Multi-omics analysis reveals a unique epigenetic signature in MYD88 wild-type Waldenström macroglobulinemia

MYD88 L265P mutations (MUT) are found in approximately 90% of patients with Waldenström macroglobulinemia (WM) and lead to the activation of pro-survival signaling.^1,2 Prior studies have demonstrated that patients with MYD88 wild-type (WT) WM have inferior responses to conventional therapies and a higher likelihood of histological transformation to an aggressive lymphoma.³ The importance of the role of non-coding genomics in lymphomagenesis is being increasingly appreciated, but the epigenetic landscape of MYD88^WT WM remains largely underexplored.⁴ Previously, our group showed how DNA methylation and non-coding RNA may alter the control of critical genes which may be responsible for the progression of IgM-monoclonal gammopathy to WM disease.^5-7 Given the demonstrated importance of these non-coding epigenetic regulators, the aim of this study was to utilize a multi-omics approach to assess the role of microRNA (miRNA) and DNA methylation in driving MYD88^WT WM biology through comparison with MYD88^MUT WM patients and healthy controls.

Bone marrow samples were prospectively collected from 20 subjects (3 with MYD88^WT/CXCR4^WT WM, 12 with MYD-88^MUT WM [3 had CXCR4-mutated disease], and 5 healthy controls). WM was defined as ≥10% bone marrow involvement by lymphoplasmacytic lymphoma and serum IgM monoclonal protein of any size. Positive selection for CD19⁺ and/or CD138⁺ B cells was performed on samples utilizing STEMCELL CD19⁺ and CD138⁺ selection kits (17854, 17877). Clonal sorting was not performed. Analyses were conducted as per prior reports and focused on comparisons between MYD88^WT WM and normal controls and between MYD88^WT WM and MYD88^MUT WM patients.^5-7 Briefly, RNA differential expression for miRNA and mRNA was determined using a log₂ fold change (FC) >0.5 or <-0.5 and false discovery rate (FDR) <0.05. Genome-wide DNA methylation analysis was performed, and CpG methylation ratios were segmented into 200 bp regions. Differentially methylated regions (DMR) were determined using a Q-value <0.05 and an absolute methylation difference ≥10%. The absolute methylation difference of ≥10% was established after investigation of different threshold values, as it allowed sufficient identification of DMR while effectively reducing background noise. Next, RNA-sequencing data from matched samples were integrated to the methylome and analyzed to identify differentially expressed genes (log₂FC >0.5 or <-0.5 FDR <0.05) with corresponding promoter hypo- or hypermethylated DMR. For miRNA analysis, miRNA-mRNA target analysis was performed, and differentially expressed miRNA-mRNA pairs with a correlated biological expression (i.e., either upregulated miRNA experimentally predicted to regulate downregulated mRNA or vice versa) were selected. Ingenuity Pathway Analysis was used to determine potentially epigenetically regulated pathways using a corrected P<0.05. This study received approval from the Mayo Clinic Institutional Review Board.

The patients’ characteristics for the whole cohort and stratified by MYD88 mutation status are included in Table 1. Baseline characteristics were comparable between MYD-88^MUT and MYD88^WT WM patients. All MYD88^WT WM patients were CXCR4^WT, and three patients from the MYD88^MUT WM cohort were CXCR4^MUT. The characteristics of the healthy control samples were not available because of institutional policy.

Comparing MYD88^WT WM samples to controls, a greater number of hypermethylated DMR were observed across genomic regions (hyper- vs. hypomethylated): promoter (29,129 vs. 6,300), 3’ untranslated region (UTR) (2,068 vs. 1,551), 5’UTR (10,169 vs. 735), intron (30,344 vs. 24,511), exon (22,224 vs. 6,855), CpG island (30,372 vs. 1,240) and CpG shore (14,477 vs. 5,872) (all P<0.0001) (Figure 1A). Next, mRNA analysis demonstrated that 352 differentially expressed genes had concordant up-/downregulation with promoter hypo-/hypermethylated DMR.

On the assessment of promoter-methylation-based pathways, comparing MYD88^WT WM to controls, multiple pathways involved in cell signaling, proliferation, gene regulation, cytokine and immune signaling, and metabolism were found to be differentially regulated (Figure 1B). Most pathways were observed to be negatively enriched, including several members of interleukin (IL) signaling (IL-2, IL-3, IL-7) and regulators of cell proliferation and signaling, including mTOR, PI3K/AKT, and p53. Underlying multiple pathways were promoter hypermethylation and mRNA downregulation of AKT3, MAP3K1, and PIK3CA genes.

Next, methylation studies comparing MYD88^WT to MYD88^MUT WM samples demonstrated a similar hypermethylated profile in MYD88^WT with a statistically greater number of hypermethylated DMR observed across genomic regions (hyper- vs. hypomethylated): promoter (24,829 vs. 2,544), 3’UTR (3,534 vs. 378), 5’UTR (6,111 vs. 605), intron (52,292 vs. 5,360), exon (21,066 vs. 2,187), CpG island (18,334 vs. 1,551) and CpG shore (16,351 vs. 1,957) (all P<0.0001) (Figure 1C). Assessing mRNA expression, 212 differentially expressed genes had concordant up-/downregulation with promoter hypo-/hypermethylated DMR.

Methylation-based pathway analysis comparing MYD88^WT to MYD88^MUT WM samples showed downregulation of cytokine signaling, including IL-8, IL-17, and regulators of intracellular signaling, including RAC (Figure 1D). Of note, we have previously demonstrated reduced levels of IL-8 in WM and, in the present analysis, we found hypermethylation with associated downregulation of downstream RAS superfamily genes, including RHOC and RRAS2 (Figure 1E).⁸ Additionally, underlying multiple downregulated cytokine pathways (IL-8, IL-4, IL-13) was promoter hypermethylation with associated mRNA downregulation of BCL2, an important regulator of apoptosis, and a current therapeutic target in WM.⁹ In relation to RAC inactivation, we observed potential upstream inhibition of this pathway via promoter hypermethylation of the integrin genes ITGAM (integrin alpha M) and ITGB1 (integrin beta 1), both found to be downregulated.¹⁰

On miRNA analysis, comparing MYD88^WT WM to control samples, we observed 192 differentially expressed transcripts (70 downregulated; 122 upregulated) experimentally targeting 6,255 mRNA (Figure 2A). After target analysis and filtering, pathway analysis demonstrated multiple intracellular signaling, and metabolism pathways were downregulated (Figure 2B). Here, we observed miRNA-based downregulation of the RAC signaling pathway, a pathway also downregulated in our methylation analysis. Underlying this pathway was upregulation of miR-98-3p (FC: 2.4; FDR <0.005) and let-7a-5p (FC: 1.1; FDR <0.005) targets of PIK3C3 and MAP3K1, respectively, found to be downregulated. These targets are members of the PI3K/AKT and MAPK family, and prior reports have described the role of miR-98 in regulating tumor progression via targeting of these pathways.¹¹ Relating back to our methylation analysis, we observed similar downregulation of PI3K/AKT and MAPK signaling comparing MYD88^WT WM to controls (Figure 1B); thus, these findings indicate multiple epigenetic regulators may influence key intracellular signaling pathways that regulate cytokine and immune function.

In comparing MYD88^WT to MYD88^MUT WM samples, our miR-NA analysis identified 11 differentially expressed transcripts (all downregulated) experimentally targeting 3,011 differentially expressed mRNA (Figure 2C). Here, pathway analysis revealed upregulation of chromatin organization and histone modification signaling (Figure 2D). Underlying both pathways was the downregulation of miR-138-5p (FC: -9.8; FDR: 0.015), which targets DNMT3A and DOT1L, both of which were observed to be upregulated on mRNA analysis. Of relevance, miR-138-5p has been demonstrated to act as a tumor suppressor, and DOT1L is a histone methyltransferase involved in the regulation of mRNA transcription.^12,13 Furthermore, DNMT3A is a critical epigenetic regulator responsible for the addition of methyl groups at CpG sites; thus, upregulation of this gene corresponds with the CpG hypermethylation seen across genomic regions in MYD88^WT WM patients (Figure 2E).¹⁴

Overall, we identified for the first time a distinct epigenetic signature in MYD88^WT WM. An important limitation of this study is the smaller sample size of patients with MYD88^WT WM given the rarity of this diagnosis. To address this limitation, we sought to compare our findings with a previously published cohort of 18 patients with MYD88^WT disease.¹⁵ In this prior report, utilizing gene set enrichment analysis comparing MYD88^WT to MYD88^MUT WM patients, significant enrichment for the upregulation of E2F, MYC, PI3K-AKT-MTOR, and G2M checkpoint signaling targets and the downregulation of inflammatory response genes and TNFA signaling via NF-κB was found.¹⁵ Performing a similar mRNA-based gene set analysis comparing our MYD88^WT to MYD88^MUT WM samples, we observed a concordant pattern (Online Supplementary Table S1) with positive enrichment of E2F, MYC, MTOR and G2M checkpoint signaling, and negative enrichment of TNFA signaling via NF-κB and inflammatory response signaling (all P<0.05). These consistent patterns offer external support for the observations presented in the current study.

Table 1.Patients’ characteristics of Waldenström macroglobulinema samples.

Figure 1.Genome-wide methylation analyses with promoter-methylation-based pathway integration. (A, B) Analysis evaluating differentially methylated regions (DMR) comparing patients with MYD88^WT Waldenström macroglobulinemia (WM) to controls (A) and methylation-based pathway analysis evaluating promoter methylation and mRNA expression (B). (C, D) DMR analysis comparing WM patients with MYD88^WT to those with MYD88^MUT (C) with corresponding pathway analysis (D). (E) IL-8 pathway analysis demonstrating downstream involvement of RHOC and RRAS2.

Figure 2.MicroRNA (miRNA) differential expression with miRNA-based pathway analysis. (A, B) Volcano-plot demonstrating differential expression miRNA between patients with MYD88^WT Waldenström macroglobulinemia (WM) and controls (A) and miRNA-based pathway analysis integrating mRNA differential expression (B). (C, D) miRNA differential expression comparing patients with MYD88^WT and MYD88^MUT WM (C) with miRNA-based pathway analysis (D). (E) A postulated mechanism whereby downregulation of miR-138-5p permits increased DNMT3A expression and consequent CpG hypermethylation. Volcano-plots demonstrate log₂(fold change) data and false discovery rate (FDR).

A central feature of our analysis of MYD88^WT WM patients was the epigenetic inactivation of cytokine and intracellular pathways involved in immune regulation, including the IL-8 and the PI3K/AKT/mTOR pathways. Although constitutional activation of MYD88 serves a critical role in the pro-inflammatory signaling observed in MYD88^MUT WM patients, these findings suggest that both methylation and miRNA-based mechanisms may epigenetically regulate the reduced inflammatory signaling in MYD88^WT WM beyond only the mutation in MYD88.^1,2 In relation to treatment, given that BTK inhibitor efficacy relies on suppression of activated inflammatory pathways, these findings also suggest a potential non-coding mechanism for the reduced response to BTK inhibitors in patients with MYD88^WT WM. Furthermore, our observation of BCL2 promoter hypermethylation with mRNA downregulation indicates that an additional epigenetic mechanism may confer resistance to venetoclax in MYD88^WT WM, which has been predominantly evaluated in patients with MYD88^MUT disease.⁹

Overall, the results of this study offer novel insights into the biology of MYD88^WT WM, suggest its epigenetic distinction from MYD88^MUT WM, and propose how DNA methylation and non-coding RNA may contribute to the unique clinical phenotype of patients with this disease. Some of the key epigenetic findings identified, such as the non-coding downregulation of cytokine and inflammatory signaling pathways, may play a role in the inferior outcomes and reduced response to conventional WM-directed therapies in patients with MYD88^WT WM. Larger studies are needed to investigate the epigenetic similarities between MYD88^WT WM and other B-cell lymphomas, including marginal zone lymphoma, which may help to improve biological subclas-sification. Furthermore, functional studies are needed to validate these findings at the proteomic level and explore the potential utility of epigenetic therapies for the treatment of MYD88^WT WM.

Footnotes

• Received August 28, 2025
• Accepted November 24, 2025

Correspondence

*KC and JP contributed equally as first authors

S.M. ANSELL - ansell.stephen@mayo.edu

Disclosures

SMA receives research funding from Bristol-Myers Squibb, Seattle Genetics, Affimed Therapeutics, Regeneron, Trillium Therapeutics, AI Therapeutics and ADC Therapeutics. The other authors have no conflicts of interest to disclose.

Contributions

Funding

This study was supported by research funding from the BINK Foundation, and the International Waldenström Macroglobulinemia Foundation (IWMF) to SMA and the Mayo Clinic. PM is supported by the NCI K08CA279652, LRF Clinical Investigator Award (CDA 1020588), the Gerstner Family Career Development Award, the ASH Junior Faculty Scholar Award, the IWMF/LLS Strategic Research Roadmap Initiative, and the IWMF/LLS Enhanced Research Roadmap Initiative.