CREBBP/EP300 mutations promoted tumor progression in diffuse large B-cell lymphoma through altering tumor-associated macrophage polarization via FBXW7-NOTCH-CCL2/CSF1 axis

Abstract

Epigenetic alterations play an important role in tumor progression of diffuse large B-cell lymphoma (DLBCL). However, the biological relevance of epigenetic gene mutations on tumor microenvironment remains to be determined. The core set of genes relating to histone methylation (KMT2D, KMT2C, EZH2), histone acetylation (CREBBP, EP300), DNA methylation (TET2), and chromatin remodeling (ARID1A) were detected in the training cohort of 316 patients by whole-genome/exome sequencing (WGS/WES) and in the validation cohort of 303 patients with newly diagnosed DLBCL by targeted sequencing. Their correlation with peripheral blood immune cells and clinical outcomes were assessed. Underlying mechanisms on tumor microenvironment were investigated both in vitro and in vivo. Among all 619 DLBCL patients, somatic mutations in KMT2D (19.5%) were most frequently observed, followed by mutations in ARID1A (8.7%), CREBBP (8.4%), KMT2C (8.2%), TET2 (7.8%), EP300 (6.8%), and EZH2 (2.9%). Among them, CREBBP/EP300 mutations were significantly associated with decreased peripheral blood absolute lymphocyte-to-monocyte ratios, as well as inferior progression-free and overall survival. In B-lymphoma cells, the mutation or knockdown of CREBBP or EP300 inhibited H3K27 acetylation, downregulated FBXW7 expression, activated the NOTCH pathway, and downstream CCL2/CSF1 expression, resulting in tumor-associated macrophage polarization to M2 phenotype and tumor cell proliferation. In B-lymphoma murine models, xenografted tumors bearing CREBBP/EP300 mutation presented lower H3K27 acetylation, higher M2 macrophage recruitment, and more rapid tumor growth than those with CREBBP/EP300 wild-type control via FBXW7-NOTCH-CCL2/CSF1 axis. Our work thus contributed to the understanding of aberrant histone acetylation regulation on tumor microenvironment as an alternative mechanism of tumor progression in DLBCL.

Fig. 1

Chromatin-modifying genes were frequently mutated in diffuse large B-cell lymphoma (DLBCL). a Gene mutations identified by whole-genome/exome sequencing (WGS/WES) in the training cohort of 316 patients (upper panel) and by targeted sequencing in the validation of 303 patients (lower panel) with DLBCL. b Type and a number of non-silent somatic mutations (upper panel), and the number of non-silent somatic single-nucleotide variants (SNVs) identified by WGS/WES/targeted sequencing (lower panel) in the training and validation cohorts of 619 DLBCL patients. c Circos diagram of the correlation between genes, representing the combinations of mutations in different genes

Fig. 2

CREBBP/EP300 mutations were related to tumor progression and the aberrant tumor microenvironment in diffuse large B-cell lymphoma (DLBCL). a, b Progression-free survival (PFS) and overall survival (OS) curves of DLBCL patients in the training cohort (a) and in the validation cohort (b). c, d The distribution of absolute lymphocyte-to-monocyte ratios (ALC/AMC) and T-cell subsets in terms of CREBBP/EP300 mutation status in the training cohort (c) and in the validation cohort (d). Data are presented as the mean ± SD

Fig. 3

CREBBP/EP300 mutations inhibited H3K27 acetylation and activated the NOTCH signaling pathway. a Structure prediction of the complex of CREBBP and EP300 mutations. b Protein expression of CREBBP and H3K27ac detected in vector, CREBBP^wt, CREBBP^R1392*, scramble, CREBBP^kd, protein expression of EP300 and H3K27ac in vector, EP300^wt, EP300^H1377R, scramble, EP300^kd of DB, SUDHL4, and LY10 cells by western blot. Tubulin and H3 were used as loading controls. The CREBBP/tubulin, EP300/tubulin, and H3K27ac/H3 ratio are shown. c Immunofluorescence assay of H3K27ac in CREBBP^R1392* and EP300^H1377R DB cells. d Immunohistochemical study of H3K27ac in tumor samples of diffuse large B-cell lymphoma (DLBCL) patients with and without CREBBP/EP300 mutations. e Pathway enrichment analysis in DLBCL patients with or without CREBBP/EP300 mutations according to the Kyoto Encyclopedia of Genes and Genomes (KEGG) and Reactome databases. f Gene Set Enrichment Analysis (GSEA) enriched differentially expressed genes in NOTCH signaling pathway with or without CREBBP/EP300 mutations. Enrichment scores were listed with P value. NES normalized enrichment score, FDR false discovery rate. g Immunohistochemical study of NICD (intracellular portions of NOTCH1) in tumor samples of DLBCL patients with or without CREBBP/EP300 mutations

Fig. 4

CREBBP/EP300 mutations inhibited FBXW7 and activated the NOTCH signaling pathway. a Heatmap of genes associated with the NOTCH signaling pathway in CREBBP^mut/EP300^mut patients, as compared to CREBBP^wt/EP300^wt patients. b Normalized mRNA expression of FBXW7, HEY1 and HEY2 in tumor samples of diffuse large B-cell lymphoma (DLBCL) patients with or without CREBBP/EP300 mutations as revealed by RNA sequencing data. c Relative gene expression of FBXW7, HEY1, and HEY2 in CREBBP^R1392*, CREBBP^kd, EP300^H1377R, and EP300^kd DB cells, as compared to CREBBP^wt, EP300^wt or scramble DB cells by quantitative real-time PCR (RT-PCR). Data are presented as the mean ± SD (N = 3). d Protein expression of FBXW7 and NICD detected in vector, CREBBP^wt, CREBBP^R1392*, scramble, CREBBP^kd DB, SUDHL4, and LY10 cells, and in vector, EP300^wt, EP300^H1377R, scramble, EP300^kd DB, SUDHL4 and LY10 cells by western blot. The same lysates were used as in Fig. 3b. Tubulin was used as a loading control. The FBXW7/Tubulin and NICD/tubulin ratio were shown. e Occupancies of H3K27ac in the proximal promoter areas of FBXW7 in CREBBP^wt, CREBBP^R1392*, scramble, CREBBP^kd DB cells, and in EP300^wt, EP300^H1377R, scramble, EP300^kd DB cells by chromatin immunoprecipitation (ChIP) assay. Data are presented as the mean ± SD (N = 3). f Expression of NICD in CREBBP^R1392* and CREBBP^kd DB cells, as well as EP300^H1377R and EP300^kd DB cells with or without reintroduction of FBXW7 protein by western blot

Fig. 5

CREBBP/EP300 mutations contributed to macrophage activation and polarization. a Distribution of immune subpopulations in diffuse large B-cell lymphoma (DLBCL) patients with or without CREBBP/EP300 mutations analyzed by computational deconvolution of transcriptomics data using Cibersort. b Prediction of immune subpopulations normalized enrichment scores for DLBCL patients with CREBBP/EP300 mutations using GSEA. Immune cells with P < 0.05 are listed. c GSEA showing association of M2 macrophage gene signatures with CREBBP/EP300 mutations in DLBCL patients. NES normalized enrichment score, FDR false discovery rate. d Confocal analysis of CD68 and CD163 in DLBCL patients with or without CREBBP/EP300 mutations. The cells were counted from five randomly selected visions and subjected for statistical analysis. Data are presented as the mean ± SD. e Chemiluminescence immunoassay of IL-10 and IL-1β in the training and validation cohorts of DLBCL patients according to CREBBP/EP300 mutations. Data are presented as the mean ± SD. f Normalized mRNA expression of immune inhibitors (B7-H4 and CSFR1) in DLBCL patients with or without CREBBP/EP300 mutations as revealed by RNA sequencing data

Fig. 6

CREBBP/EP300 mutations promoted macrophage polarization via the FBXW7-NOTCH-CCL2/CSF1 axis in vitro. a The viability of CREBBP^R1392*, EP300^H1377R DB cells (left panel), and CREBBP^kd, EP300^kd DB cells (right panel) when co-cultured with peripheral blood mononuclear cells (PBMCs) at 1:1 ratio or 1:5 ratio for 72 hours. Data are presented as the mean ± SD (N = 3). *P < 0.05 comparing with vector or scramble DB cells. b Main sub-clones shown by t-distributed stochastic neighbor embedding (tSNE) mapping of CREBBP^wt, CREBBP^R1392*, and CREBBP^kd DB cells, co-cultured with PBMCs of 1:5 ratio for 72 h. c Fraction of relative proportions of PBMC sub-clones. d THP-1 differentiation assay using tumor-conditioned media from CREBBP^wt, CREBBP^R1392*, scramble, and CREBBP^kd DB cells. Ratios relative to the control (treated with 320 nM phorbol myristate acetate (PMA) for 24 h) are shown. Data are presented as the mean ± SD (N = 3). e Gene expression of cytokines IL-10 and IL-1β derived from THP-1 cells cultured by tumor-conditioned media from CREBBP^wt, CREBBP^R1392*, scramble, and CREBBP^kd DB cells by quantitative RT-PCR. Data are presented as the mean ± SD (N = 3). f CCL2 and CSF1 expression in CREBBP^wt, CREBBP^R1392*, scramble, and CREBBP^kd DB cells by quantitative RT-PCR. Data are presented as the mean ± SD (N = 3). g Protein expression of CCL2 and CSF1 in CREBBP^wt, CREBBP^R1392*, scramble, and CREBBP^kd DB cells by western blot. h Protein expression of NICD, CCL2, and CSF1 in CREBBP^R1392* or CREBBP^kd DB cells by western blot with and without γ-secretase inhibitor (GSI-I, 50 μm) treatment for 48 h. i Flow cytometry analysis of macrophage markers (CD68 and CD163) in PBMCs, co-cultured with CREBBP^R1392* or CREBBP^kd DB cells with or without GSI-I treatment for 48 h. The results were summed up in the bar graph, presented as the mean ± SD (N = 3)

Fig. 7

CREBBP/EP300 mutations promoted macrophage polarization in murine patient-derived xenografted (PDX) models. a Tumor volume of PDX models injected subcutaneously with CREBBP^Q929* or EP300^I997V tumor tissue from diffuse large B-cell lymphoma (DLBCL) patients. Error bars represent SD (N = 7). **P < 0.01; ***P < 0.001 comparing with those of CREBBP^wt/EP300^wt. b Immunohistochemical assay of H3K27ac in CREBBP^wt/EP300^wt and CREBBP^Q929*/EP300^I997V tumors of PDX models. c Ratio of tumor volume of PDX models upon HDAC inhibitor chidamide (CHID) treatment (day 14 vs day 0). Data are presented as the mean ± SD (N = 7). d Gene expression of FBXW7, HEY1 and HEY2 derived from PDX models with or without CREBBP/EP300 mutations by quantitative RT-PCR. Data are presented as the mean ± SD (N = 3). e Gene expression of CCL2, CSF1 and IL-10 derived from PDX models with or without CREBBP/EP300 mutations by quantitative RT-PCR. Data are presented as the mean ± SD (N = 3)