WTAP goes through phase transition with IFN-β stimulation.

(A) THP-1-derived macrophages were infected with Vesicular Stomatitis Virus (VSV) (m.o.i. = 0.1) for 24 hours together with or without 5% 1,6-hexanediol (hex) and 10 μg/mL digitonin for 2 hours or left untreated (UT). Endogenous WTAP was stained and imaged using confocal microscopy. The number of WTAP condensates that diameter over 0.4 μm of n = 20 cells were counted through ImageJ and shown. Scale bars indicated 5 μm.

(B) HeLa cells were placed on the dishes and stimulated with or without 10 ng/mL IFN-β for 1 hour at 37 °C. Endogenous WTAP was stained and imaged using confocal microscopy. The number of WTAP condensates that diameter over 0.4 μm of n = 80 cells were counted through ImageJ and shown. Scale bars indicated 5 μm.

(C) Domain structures (top) and distribution of amino acids (bottom) of WTAP protein. WT, wild type. NTD, N-terminal domain. CTD, C-terminal domain. NLS, nuclear localization signal. IDR, intrinsically disordered regions. Gln, Glutamine. Ser, Serine. Lue, Leucine. Glu, Glutamic acid.

(D) Tubes containing physiological buffer with recombinant mCherry (10 μM) or mCherry-WTAP (10 μM) were compared, in which recombinant mCherry-WTAP underwent phase separation.

(E) Foci formation of recombinant mCherry-WTAP (10 μM) with or without 5% 1,6-hexanediol (hex) in vitro was observed through confocal microscopy. Scale bars indicated 5 μm.

(F) Phase separation of mCherry-WTAP in mCherry-WTAP-rescued HeLa cells treated with or without 5% hex and 20 μg/mL digitonin were observed through confocal microscopy. Representative images of n = 20 cells were shown. Scale bars indicated 5 μm.

(G) mCherry-WTAP-rescued HeLa cells were placed on dishes and treated with or without 10 ng/mL IFN-β for 1 hour at 37 °C. After stimulation, bleaching of the WTAP foci was performed and quantification of fluorescence recovery after photobleaching (FRAP) of mCherry-WTAP aggregates was analyzed. The start time point of recovery after photobleaching was defined as 0 second. Representative images of n = 10 cells were shown. Scale bars indicated 5 μm.

(H) Recombinant mCherry-WT WTAP, NTD and CTD (10 μM) were mixed with the physiological buffer and placed on the dishes at 37°C (Prebleach). After incubation, bleaching was performed and quantification of FRAP of recombinant mCherry-WT WTAP, NTD and CTD were analyzed. Representative images of n = 6 condensates were shown, and the normalized intensity was measured and analyzed. The start time point of recovery after photobleaching was defined as 0 second. Scale bars indicated 2 μm.

All error bars, mean values ± SD, P-values were determined by unpaired two-tailed Student’s t-test of n = 20 cells in (A) and n = 80 cells in (B). For (A-B and D-H), similar results were obtained for three independent biological experiments.

IFN-β-mediated dephosphorylation of WTAP induces its phase transition.

(A) THP-1-derived macrophages were treated with 10 ng/mL IFN-β for 1 hour or left untreated. Whole cell lysate (WCL) was collected and immunoprecipitation (IP) experiment using anti-WTAP antibody or rabbit IgG was performed, followed with immunoblot. pWTAP was detected by anti-phosphoserine/threonine/tyrosine antibody (pan-p). Relative protein level was shown.

(B) THP-1-derived macrophages were treated with 10 ng/mL IFN-β for 1 hour or left untreated. WCL was collected and IP experiment using anti-WTAP antibody or rabbit IgG was performed, followed with immunoblot.

(C) THP-1-derived macrophages were transfected with scramble (scr) or PPP4-targeted siRNA and stimulated with or without 10 ng/mL IFN-β for 1 hour. WCL was collected and IP experiment with anti-WTAP antibody or rabbit IgG was performed, followed with immunoblot. pWTAP was detected by anti-phosphoserine/threonine/tyrosine antibody (pan-p). Relative protein level was measured and analyzed.

(D) mCherry-WTAP was purified from HEK 293T cells expressing mCherry-WTAP using anti-WTAP antibody and analyzed by mass spectrometry (MS) for the phosphorylation. Six phosphorylated residues were identified. Data of MS assay for five phosphorylated sites within CTD were shown in Figure 2-figure supplement 1D. (E-F) Recombinant mCherry-WTAP mutants with indicated serine/threonine (S/T) to aspartic acid (D) or alanine (A) mutation were listed in (E), while representative images of the phase-separated mCherry-WT WTAP, 5ST-D and 5ST-A mutants were shown in (F). Images of phase separation of other mCherry-WTAP mutants (10 μM) were shown in Figure 2-figure supplement 1E. Scale bars indicated 5 μm.

(G) Recombinant mCherry-WT WTAP (10 μM), 5ST-D mutant (10 μM) or 5ST-A mutant (10 μM) was mixed with the physiological buffer and placed on the dishes at 37°C (Prebleach). After incubation, bleaching was performed and quantification of fluorescence recovery after photobleaching (FRAP) of recombinant mCherry-WT WTAP, 5ST-D or 5ST-A mutant were analyzed. Representative images of n = 5 condensates were shown and the normalized intensity was measured and analyzed. The start time point of recovery after photobleaching was defined as 0 second. Arrow indicated the FRAP area while scale bars indicated 2 μm.

(H) mCherry-WT WTAP, 5ST-D or 5ST-A mutants-rescued HeLa cells were placed on the dishes and stimulated with or without 10 ng/mL IFN-β for 1 hour at 37°C. After seeding, bleaching of the WTAP foci was performed and quantification of FRAP of mCherry-WTAP aggregates was analyzed. Representative images of n = 10 cells were shown, and the normalized intensity was measured and analyzed. The start time point of recovery after photobleaching was defined as 0 second. Scale bars indicated 5 μm.

(I) Schematic figure of phase separation of WT WTAP, 5ST-D and 5ST-A in vitro (above) and WT WTAP with IFN-β stimulation or left untreated (UT), 5ST-D and 5ST-A in cells (below).

All error bars, mean values ± SD, P-values were determined by unpaired two-tailed Student’s t-test of n = 3 independent biological experiments in (A and C). For (A-C and E-H), similar results were obtained for three independent biological experiments.

WTAP is crucial for the m6A modification and expression of ISG mRNAs.

(A) Transcriptome sequencing analysis of control (GFPsgRNA) and WTAPsgRNA #2 THP-1-derived macrophages stimulated with 10 ng/mL IFN-β for 0, 6, 12 hours. The count-per-million (CPM) value of the ISGs was drawn by Heatmapper and clustered using Centroid Linkage approach.

(B) Volcano plots showing the changes in transcripts level of IFN-β up-regulated genes in WTAPsgRNA THP-1-derived macrophages versus control (GFPsgRNA) cells under IFN-β stimulation for 6 and 12 hours. Red dots indicated the significantly up-regulated genes in WTAPsgRNA cells (log2(fold change) > 1 while adjusted P-value (padj) < 0.05).

(C) Gene ontology analysis for the WTAP down-regulated ISGs.

(D) TPMs showing the changes in transcripts level of IFIT1, IFIT2, OAS1 and OAS2 in control (GFPsgRNA) and WTAPsgRNA #2 THP-1-derived macrophages stimulated with 10 ng/mL IFN-β for 6 or 12 hours or left untreated.

(E-F) Control (GFPsgRNA) and WTAPsgRNA #2 THP-1-derived macrophages were treated with 10 ng/mL IFN-β for 4 hours, m6A modification analyzed by m6A methylated RNA immunoprecipitation (IP) followed by deep sequencing (MeRIP-seq). Distribution (E) and topology (F) analysis (5’-untranslated regions (UTR), coding sequences (CDS) and 3’-UTR) of total m6A sites in control (GFPsgRNA) cells or WTAP-dependent m6A sites.

(G) Percentage of m6A-modified ISGs including core ISGs or WTAP down-regulated ISGs were calculated and presented.

(H) Topology analysis (5’-UTR, CDS and 3’-UTR) of ISGs m6A sites and non-ISGs m6A sites.

(I) Control (GFPsgRNA) and WTAPsgRNA #2 THP-1-derived macrophages were treated with 10 ng/mL IFN-β for 4 hours. MeRIP-quantitative real time-polymerase chain reaction (qPCR) assay of IFIT1, IFIT2, OAS1 and OAS2 was performed and ratios between m6A-modified mRNA and input were shown (m6A IP/input).

(J) Control (GFPsgRNA), WTAPsgRNA #1 and WTAPsgRNA #2 THP-1-derived macrophages were treated with 10 ng/mL IFN-β for 2 hours. After IFN-β treatment, medium with stimuli was replaced by 5 μM actinomycin D (Act D) for indicated time points. RNA was collected and detected by qPCR assay.

(K) Control (GFPsgRNA), WTAPsgRNA #1 and WTAPsgRNA #2 THP-1-derived macrophages were treated with 10 ng/mL IFN-β for indicated time points. Expression of IFIT1, IFIT2, OAS1 and OAS2 mRNA was analyzed through qPCR assays.

All error bars, mean values ± SEM, P-values were determined by unpaired two-tailed Student’s t-test of n = 3 independent biological experiments in (I). All error bars, mean values ± SEM, P-values were determined by two-way ANOVA test of n = 3 independent biological experiments in (J-K).

Liquid-phase separation of WTAP mediates m6A modification and ISGs expression.

(A) Control (GFPsgRNA) and WTAPsgRNA #2 THP-1-derived macrophages were treated with 10 ng/mL IFN-β together with or without 5% 1,6-hexanediol (hex) and 20 μg/mL digitonin for 4 hours or left untreated. MeRIP-qPCR analysis of IFIT1, IFIT2, OAS1 and OAS2 was performed and ratios between m6A-modified mRNA and input were shown (m6A IP/input).

(B) THP-1-derived macrophages transfected with scramble (scr) siRNA or PPP4 siRNA were treated with 10 ng/mL IFN-β for 4 hours. MeRIP-qPCR assay of IFIT1, IFIT2, OAS1 and OAS2 was performed and ratios between m6A-modified mRNA and input were shown (m6A IP/input).

(C) THP-1-derived macrophages were pre-treated with 2 nM or 5 nM fostriecin for 24 hours or left untreated, followed with 10 ng/mL IFN-β for 4 hours. MeRIP-qPCR analysis of IFIT1, IFIT2, OAS1 and OAS2 was performed and ratios between m6A-modified mRNA and input were shown (m6A IP/input).

(D) Control (GFPsgRNA) and WTAPsgRNA #2 THP-1-derived macrophages with or without expression of Flag-tagged wild type (WT) WTAP, and its 5ST-D or 5ST-A mutants were treated with 10 ng/mL IFN-β for 4 hours. MeRIP-qPCR assay of IFIT1, IFIT2, OAS1 and OAS2 was performed and ratios between m6A-modified mRNA and input were shown (m6A IP/input).

(E) Wild type (WT) WTAP, 5ST-D or 5ST-A mutant-rescued WTAPsgRNA THP-1-derived macrophages were treated with 10 ng/mL IFN-β for 4 hours. MeRIP-qPCR assay of IFIT1, IFIT2, OAS1 and OAS2 was performed and ratios between m6A-modified mRNA and input were shown (m6A IP/input).

All error bars, mean values ± SEM, P-values were determined by two-way ANOVA test of n = 3 independent biological experiments in (A, E). All error bars, mean values ± SEM, P-values were determined by unpaired two-tailed Student’s t-test of n = 3 independent biological experiments in (B-C). All error bars, mean values ± SEM, P-values were determined by one-way ANOVA test of n = 3 independent biological experiments in (D).

Liquid-phase separated WTAP bridges STAT1 and m6A modification methyltransferase complex to direct m6A modification on ISG mRNAs.

(A) Occupancy of the promoter of the WTAP down-regulated ISGs with lower m6A level in WTAPsgRNA #2 cells (genes were identified in Figure 3A and both −2000-0 bp and −500-0 bp upstream of transcription start site were analyzed) by STAT1, STAT2 and IRF9 was predicted by AnimalTFDB 3.0.

(B) THP-1-derived macrophages were treated with 10 ng/mL IFN-β for indicated time points. Whole cell lysate (WCL) was collected and immunoprecipitation (IP) experiment using anti-STAT1 antibody or rabbit IgG was performed, followed with immunoblot.

(C) Control (GFPsgRNA) and WTAPsgRNA #2 THP-1-derived macrophages were treated with 10 ng/mL IFN-β for 1 hour or left untreated. WCL was collected and IP experiment using anti-METTL3 antibody or rabbit IgG was performed, followed with immunoblot.

(D) Control (GFPsgRNA) and WTAPsgRNA #2 THP-1-derived macrophages were treated with 10 ng/mL IFN-β for 2 hour or left untreated. RNA immunoprecipitation (RIP) experiments using anti-METTL3 antibody or rabbit IgG control were performed in control and WTAPsgRNA #2 THP-1-derived macrophages treated with 10 ng/mL IFN-β for 2 hours. RNA of ISGs enriched by RIP was analyzed by quantitative real time-polymerase chain reaction (qPCR) assay, and the ratios between RIP-enriched RNA and input were shown (RIP enrichment/input).

(E) THP-1-derived macrophages were treated with 10 ng/mL IFN-β for indicated time points together with or without 5% 1,6-hexanediol (hex) and 20 μg/mL digitonin. WCL was collected and IP experiment with anti-STAT1 antibody or rabbit IgG was performed, followed with immunoblot.

(F) Recombinant GFP-STAT1 (10 μM) and mCherry-WTAP (10 μM) were incubated with or without 5% hex in vitro. IP experiment with anti-STAT1 antibody or rabbit IgG was performed, followed with immunoblot.

(G) THP-1-derived macrophages were treated with 10 ng/mL IFN-β only or with 5% hex and 20 μg/mL digitonin for 1 hour or left untreated (UT). Interaction between STAT1 (green) and WTAP (red) were imaged using confocal microscope. Pearson’s correlation coefficient was analyzed and calculated from n = 80 cells through ImageJ. Scale bars indicated 5 μm.

(H-I) Recombinant GFP-STAT1 (10 μM), mCherry-WTAP (10 μM) and CFP-METTL3 (10 μM) were incubated using physiological buffer at 37°C in vitro. After incubation, images were captured using confocal microscope (H). Relative fluorescence intensity of proteins in n = 10 condensates were analyzed by ImageJ software (I).

All error bars, mean values ± SEM, P-values were determined by unpaired two-tailed Student’s t-test of n = 3 independent biological experiments in (D). All error bars, mean values ± SD, P-values were determined by unpaired two-tailed Student’s t-test of n = 80 cells in (G) and n = 3 independent biological experiments in (I). For (B-D, E-G and H), similar results were obtained for three independent biological experiments.

Phase transition of WTAP mediates its interaction with ISGs promoter regions to regulate m6A modification on ISG mRNAs.

(A) Chromatin immunoprecipitation (ChIP) experiments using anti-STAT1 antibody or rabbit IgG control were performed in THP-1-derived macrophages treated with 10 ng/mL IFN-β for 2 hours or left untreated. Binding between the promoter regions of IFIT1, IFIT2, OAS1 and OAS2 with WTAP was detected by quantitative real time-polymerase chain reaction (qPCR) assay. Ratios between ChIP-enriched DNA and input were shown (ChIP enrichment/input).

(B) ChIP experiments using anti-WTAP antibody or rabbit IgG control were performed in THP-1-derived macrophages treated with 10 ng/mL IFN-β for 2 hours or left untreated. Binding between the promoters of IFIT1, IFIT2, OAS1 and OAS2 with WTAP were detected by qPCR assay. Ratios between ChIP-enriched DNA and input were shown (ChIP enrichment/input).

(C) ChIP experiments using anti-WTAP antibody or rabbit IgG control were performed in THP-1-derived macrophages transfected with scramble (scr) siRNA, STAT1 siRNA or PPP4 siRNA treated with 10 ng/mL IFN-β for 2 hours. Binding between the promoter regions of IFIT1, IFIT2, OAS1 and OAS2 with WTAP was detected by qPCR assay. Ratios between ChIP-enriched DNA and input were shown (ChIP enrichment/input).

(D) ChIP experiments using anti-WTAP antibody or rabbit IgG control were performed in control (GFPsgRNA) and WTAPsgRNA #2 THP-1-derived macrophages with or without expression of Flag-tagged wild type (WT) WTAP, and its 5ST-D or 5ST-A mutants treated with 10 ng/mL IFN-β for 2 hours. Binding between the promoter regions of IFIT1, IFIT2, OAS1 and OAS2 with WTAP was detected by qPCR assay. Ratios between ChIP-enriched DNA and input were shown (ChIP enrichment/input).

(E) ChIP experiments using anti-WTAP antibody or rabbit IgG control were performed were performed in WT WTAP, 5ST-D or 5ST-A mutant-rescued WTAPsgRNA THP-1-derived macrophages treated with 10 ng/mL IFN-β along with 5% 1,6-hexanediol (hex) for 2 hours. Binding between the promoters of IFIT1, IFIT2, OAS1 and OAS2 with WTAP were detected by qPCR assay. Ratios between ChIP-enriched DNA and input were shown (ChIP enrichment/input).

(F) Schematic figure of IFN-β-induced phase transition of WTAP regulates the m6A modification of ISG mRNAs.

All error bars, mean values ± SEM, P-values were determined by unpaired two-tailed Student’s t-test of n = 3 independent biological experiments in (A-C). All error bars, mean values ± SEM, P-values were determined by one-way ANOVA test of n = 3 independent biological experiments in (D). All error bars, mean values ± SEM, P-values were determined by two-way ANOVA test of n = 3 independent biological experiments in (E).