Extended Data Fig. 14: NR3C1 binding to the CXCR4-SE is abrogated by somatic mutations.

a. NR3C1 ChIP–qPCR of the region encompassing the CXCR4-SE mutational hotspot or the NFKBIA control region and a negative control region in the WT SUDHL16 cell line (mean +/− SD; n = 4 technical replicates, from one representative experiment out of 2 independent experiments that gave analogous results, one-way ANOVA with Bonferroni correction). Data are expressed as fold enrichment vs. control IgG IP. b. NR3C1 ChIP–qPCR (left) and allelic quantification in the mutant (right) HLY1 cell line (mean +/− SD; n = 3 technical replicates, from one representative of 2 independent experiments, one-way ANOVA with Bonferroni correction and two-tailed Fisher’s exact test). c. Simplified schematic of the CXCR4 locus; the recurrently hypermutated SE is shown in red, and the primers used for NR3C1-ChIP are approximately positioned below the map. d. Gel electrophoresis of NR3C1 ChIP–PCR amplicons from the indicted cell lines, as compared to input and control IgG ChIP. Band quantification was obtained by densitometry and the relative values are provided below the image, with input set as 1 (data shown are representative of 2 independent experiments). e. Sequencing analysis of the PCR products shown in d. On the top panel, the reference CXCR4 genomic sequence (NM_003467) and the sequence of the mutated allele are aligned to the predicted NR3C1 binding motif (reverse strand). Sequencing traces of ChIP–PCR amplicons document reduced signal for the mutant allele, as compared to the input (arrow), indicating abrogation of NR3C1 binding by the mutations. f. Sequencing traces of DNA and cDNA amplicons obtained from the same clones shown in e. Arrows indicate the mutated position. g. Relative changes in CXCR4 expression between unedited and CXCR4-corrected isogenic clones (n = 8 each for HLY,1 and n = 6 each for BJAB), colour coded as in Fig. 6e. For each cell line, the mean value of unedited clones is set as 1 (two-tailed unpaired t-test). h. CXCR4 allelic expression in clones surviving correction of the CXCR4-SE mutation. Top: Schematic diagram of the wild-type (WT) and mutant (M) CXCR4 alleles in the HLY1 cell line, carrying the A413G nucleotide substitution. Additional SNVs segregating with the two alleles and used to track allele-specific expression are also indicated, in red. The nucleotide sequence of the two alleles with the predicted NR3C1 binding motif, is shown below the diagram and is aligned to the sequencing tracks (reverse strand) of representative DNA and cDNA amplicons obtained from isogenic HLY1 clones (control, clones edited in the neutral genomic region; corrected, clones corrected in the 413 position within the CXCR4-SE). Arrow points to the mutated/corrected nucleotide. i. Relative distribution of genetic lesions affecting the CXCR4:NR3C1 axis in DLBCL. Overlap between cases of DLBCL with mutations in the CXCR4-NR3C1 binding site (red) and/or in the NR3C1 coding exons (blue). Data are from 315 cases analysed by WGS or Sanger sequencing. j. CXCR4 expression levels in cases of primary DLBCL with WT vs. mutated CXCR4-SE sequences (n = 181 cases with available WGS and RNA-seq data). Data are expressed as TPM. k. Percentage of cases harbouring CXCR4-SE mutations in DLBCL COO subtypes. The total number of cases analysed within each subtype is provided in brackets, and the number of mutated cases is shown on the top. l. Percentage of cases harbouring CXCR4-SE mutations in different LymphGen genetic classes. P-values were calculated by two-tailed Fisher’s exact test for enrichment of a genetic lesion in a specific LymphGen class versus the other classes combined. The total number of cases analysed within each class is provided in brackets, and the number of mutated cases is shown on the top. Data indicate a significant enrichment in the BN2 subtype (two-tailed Fisher’s exact test).

Source data