AUTHORS
Zhiquan Wang, Ph.D, Huihuang Yan, PhD2, Mingma Hoel3, Daniel Bihnam, MS4, Cristine Allmer, BS5, Chantal McCabe6, Dennis P. Robinson7, James R. Cerhan, MD, PhD8, Richard R Furman, MD9, Neil E Kay, M.D.10, Sameer A. Parikh, MBBS11, Susan L. Slager, PhD12, Esteban Braggio, PhD13
AFFILIATIONS
1 Mayo Clinic, Rochester, Minnesota, United States
2 Mayo Clinic, Rochester, Minnesota, United States
3 Mayo Clinic, Rochester, Minnesota, United States
4 Mayo Clinic, Rochester, Minnesota, United States
5 Mayo Clinic, Rochester, Minnesota, United States
6 Mayo Clinic, Rochester, Minnesota, United States
7 Mayo Clinic, Rochester, Minnesota, United States
8 Mayo Clinic, Rochester, Minnesota, United States
9 Weill Cornell Medical College, New York, United States
10 Mayo Clinic, Rochester, Minnesota, United States
11 Mayo Clinic, Rochester, Minnesota, United States
12 Mayo Clinic, Rochester, Minnesota, United States
13 Mayo Clinic, Phoenix, Arizona, United States
ABSTRACT
Background: Chronic lymphocytic leukemia (CLL) shows ancestry-associated differences in incidence, clinical presentation, and outcome, yet the regulatory mechanisms underlying these differences remain poorly defined. Most genomic and epigenomic studies of CLL have been performed in cohorts dominated by individuals of European ancestry (EA), limiting their generalizability. We integrated chromatin accessibility, histone modifications, and gene expression in purified CLL B cells from genetically confirmed African ancestry (AA) and EA patients to define ancestry-associated regulatory programs and determine whether transcriptional differences are supported by coordinated changes across multiple chromatin layers.
Methods: ATAC-seq was performed in 15 AA and 51 EA patients. CUT&Tag sample sizes varied by mark, with at least 7 AA and 8 EA cases per mark for H3K27ac, H3K4me1, H3K4me3, H3K36me3, H3K27me3, and H3K9me3; RNA-seq was generated for all samples. Differential regulatory regions were identified using edgeR and linked to expressed genes using nearest-gene assignment, ENCODE regulatory annotations, and Hi-C datasets. We first characterized ancestry-associated differences in chromatin accessibility and individual histone modifications. We then applied a stringent multi-mark concordance framework requiring consistent regulatory direction across ATAC-seq, H3K27ac, H3K4me1, H3K4me3, and H3K36me3, with H3K27me3 incorporated as an additional repressive-chromatin feature. Concordant genes were stratified as AA-up or AA-down by RNA expression, and pathway enrichment was assessed using GO Biological Process, KEGG, Reactome, and MSigDB Hallmark gene sets with Benjamini-Hochberg FDR <=0.05.
Results: ATAC-seq identified 4,955 regions with increased and 3,000 regions with decreased accessibility in AA compared with EA CLL (fold change >2; FDR <=0.05). Increased-accessibility regions mapped to 2,313 genes, including 599 AA-up genes enriched for TNF-alpha/NF-kB and apoptosis-related programs. Decreased-accessibility regions mapped to 1,953 genes, including 317 AA-down genes enriched for interferon-gamma response and IL-6/JAK/STAT3 signaling. Histone profiling demonstrated ancestry-associated remodeling across active, transcription-associated, and repressive chromatin features. Reduced H3K27ac in AA CLL was linked to 68 downregulated genes enriched for interferon-gamma response and IL-6/JAK/STAT3 signaling, whereas increased H3K27me3 was linked to 399 genes involving related immune-regulatory programs. These patterns were consistent with distinct regulatory mechanisms, including loss of active chromatin, gain of repression, and coordinated multi-mark remodeling.
The stringent multi-mark analysis identified 148 AA-up and 83 AA-down concordant genes whose expression direction was supported across accessibility and histone modification profiles. Significant pathway enrichment was restricted to the AA-up set, with 7 GO Biological Process and 3 Hallmark terms. The strongest Hallmark signal was TNF-alpha signaling via NF-kB (FDR=8.17x10^-8), accompanied by IL2/STAT5 signaling (FDR=0.0175) and inflammatory response (FDR=0.0497). GO terms converged on extrinsic apoptotic signaling (best FDR=0.00294). No pathway reached FDR <=0.05 among AA-down concordant genes. Sensitivity analyses using less stringent 4-mark and 5-mark definitions showed the same directional pattern, supporting the robustness of the AA-up regulatory program.
Conclusions: Ancestry-associated differences in CLL extend across chromatin accessibility, active and repressive histone modifications, and transcription, and converge into coordinated multi-layer regulatory programs. These findings establish epigenomic remodeling as a mechanistic component of ancestry-associated CLL biology, broaden molecular models developed largely in EA cohorts, and identify regulatory circuits that may contribute to disease heterogeneity and inform future functional and therapeutic investigation.
DISCLOSURES
Zhiquan Wang, Ph.D: No relevant conflicts of interest to declare.