Abstract Introduction: Ribosomal DNA (rDNA) encodes ribosomal RNA (rRNA), which undergoes processing to become ribosomes. Although rDNA dysregulation has been observed in various pathologies, comprehensive epigenetic profiling in cancer has been limited due to its repetitive sequence. To overcome this challenge, a customized genome has been developed to map human and mouse rDNA using high-throughput sequencing data. Using this genome and more than 2,200 ChIP-seq datasets, Antony et al performed a systematic analysis to identify the transcription factors, e.g., C/EBP-A, that support rRNA production in hematopoietic cells. Our previous publications suggested that DNA cytosine modifications influence genomic binding of C/EBP. Accumulated studies, including our own, have reported aberrant DNA methylation and hydroxymethylation in myeloid malignancies. Yet, it remains unclear whether the leukemia-associated DNA methylation dysregulation extends to genomic regions encoding rRNA (rDNA), and how these epigenetic changes in rDNA affect downstream targets. Methods: To address these critical knowledge gaps, we employed innovative rDNA assembly methods, along with a DNA methylation analysis pipeline, to evaluate the whole-genome bisulfite sequencing (WGBS) data from 136 cancer patients covering 9 cancer types, aiming to uncover DNA methylation changes in rDNA in cancer. To further investigate unique patterns observed in hematological malignancies, we expanded our analysis to a dataset containing DNA methylation profiles from 292 AML patients. Additional AML datasets were analyzed to validate our initial findings with integrative methylation analysis methods. Furthermore, we analyzed published nanopore long-read data to independently confirm the methylation status of rDNA. Results: Our pan-cancer analysis revealed several patterns of rDNA methylation across cancer types, often at odds with non-rDNA methylation patterns. Intriguingly, we discovered that hematological malignancies exhibited distinct DNA methylation signatures in rDNA, with higher DNA methylation in the rDNA-coding region compared with other cancer types. This finding was validated in multiple independent datasets totaling 484 AML samples. Additionally, long-read nanopore sequencing verified these methylation patterns across the entire rDNA tandem array. Further characterization of this hypermethylation in AML patient samples revealed CpG loci that show highly conserved changes. A significant correlation was observed between hypermethylation at these loci and decreased time-to-relapse, suggesting that increased rDNA coding region methylation may serve as a biomarker. Interestingly, along with prognostic significance related to time-to-relapse, we revealed distinct patterns of rDNA methylation in patients treated with hypomethylating agents (HMA). To study potential mechanisms, we first validated published findings that many transcription factors exhibit consistent binding patterns on rDNA. To elucidate the connection between DNA methylation and transcription factor binding within rDNA, we first performed a motif analysis of the rDNA sequence, revealing 4000+ canonical binding motifs. We then used our large AML datasets to identify motifs with differential methylation between healthy bone marrow and AML patients, both at diagnosis and relapse. Across both datasets, we identified several potential transcription factors, e.g., CEBP families, for downstream analysis. We will use multiple leukemia cell lines to validate the correlation between DNA methylation and TF binding at rDNA. We will further use catalytically dead Cas9 (dCas9) mediated epigenome editing to probe the causal relationship to support our hypothesis that rDNA methylation can influence TF binding. Conclusions: Our data revealed distinct DNA methylation signatures in rDNA specific to myeloid leukemia. Using multiple published DNA methylation datasets from AML patients, we identified unique DNA methylation changes in rDNA, highlighting its potential diagnostic and prognostic value. At the molecular level, we observed differential DNA methylation levels within several TF binding motifs, which might alter TF binding and impact their transcriptional activity in rDNA, ultimately contributing to aberrant ribosomal biogenesis. Overall, our study provides a comprehensive analysis of previously neglected DNA methylation changes in rDNA in AML, uncovering novel epigenetic mechanisms with potential clinical relevance.
Yun Huang (Mon,) studied this question.