Fig 1: FGF4 is essential for upregulating Xist expression and initiating random XCI.(A) The protein level of pFGFR in WT ES cells after addition of 0, 25, 75, 125, 200, and 250 nM pan-FGFR inhibitor (BGJ398) on day 5 of differentiation. (B, C) Immunofluorescence staining (B) and quantification (C) of H3K27me3 domains in WT, Fgf4 KO-84#, Fgf4 KO-84# + FGF4 ES cells on day 5 of differentiation during BGJ398 treatment. n = 5–6; scale bars: 10 μm; FGF4: 10 ng/mL. The white arrow denotes a nuclear H3K27me3 domain. % H3K27me3 = number of cells with nuclear H3K27me3 domains/total number of analyzed cells. Cells with one H3K27me3 domain were considered as normal XCI and included in the statistical unit. (D) The presentation of the spatial pattern of gene expression of Fgf4 and Fgfr1 in gastrulating mouse embryos. The upper panel: the schematic diagram of E5.5, E6.0, and E6.5 embryos. The boxed region indicates the epiblast, and the schematic model (right) depicts the sample section and captured regions. The position of each sectional sample is labeled by the number on the distal-proximal axis. The captured regions are labeled in the section, and the corresponding abbreviations in each section are indicated below. The lower panel: spatial expression pattern of Fgf4 and Fgfr1 in the epiblast at the indicated stage, which is presented as the corn plot (so named for its resemblance to a corn cob). Each dot in the plot represents the cell sample at the specific positional address indicated by the number next to the plot and the abbreviation at the top. The color indicates the level of gene expression computed from the transcript counts in the 3D gene expression database of gastrulating mouse embryos. Data source: eGastrulation (http://egastrulation.sibcb.ac.cn). (E) Immunofluorescence staining of FGF4 and FGFR1 in E6.5 epiblasts; scale bars: 75 μm. OCT3/4 was used as the marker specific to the epiblast. (F) Dynamic expression levels of Xist in WT, Fgf4 KO-84#, and Fgf4 KO-84# + FGF4 ES cells on day 0, 1, 3, and 5 of differentiation. Data were represented as means ± SEM; n ≥ 6. The schematic diagram below the X-axis indicates the timeline of differentiation. The day when ES cells were transferred to the differentiation medium was designated as day 0. (G, H) FISH staining (G) and quantification (H) of Xist clouds in WT, Fgf4 KO-84#, Fgf4 KO-84# + FGF4 ES cells on day 3, 5, 7, and 9 of differentiation; n = 4; scale bars: 10 μm. The white arrow denotes a nuclear Xist cloud. % Xist-cloud = number of cells with Xist cloud/total number of analyzed cells. Cells with one Xist cloud signal were considered normal XCI and included in the statistical unit. (I, J) Immunofluorescence staining (I) and quantification (J) of H3K27me3 domain in WT, Fgf4 KO-84#, Fgf4 KO-84# + FGF4 ES cells on day 3, 5, 7, and 9 of differentiation; n = 3–4; scale bars: 10 μm. The white arrow denotes a nuclear H3K27me3 domain. % H3K27me3 = number of cells with nuclear H3K27me3 domains/total number of analyzed cells. Cells with one H3K27me3 domain were considered as normal XCI and included in the statistical unit. (K, L) Immunofluorescence staining (K) and quantification (L) of H3K27me3 domains in Fgf4 KO-84# after FGF4 addition at different time windows; n = 4–6. Control, cells without FGF4 supplement during the process of differentiation; 0–24, 0–48, 0–72, 0–96, and 0–120 h, cells are treated with FGF4 for 24, 48, 72, 96, and 120 h, respectively, upon differentiation, then transferred to the differentiation medium without FGF4 addition until the H3K27me3 staining on day 5 of differentiation. (C, L) Data are shown as means ± SEM. P value was calculated by one-way ANOVA test with multiple comparisons. For more details, please see the “Methods”. Source data are available online for this figure.
Fig 2: FGF4 regulates Xist upregulation via the ERK–YY1 axis.(A) Phos-tag analysis of pYY1 and Western blot analysis of β-TUBULIN in WT ES cells on day 0–0 h, day 0–2 h, day 1, day 3, and day 5 of differentiation (upper panel), along with the quantitative analysis of pYY1 phosphorylation levels (down panel), n = 4. (B) Phos-tag analysis of pYY1 and Western blot analysis of pERK, ERK in WT, Fgf4 KO-84#, Fgf4 KO-84# + FGF4 ES cells on day 0–2 h of differentiation; FGF4 10 ng/mL. (C, D) Co-IP assays (C) and quantification (D) of interaction between YY1 and pERK, in differentiating WT, Fgf4 KO-84#, and Fgf4 KO-84# + FGF4 ES cells on day 1 of differentiation; n = 3. (E) Western blot analysis of pYY1, pERK, ERK in WT treated with or without MEK/ERKi on day 0–2 h of differentiation. MEK/ERKi, PD0325901, 1 μM. (F) ChIP analysis of YY1 enrichment at the binding site within the Xist’s 5’ region in WT, Fgf4 KO-84#, and WT ES cells treated with MEK/ERKi on day 1 of differentiation; n = 3. (G) ChIP analysis of YY1 enrichment at the binding site within the Xist’s 5’ region in Fgf4 KO-84# after FGF4 addition on day 1 of differentiation; n = 3. (H) Average profile (metaplot) of normalized YY1 CUT&Tag signal centered on transcription start sites in WT and Fgf4 KO-84# ES cells. Signals were plotted across a ± 3 kb window around the TSS. The x-axis indicates distance from the TSS (kb), and the y-axis indicates normalized YY1 enrichment. (I) Heatmaps of YY1 binding enrichment and corresponding average CUT&Tag signal profiles in WT and Fgf4 KO-84# ES cells. CUT&Tag signals are plotted around TSS ( ± 3 kb) associated with YY1 peaks. YY1 peaks are categorized as gained (n = 187), lost (n = 191), or shared (n = 1809) in Fgf4 KO-84# compared with WT. (J) The UCSC browser view showing YY1 enrichment at representative promoter regions in WT, Fgf4 KO-84# ES cells. Gene models are shown below; arrows indicate the direction of transcription. (K) Schematic diagram of DNA mutations via the substitution of serine (Ser, S) 120 and 247 with alanine (Ala, A, non-phosphorylation-mimic mutation) or aspartic acid (Asp, D, phosphorylation-mimic mutation) with the coding region of Yy1 gene. For the Ser→Ala substitution at the positions 120 or 247, the triplet codons TCG and TCA were respectively replaced by CGG and CGA; For the Ser→Asp substitution at the position 120 or 247, the triplet codon TCG and TCA were replaced by CTG and TCG, respectively. (L) Relative luciferase activity of the reporters containing Xist promoter, which were co-transfected with an empty vector (pCAGGS), or overexpression vector of WT or mutant Yy1 in WT ES cells; n = 6–9. Each dot indicates the result of each reaction, and assays were replicated independently at least three times. (M) Relative expression levels of endogenous Xist in YY1-dTAG ES cells transfected with an empty vector (pCAGGS), or an overexpression vector of WT or mutant Yy1; n = 4. (D, F, G, L, M) Data are shown as means ± SEM. P value was calculated by one-way ANOVA test with multiple comparisons. For more details, please see the “Methods”. Source data are available online for this figure.
Fig 3: FGF4 is critical for the timely decline of pluripotency factors that are major repressors of Xist upregulation.(A–C) Dynamic expression levels of Nanog (D), Prdm14 (E), Rex1 (F) in WT, WT+FGFRi, Fgf4 KO-84#, Fgf4 KO-84# + FGF4 ES cells at days 0, 1, 3, and 5 of differentiation; n = 3. The schematic diagram below the X-axis in D indicates the timeline of differentiation. The day when ES cells were transferred to the differentiation medium was designated as day 0. (D–F)Xist FISH staining and immunofluorescence staining of H3K27me3, and corresponding quantification of Xist clouds (B) and H3K27me3 domains (C) in Fgf4 KO-84 ES cells treated with retinoic acid (RA), FGF4, and pan-FGFR inhibitor, either alone or in combination; RA, 100 nM; FGF4, 10 ng/mL; FGFRi: pan-FGFR inhibitor (BGJ398), 75 nM; n = 4–7 for (B); n = 3–4 for (C); scale bars: 10 μm. In the upper panels, the white arrow in the upper panels denotes a nuclear Xist cloud. % Xist-cloud = number of cells with Xist cloud/total number of analyzed cells. Cells with one Xist cloud signal were considered as normal XCI and included in the statistical unit. In the lower panels, the white in the nuclear H3K27me3 domain. % H3K27me3 = number of cells with nuclear H3K27me3 domains/total number of analyzed cells. Cells with one H3K27me3 domain were considered normal XCI and included in the statistical unit. (G, H) Immunofluorescence staining (G) and quantification (H) of H3K27me3 domains in WT, Fgf4 KO, Fgf4-Nanog DKO, Fgf4-Prdm14 DKO, Fgf4-Rex1 DKO ES cells on day 5 of differentiation; scale bars: 10 μm. (I) A model illustrating the mechanism underlying FGF4-induced Xist upregulation and XCI initiation. The white arrow denotes a nuclear H3K27me3 domain. % H3K27me3 = number of cells with nuclear H3K27me3 domains/total number of analyzed cells. Cells with one H3K27me3 domain were considered as normal XCI and included in the statistical unit. (B, C) Data are shown as means ± SEM. P value was calculated by one-way ANOVA test with multiple comparisons. For more details, please see the “Methods”. Source data are available online for this figure.
Fig 4: FGF4 initiates random XCI in a MEK/ERK-dependent manner.(A) Schematic of inhibitor treatment during differentiation from D0 to D5 (upper panel), and Immunofluorescence staining of H3K27me3 and Xist FISH staining (down panel) in WT ES cells treated with the inhibitor specific to PLCγ, STAT, PI3K, and MEK/ERK on day 5 of differentiation. (B, C) Immunofluorescence staining of H3K27me3 and Xist FISH staining, and corresponding quantification of H3K27me3 domains (B) and Xist clouds (C) in WT ES cells treated with the inhibitor specific to PLCγ, STAT, PI3K, and MEK/ERK on day 5 of differentiation; n = 3–5 for B; n = 4–5 for (C); scale bars: 10 μm. PLCγi: U73122 (U), 5 μM; STATi: Fludarabine (Flu), 1 μM plus SH-4-54 (SH), 1 μM; PI3Ki: LY294002 (LY) 1 μM; MEK/ERKi: PD0325901 (PD), 1 μM. In the upper panels, the white arrow in the upper panels denotes a nuclear Xist cloud. % Xist-cloud = number of cells with Xist cloud/total number of analyzed cells. Cells with one Xist cloud signal were considered as normal XCI and included in the statistical unit. In the lower panels, the white in the nuclear H3K27me3 domain. % H3K27me3 = number of cells with nuclear H3K27me3 domains/ total number of analyzed cells. Cells with one H3K27me3 domain were considered as normal XCI and included in the statistical unit. (D) Western blot analysis of ERK and pERK in WT ES cells treated with 0.01, 0.1, and 1 μM MEK/ERK inhibitor (MEK/ERKi) on day 5 of differentiation. (E) Quantification of H3K27me3 domains in WT ES cells treated with 0.01, 0.1, and 1 μM MEK inhibitor on day 5 of differentiation; n = 3–4. A two-tailed unpaired Student’s t test was used to assess the differences between the control and experimental groups. (F) Quantification of H3K27me3 domains in WT and Fgf4 KO-84# ES cells treated with or without 5, 20, 100, and 500 nM PMA on day 5 of differentiation; n = 4–5. (G) Quantification of H3K27me3 domains in WT ES cells subject to FGFR inhibition or in combination with PMA-induced ERK activation on day 5 of differentiation; n = 4–5. FGFRi: BGJ398, 75 nM; PMA: 100 nM. (H) Schematic diagram of the experimental workflow. Pregnant mice received daily intraperitoneal injections of the MEK/ERK inhibitor, FGFR inhibitor, or in combination with the ERK-activating agent PMA, during the window of random XCI. The female epiblast was collected and tested for Xist expression and iXCI status. (I) Relative expression levels of Xist in in vivo E6.5 epiblast from females undergoing intraperitoneal injections with DMSO, MEK/ERKi; n = 7–10. MEK/ERKi: PD0325901, 10 mg/kg. (J) Immunofluorescence staining (left) and quantification (right) of H3K27me3 domains in in vivo E6.5 epiblasts from females undergoing intraperitoneal injection with DMSO, MEK/ERKi; n = 4–6; scale bars: 10 μm. The white arrow denotes a nuclear H3K27me3 domain. % H3K27me3 = number of cells with nuclear H3K27me3 domains/total number of analyzed cells. Cells with one H3K27me3 domain were considered as normal XCI and included in the statistical unit. (K) Relative expression levels of Xist in in vivo E6.5 epiblast from females undergoing intraperitoneal injection with DMSO, FGFRi, and FGFRi+PMA, respectively; n = 7–17. FGFRi: BGJ398, 10 mg/kg; PMA, 2.5 mg/kg. (L) Immunofluorescence staining (left) and quantification (right) of H3K27me3 domains in in vivo E6.5 epiblasts from females undergoing intraperitoneal injection of DMSO, MEK/ERKi, FGFRi and FGFRi+PMA, respectively; n = 4–8; scale bars: 10 μm. The white arrow denotes a nuclear H3K27me3 domain. % H3K27me3 = number of cells with nuclear H3K27me3 domains/total number of analyzed cells. Cells with one H3K27me3 domain were considered as normal XCI and included in the statistical unit. (B, C, E–G, I–L). Data are shown as means ± SEM. P value was calculated by one-way ANOVA test with multiple comparisons (B, C, E–G, K, L) and unpaired two-tailed Student’s t test (I, J). For more details, please see the “Methods”. Source data are available online for this figure.
Supplier Page from R&D Systems, a Bio-Techne Brand for Recombinant Mouse FGF-4 Protein