Fig 2: Nuclear TRIM59 promotes macroH2A1 ubiquitination and degradation. a, b WB and qRT-PCR assays of effects of TRIM59 knockdown on macroH2A1 protein (a) and mRNA (b) levels in LN229/EGFR and U87/EGFR cells stimulated with EGF. c, e Effects of TRIM59 knockdown on macroH2A1 (mH2A1) stability in LN229/EGFR (c) and U87/EGFR (e) cells. After stimulation with EGF (100 ng/ml) for 6 h, cells were treated with cycloheximide (CHX, 20 mg/ml) for the indicated time points. d, f Quantification of mH2A1 protein levels in c and e, respectively. g Effects of TRIM59 WT, S308A, or S308D on mH2A1.1 ubiquitination. His-tagged ubiquitin (His-Ub) was co-transfected into LN229/EGFR/shTRIM59 and U87/EGFR/shTRIM59 cells with TRIM59 constructs or empty vector control. h Effects of Roscovitine on mH2A1.1 ubiquitination. His-Ub was co-transfected into LN229/EGFR/shTRIM59 and U87/EGFR/shTRIM59 cells with TRIM59 constructs or empty vector control. i HA-TRIM59 proteins purified from HEK293 cells were incubated with ATP, E1, and E2 along with Flag-mH2A1.1 proteins isolated from HEK293 cells for in vitro ubiquitination assay. Data are representative of three independent experiments with similar results. Data were expressed as means ± SD. **P < 0.01 by two-tailed Student’s t test. Source data are provided as a Source Data file

Fig 3: QKI5 regulates macroH2A1 alternative splicing in GC cellsA. QKI5 levels were detected in HGC-27 cells after treatment with pcDNA-QKI5 or pcDNA empty vector, and in SGC-7901 cells after treatment with si-QKI5s or siRNA controls by immunoblotting; B. Relative expression of macroH2A1.1, macroH2A1.2, and total macroH2A1 in QKI overexpressing and knockdown cells as described in A; C. Immunoblotting of the protein levels of macroH2A1.1, macroH2A1.2, and total macroH2A1 in QKI overexpressing and knockdown cells as described in A; D. Expression level of macroH2A1.1 mRNA was detected in 100 pairs of GC tissues and adjacent normal regions by RT-qPCR. E. The normalized relative expression of QKI5, mH2A1.1 and mH2A1.2 from the 100 pairs of gastric cancer samples were plotted in a radar graph on a logarithmic scale. F. Statistical analysis of the survival probability of macroH2A1.1 high- (n = 10) and low-expressing (n = 50) GC patients using Kaplan–Meier analysis. Statistical analysis is described in the Materials and Methods.

Fig 4: TRIM59-mediated macroH2A1 degradation enhances STAT3 signaling. a Venn diagram shows the overlap of mH2A1-bound sites in the genome in LN229/EGFRvIII cells treated with vehicle (0.1% DMSO), erlotinib (10 µM), or Roscovitine (20 µM) for 16 h using ChIP-Seq. b Heatmap of mH2A1 ChIP-Seq. c Gene ontology analysis of effects of both the treatment of erlotinib and Roscovitine on mH2A1 binding in distance from center of peak (±2 kb) at top five EGFR-driven pathways. d mH2A1 ChIP-Seq peaks at STAT3-targeted genes using IGV (Integrative Genomics Viewer). e ChIP-qPCR of effects of erlotinib or Roscovitine on mH2A1 binding at STAT3-targeted genes. f–h Effects of mH2A1 knockdown in LN229/EGFR and U87/EGFR cells with EGF stimulation. WB for expression of TRIM59, STAT3, and p-STAT3 (f); ChIP-qPCR analysis for STAT3 binding with PIM1 promoter (g); qRT-PCR analysis for STAT3-targeted gene PIM1 transcription (h). i–k Effects of reconstituted TRIM59 WT, S308A, ?R, or Y218F with or without EGF stimulation. WB for expression of TRIM59, mH2A1, and p-STAT3 (i); ChIP-qPCR analysis for STAT3 binding with PIM1 promoter (j); qRT-PCR analysis for PIM1 transcription (k). Data are representative of three independent experiments with similar results. Data were expressed as means ± SD. **P < 0.01, ***P < 0.001, by two-tailed Student’s t test. Source data are provided as a Source Data file

Fig 5: a Left representative pictures from hematoxylin and eosin staining of visceral white adipose tissue sections in wild-type (WT) and macroH2A1.2 Tg (Tg) mice fed with a HF diet. Right quantification of adipocyte area (µm2). b Leptin blood levels, measured by ELISA (Milliplex) in the four mice groups upon killing. Data are expressed as mean ± S.E.M. (n = 8–9 per group). *p < 0.05, ***p < 0.001 change versus WT fed a chow diet; # p < 0.01 change versus WT fed a HF diet. c Adipose gene expression profiling expressed in fold change and relative to HF diet compared to chow diet either in WT or in Tg mice. Results are represented as means of 3 mice per condition (WT, WT + HF diet, Tg, Tg + HF diet). All differences, with the exception of BMP7 and WNT5B (not significant), are highly significant (p < 0.0001)