Comprehensive Epigenome-Wide Profiling Reveals Distinctive DNA Methylation Signatures and Potential Prognostic Biomarkers in Mexican Pediatric B-ALL

Abstract

Acute lymphoblastic leukemia (ALL) is the most common childhood cancer. In Mexico, its higher incidence and lower survival suggest a role for epigenetic factors like DNA methylation (DNAme). We conducted an epigenome-wide association study (EWAS) to define the methylation landscape and identify the profiles associated with ALL and relapse. Bone marrow or peripheral blood samples from pediatric ALL patients at diagnosis and controls without ALL were analyzed using an Infinium MethylationEPIC v2.0 array. Differential methylation was assessed using the ChAMP package. We identified a significant hypermethylated profile in ALL patients compared to controls. Probes in MAD1L1 and RPTOR contained the most differentially methylated CpG sites. Key affected pathways included proliferation, neurotransmission, and neuronal signaling. Survival analysis revealed that hypomethylation of four specific CpGs-cg01052776 (RNH1), cg20747787, cg05001671, and cg01767116 (FBXL22)-was significantly associated with an increased risk of relapse, highlighting their potential as prognostic biomarkers. This study underscores the importance of epigenetic mechanisms in pediatric ALL.

Keywords: DNA methylation; FBXL22; MAD1L1; Mexican pediatric population; RNH1; acute lymphoblastic leukemia; relapse.

Figure 1

Case–control analysis showing differences in methylation levels between pediatric patients with ALL (yellow) and controls (green). (a) Unsupervised hierarchical clustering of differentially methylated CpGs (DMC) based on methylation differences (Δβ) between ALL cases and controls. The heatmap shows a cluster diagram of the top 500 DMCs in ALL patients versus controls. The scale on the right represents the methylation status (red: hypermethylated, blue. hypomethylated). (b) The volcano plot displays the DMCs between cases and controls, with FDR-adjusted <0.05 and absolute Δβ (|Δβ|) >0.2. The x-axis shows the methylation difference (Δβ), and the y-axis represents –log10 of the adjusted p-value. Hypermethylated (red) and hypomethylated (blue) CpGs are highlighted.

Figure 2

Box plot results of the DNA methylation levels derived from ALL cases are shown on CpG islands of promoter regions (a) and on gene biotypes (b) in ALL cases (yellow) and controls (green). The line inside each box represents the median of the methylation β values of the samples. The upper and lower edges of the boxes are the whiskers (75th percentiles and 25th, respectively), while upper and lower lines outside the boxes are the error bars. p-values < 0.05 (Mann–Whitney test) mean that the differences in DNA methylation between ALL patients and controls are statistically significant. 1stExon: first exon; 5′UTR: 5′ untranslated region; TSS: transcriptional start site; PC: protein coding; lncRNA: long non-coding RNA; miRNA: micro-RNA; TUPsdg: transcribed unitary pseudogene; TPPsdg: transcribed processed pseudogene; TUnPPsdg: transcribed unprocessed pseudogene; UPPsdg: unprocessed pseudogene; PPsdg: processed pseudogene.

Figure 3

Gene set enrichment analysis on the genes showing differentially methylated CpGs between ALL patients and controls using ingenuity pathway analysis (IPA). Plot shows the most significant terms for cellular components, biological processes, and molecular functions from the IPAs of DMC-annotated genes (p-values  <  0.05). X-axis: Statistical significance (−log₁₀(p)), where higher values indicate greater relevance. Bubble size represents the number of genes involved in each pathway. Color indicates the IPA z-score, which predicts the direction of regulation. Orange: significant activation (z-score > 2). Blue: significant inhibition (z-score < −2). Gray: statistically significant pathways (p < 0.05) with no predicted direction of change.

Figure 4

Kaplan–Meier survival curves of the most differentially methylated CpGs between ALL detected in the comparative analysis between patients and controls. Patients were categorized into low- and high-methylation groups using the median values of (a) cg01052776, (b) cg01767116, (c) 271 cg05001671, and (d) cg20747787. p-value was calculated using the Cox regression model, and HR indicates hazard ratio. Patients with hypermethylated levels (red lines) in any of these CpGs showed a significantly reduced risk of relapse compared to those with hypomethylated (blue lines).

Figure 4

Kaplan–Meier survival curves of the most differentially methylated CpGs between ALL detected in the comparative analysis between patients and controls. Patients were categorized into low- and high-methylation groups using the median values of (a) cg01052776, (b) cg01767116, (c) 271 cg05001671, and (d) cg20747787. p-value was calculated using the Cox regression model, and HR indicates hazard ratio. Patients with hypermethylated levels (red lines) in any of these CpGs showed a significantly reduced risk of relapse compared to those with hypomethylated (blue lines).

Figure 5

Methylation comparative analysis between ALL cases with early relapse (ER) (n = 7) and without relapse (n = 28) showing the hypomethylation and activation of Rho GTPase signaling in the ER group. (a) Volcano plot displaying differentially methylated CpGs (N = 2629) between ER and no-relapse showing statistical significance (p < 0.05). The x-axis shows the methylation difference (Δβ), and the y-axis represents –log10 of the adjusted p-value. Hypermethylated (red) and hypomethylated (blue) CpGs are highlighted. Gray points represent the non-significant DMP. (b) Summary graph generated by IPA comparing ER and non-relapse ALL cases. Orange color predicts activation, and blue color indicates inhibition. Octagons: biological functions; ovals: transcription regulators; squares: cytokines or growth factors; diamonds: G-protein-coupled receptors; triangles: transcription regulators with unknown or indirect roles. Solid lines: direct relationships; dashed lines: indirect relationships; dotted lines: predicted or less certain associations.