3 Restrict Developmental Potential In Vivo
(A) Scheme of sorting strategy and 2n aggregation assay with E2.5 embryos.
(B and C) Generation of chimeric embryos by aggregating TBX3-BAC-EGFP high/low ESCs infected with LvCherry lentivirus with 2n E2.5 embryos (B) and ESCs integrated into ICM of chimeric blastocysts after 24 hr in culture (E3.5). (C) Top, phase contrast; middle, LvCherry signal for lentiviral labeling of cells to ensure visualization; bottom, GFP signal of the TBX3-BAC-EGFP high and low ESCs.
(D) Embryonic development from 2n embryo aggregation with TBX3^GFP-high and -low ESCs.
(E) Phase contrast (left) and GFP fluorescence (right) of TBX3-BAC-EGFP-low chimera. fl, front limb; hl, hind limb; m, midbrain.
The scale bars represent 50 μm (B and C) and 1 mm (E).

TBX3 Is Dynamically Expressed in Mouse ESCs
(A) TBX3^+/ven pre-implantation embryos at indicated time points.
(B) Wild-type pre-implantation embryos stained for NANOG (green), TBX3 (red), and PDGFRa (purple). Hoechst 33342 is shown (nuclei, blue).
(C) Representative phase contrast and GFP fluorescent images of de-novo-derived TBX3-BAC-EGFP mESCs cultured under indicated conditions. The scale bars represent 20 μm.
(D) TBX3-BAC-EGFP expression fluctuates under indicated culture conditions.
(E) qPCR for Tbx3 of TBX3-BAC-EGFP low, mid, and high sorted cells (according to D). n = 4; two individual clones from two independent experiments)
(F and G) Experimental sorting strategy and schematic results (F) of (G); n = 2, two independent clones.
(H and I) Sorting strategy and schematic results (H) of (I). Two out of four independent experiments with similar results are shown. For 2i condition, n = 2.
(J) Hierarchical clustering of gene expression profiles of TBX3^GFP-high and -low ESCs with published pluripotent and somatic cells (GSE11274 and GSE58735).
(K and L) Gene set enrichment analysis shows no bias of pluripotent marker genes between TBX3^GFP-high (high) and -low (low) ESCs (J) but a bias to epiblast (K) in TBX3^GFP-low ESCs.
(M) Hierarchical clustering shows that TBX3^GFP-low ESCs (TBX3 low) cluster close to EpiSCs and epiblast (E5.5/6.5), TBX3^GFP-high ESCs (TBX3 high) are close to published ESCs (J1, R1), 2i-treated ESCs (J1_2i), and E3.5 inner cell mass (ICM cells E3.5; GSE58735 and GSE35416).
See also and .

Characterization of TBX3-Null ESCs
(A) BrdU-FACS of TBX3^ven/ven and TBX3^+/+ ESC lines. Gate at y axis, apoptotic cells; lower left circle, G0/G1 population; lower right circle, G2/M population; upper big gate, S phase.
(B) Quantification of BrdU incorporation summarized from three TBX3^ven/ven and a TBX3^+/+ ESC line. n = 3 (three individual clones). DAPI-based cell-cycle analysis revealed no difference between TBX3^ven/ven and TBX3^+/+ ESCs (data not shown).
(C) Quantified AP+ colonies after removing individual LIF/2i components. N2B27 backbone medium. n = 3, for two independent clones with duplicate technical replicates.
(D) Co-culture strategy to compare self-renewal capacity of TBX3^+/+ and TBX3^ven/ven mESC under LIF/2i. Analysis at start (P0) and end of experiment (P3) for venus expression via FACS is shown. n = 3, for two independent clones with duplicate technical replicates.
(E) Quantification of cell populations from (D) after three passages.
(F) Phase images of TBX3^ven/ven and TBX3^+/+ mESC at 0 hr, 24 hr, 48 hr, and 72 hr after withdrawal of LIF and 2i. Delayed differentiation of TBX3^ven/ven as evident by preserved ESC morphology. The scale bars represent 20 μm.
(G–I) mRNA levels of pluripotency markers (Oct3/4, Sox2, and Nanog), early endodermal markers (Foxa2, Eomes, and Hnf4a), T (brachyury) mesodermal, and Pax6 as ectodermal marker in TBX3^+/+ and TBX3^ven/ven mESC at respective time points of LIF/2i withdrawal. Two clones per genotype are shown.
Two independent experiments performed with triplicate technical replicates are shown. See also .

Tbx3 Is Dispensable to Induce and Maintain Pluripotency
(A and B) Quantification (A) and representative AP+ colonies (B) day 12 after reprogramming of TBX3^+/+ and TBX3^−/− MEFs. Three independent experiments performed with triplicate technical replicates are shown.
(C) Morphology of OKS-infected TBX3^−/− MEFs (top) picked and expanded TBX3^−/− iPSCs (bottom).
(D) Tbx3 expression in TBX3^+/− and (TBX3^−/−) iPSCs. Two independent experiments performed with triplicate technical replicates are shown.
(E) TBX3 western blot in iPSCs (two different antibodies). Representative images of at least three independent experiments are shown.
(F) Pluripotency marker IHC in generated iPSCs of the indicated genotypes.
(G) Hierarchical clustering of gene expression profiles shows that iPSCs cluster with published ESC/ICM datasets.
(H) Crossing scheme to obtain TBX3^ven/ven-null mice and overview of derivation frequency; ND, not determined.
(I) Pluripotency marker IHC in de novo ESCs of the indicated genotypes.
(J and K) TBX3 mRNA (J; three independent experiments performed with triplicate technical replicates) and protein (K) expression confirms TBX3-null phenotype in TBX3^ven/ven ESCs (two TBX3 antibodies). Representative images of at least three independent experiments are shown.
(L) TBX3^−/− iPSCs form teratomas with all three germ layers (ecto-, meso-, and endoderm) in NMRI mice.
(M) Generation of “all-PSC”-derived TBX3-null embryos using 4n embryo aggregation from TBX3^−/− iPSCs. Developmental stage as indicated. To ensure visualization of TBX3-null progeny, TBX3^−/− iPSCs were constitutively labeled with a tdTomato-expressing lentivirus (LVtdTomato).
(N) Germline contribution of TBX3^−/− iPSCs carrying an OCT4-EGFP reporter and LVtdTomato after 2n aggregation assay in female (left) and male (right) gonads at E12.5.
(O) Representative fluorescent image of newborn chimeric mouse testis after 2n aggregation assay of OCT4-EGFP and LVtdTomato-infected TBX3^−/− iPSCs.
(P) TBX3^−/− iPSCs contribute to adult chimeric mice after 2n aggregation assay (light coat color).
(Q) Representative fluorescent image of adult chimeric mouse testis after 2n aggregation assay of OCT4-EGFP and LVtdTomato-infected TBX3^−/− iPSCs.
The scale bars represent 50 μm. See also and .

TBX3 Depletion Leads to Complex Compensational Processes
(A) Molecular circuitry of pluripotency markers (inner layer) and TBX3 target genes (outer layer) generated by the STRING database. The color of a node reflects the change in gene expression level between wild-type (WT) and TBX3-null (KO) ESCs. Black edge shows an evident interaction (confidence score > 0.7) between two connected genes from the STRING database. The gray edge connects TBX3 with its target genes identified by ChIP-seq data (GSE19219; ).
(B) Venn diagram of upregulated genes in TBX3-null ESCs and Tbx3 knockdown ESCs (GSE26520), compared to respective wild-type controls, shows 20 candidate genes.
(C) Heatmap of gene expression level of these candidate genes. Light green, below median expression; dark green, above median.
(D) The expression level of Dppa3 in TBX3^+/+and TBX3^−/− iPSCs, wild-type (+/+), and TBX3-null (ven/ven) ESCs.
(E) Clustering of single-cell Dppa3 expression (n = 26) suggests the reciprocal expression between Tbx3 and Dppa3 (hypergeometric test; p = 0.023). Dark green, group of low Dppa3 expression; light green, group of high Dppa3 expression. TBX3-low-sorted single cell is shown (GFP-L no. X). TBX3-high-sorted single cell is shown (GFP-H no. X).
(F) TBX3 occupancy (GSE19219) and H3K4me3/H3K27me3 bivalent modifications (GSE31039) of DPPA3 promoter region in ESCs.
(G) Dppa3 expression in WT or TBX3^−/− iPSCs upon transduction with scramble control virus or various combinations of different shRNAs targeting Dppa3. Two independent experiments performed with triplicate technical replicates are shown.
(H) Colony formation of TBX3^−/− iPSC lines with shRNAs against Dppa3 upon clonal density seeding. Top, representative AP-stained plates; bottom, quantification of three independent experiments performed with triplicate technical replicates. Black bar, AP positive; grey bar, partially AP positive; light gray bar, AP negative.
(I) qPCR of pluripotency markers in Dppa3 knockdown and scramble control TBX3^−/− iPSC lines (three independent experiments performed with duplicate technical replicates).
(J) Comparisons of differentially regulated genes between TBX3-high and -low ESC and between TBX3 WT (+/+) and TBX3-null (ven/ven) ESCs.
(K and L) The GO terms of DNA methylation (GO0006306) show no bias toward TBX3-high or -low ESCs (K) but were significantly enriched in TBX3-null ESCs (L).