It has been challenging to understand how recurrent somatic mutations of genes encoding major chromatin factors contribute to B cell non-Hodgkin lymphoma pathogenesis. In this issue of Blood, Yang et al1 demonstrate that, in the case of missense mutations targeting the gene CREBBP, it all depends on a stoichiometric tug-of-war between wild-type and mutant enzymes at a small set of CD40 target genes.
Mutations in chromatin-modifying genes are key drivers of epigenetic deregulation in germinal center (GC)–derived B-cell lymphomas (BCLs).2,3 Recurrent CREBBP mutations occur in ∼60% of follicular lymphoma (FL) and ∼30% of diffuse large BCL (DLBCL) cases,4 as well as in ∼50% of FL cases that undergo transformation to aggressive disease,5 typically presenting as DLBCL. Approximately half of these mutations affect the lysine-acetylase domain (KAT), reducing catalytic activity and resulting in a partial or total loss of function. These mutations are heterozygous, indicating that pathogenicity likely arises from an overall reduction in acetylation.4 Indeed, mutually exclusive recurrent mutations in EP300, a CREBBP paralog, occur in up to ∼40% of FL and DLBCL cases (15% of transformed FL). Understanding how these mutations contribute to lymphomagenesis is crucial for therapeutic targeting.3
CREBBP and EP300 function as transcriptional coactivators that regulate gene expression via histone acetylation, modifying chromatin accessibility. A key acetylation target is H3K27Ac, which marks gene active regulatory regions. Previous studies demonstrated that CREBBP or EP300 deletion in mouse GC B cells significantly alters chromatin at critical regulatory regions.6 Specifically, CREBBP is predominant in light zone GC B cells, whereas EP300 dominates in the dark zone, reflecting distinct transcriptional programs in these compartments.6 Mutations in either enzyme may disrupt GC homeostasis and the ability of B cells to switch between these two programs, facilitating malignant transformation. The pathways affected by these deficiencies and the functional consequences of missense mutations vs deletions remain unclear. For instance, patients with Rubinstein-Taybi syndrome, marked by germ line mutations and deletions of CREBBP and EP300, show a complex relationship between mutation type and clinical severity.7
Yang et al addressed these questions using CRISPR editing to introduce the 3 most common KAT domain CREBBP mutations (R1446C, Y1482N, and Y1503C) in the CREBBP locus of 2 DLBCL cell lines, RL and HT, which harbor monoallelic EP300 frameshift mutations. Considering the importance of acetyltranferase activity levels, a decrease in EP300 expression in these cells could affect the functional outcomes of CREBBP mutations.
Epigenetic and transcriptomic profiling of these isogenic mutant cell lines revealed that KAT mutations produced a distinct chromatin and transcriptional landscape compared to wild-type and knockout (KO) cells. Loss of H3K27Ac occurred at specific regulatory regions in KAT mutants but not in KO cells, indicating a dominant-negative effect rather than simple loss of function. Unlike KO cells, which retained EP300 recruitment to chromatin, KAT point-mutants (KAT-PM) failed to recruit EP300, preventing functional compensation. Consequently, KAT-PM cells displayed a more divergent epigenomic and transcriptomic profile from wild-type controls than KO cells.
Mutations predominantly affected enhancers dynamically regulated by CREBBP and EP300 during dynamic GC B-cell transitions. These regions exhibit nonoverlapping binding patterns for EP300 (dark zone) and CREBBP (light zone), mirroring findings from murine GC studies.6 In mutant cells, loss of CREBBP binding at these sites was not compensated by EP300, and disrupted the expression of MHC genes, light zone signature genes, and CD40-responsive genes (elements crucial for GC B-cell function). Notably, CD40 signaling was significantly impaired in KAT-cells but remained intact in KO cells. Similar results were observed in FL tumors with CREBBP mutations and in xenografts derived from patients with DLBCL, thus validating these findings in primary lymphoma samples.
CD40 signaling is pivotal in GC biology, as it captures T-cell signals that guide B cells to expand, diversify, and differentiate.8 Transcription factors responding to CD40 (eg, NF-κB and IRF4) interact with CREBBP and EP300 to activate transcription. Yang et al explored whether enhancing CD40 signaling in KAT-PM cells could restore normal transcriptional responses. In wild-type RL and HT cells, enforced CD40 signaling increased CREBBP recruitment to regulatory regions. In KAT-PM cells, CD40 stimulation recruited EP300 to sites where mutant CREBBP failed to bind, partially restoring transcriptional responses and NF-κB activity. Overexpression of IRF4 further rescued the phenotype by saturating mutant CREBBP molecules, thereby freeing EP300 for chromatin binding.
Next, the authors tested whether bispecific antibodies that engage T cells could elicit similar effects. These bispecifics, designed to facilitate B-T–cell interactions, restored EP300 loading and transcriptional activation of CD40-responsive genes in mutant cells. This response was accompanied by increased antigen presentation, enabling T cells to exert cytotoxic effects on lymphoma cells. The parallels between enforced CD40 signaling and bispecific antibody treatment suggest that lymphoma responses to these novel immunotherapies may involve not only direct cytotoxicity but also the upregulation of pathways essential for B-T cell communication. These findings highlight the potential of immunotherapeutic strategies to counteract epigenetic deregulation in lymphoma.
The selective targeting of CD40 programs by CREBBP mutants raises intriguing possibilities. Not all GC B cells may be equally susceptible to CREBBP mutations; the extent of their impact may depend on the strength of CD40 signaling, which dictates cell fate decisions in B cells and is influenced by T-cell help.8 Previous studies have proposed that epigenetic heterogeneity within the GC may affect responses to immune signals,9 and Yang et al’s findings suggest that CREBBP mutations could exploit this heterogeneity. The importance of epigenetics in GC dynamics and lymphomagenesis cannot be understated, and the role of CD40 signaling in GC-derived lymphomagenesis may be more critical than previously recognized.
Most importantly, the study highlights the dominant-negative nature of CREBBP KAT mutations, which outcompete their wild-type counterparts and prevent functional compensation by EP300. This distinction underscores the risks of equating loss-of-function mutations with genetic KOs, because their biological consequences may diverge significantly. The authors’ “zombie enzyme” model challenges the assumption that mutant enzymes are simply inactive, drawing distant parallels to pseudokinases that acquire unexpected biochemical activities.10 Whether CREBBP KAT mutants possess unrecognized functions remains an open question, but this work suggests they may not be entirely dead after all, raising intriguing possibilities for future investigation.
Conflict-of-interest disclosure: The author declares no competing financial interests.
This feature is available to Subscribers Only
Sign In or Create an Account Close Modal