1. Introduction

Chronic lymphocytic leukemia (CLL) is characterized by the accumulation of clonal B cells in peripheral blood, bone marrow, and secondary lymphoid organs. It is the most common leukemia in adults of Western countries and accounts for approximately 2–35 percent of all leukemias in the United States [1]. CLL is more common in men with a male to female ratio ranging from 1.2 : 1 to 1.8 : 1 [1, 2] and is considered to be a disease of the elderly as the median age at diagnosis is 70 years and an incidence increasing rapidly with increasing age [3]. Despite characterized by a common morphology and immunophenotype pattern, CLL shows a variable clinical behavior spectrum. At one end, some CLL patients have a very indolent disease course and may not require treatment for many years, while at the other end, some patients present a very aggressive disease early from diagnosis and require prompt treatment with lower survival rates despite therapy [4].

In terms of genetic pathobiology, at least one of the four common chromosomal abnormalities (deletion 13q14, trisomy 12, deletion 11q22-23, and deletion 17p12) can be detected by interphase fluorescence in situ hybridization in most patients [5]. Most CLL tumors have myriads of somatic gene mutations. Of these, tumor protein 53 (TP53), ataxia telangiectasia mutated (ATM), neurogenic locus, notch homolog protein (NOTCH), and subunit 1 of the splicing factor 3b protein complex (SF3B1) genes are the most frequently mutated at diagnosis [6, 7]. Several other gene mutations seen in CLL tumors are involved in important cellular signaling pathways. NOTCH signaling, B-cell receptor (BCR) signaling, Toll-like receptor, mitogen-activated protein kinase/extracellular signal-regulated kinase (MAPK-Erk) pathway, nuclear factor kappa-light-chain enhancer of activated B cells (NF-kB) signaling, chromatin modifiers signaling, cell cycle signaling, DNA damage signaling, and RNA splicing are some of the many pathways investigated [6–10]. Moreover, it is well documented that the interaction between CLL tumor cells and antigens is regulated by the somatic hypermutation load of the immunoglobulin heavy chain variable region genes [11].

In the complex domain of the tumor microenvironment, CLL cells seem to influence the kind and arrangement of the cells surrounding them, and they are highly dependent on signals from these cells for their own survival and proliferation, encouraging a leukemia-supportive and immunosuppressive microenvironment [11].

Methylation of DNA is an enzyme-mediated modification of DNA structure without interfering in the specific sequence of the base pairs for the gene encoded. Although DNA methylation is important in normal biologic processes, aberrant patterns of methylation are observed in several malignancies. Specifically, two patterns have been mainly described, namely, large sites of global hypomethylation along the genome and localized areas of hypermethylation at CpG islands, within the gene promoter regions. Ιn more and more studies in neoplastic diseases in recent years, there has been an emergence of evidence that methylation of the promoter regions of hundreds of genes, including tumor suppressor genes, results in the failure to express their normal purpose. In other cases, DNA methylation may represent an early step in the pathway by which normal tissue cells undergo tumorigenesis. Some other possible mechanisms for mutational induction secondary to DNA methylation include failure to produce DNA repair proteins that normally protect from mutations and predisposition to increased oxidative DNA damage, resulting in increased mutation potential [12].

Abnormal DNA methylation has been documented in many solid neoplasms, including breast cancer, lung, prostate, colorectal cancer, and melanoma. As described above, hypermethylation of promoter regions inhibits the expression of tumor suppressor genes while hypomethylation activates an oncogene expression. Global hypomethylation is also of crucial role, leading to genomic and chromosomal instability. For instance, BRCA1/2 genes are two of the most studied genes in breast cancer. The hypermethylation of their promoter regions is reported to result in their inactivation and consequent increased risk of breast cancer [13].

As regards CLL, a global DNA methylation pattern was reported to be relatively stable during the disease course and similar both in resting and proliferative cell compartments, implying that aberrant methylation may present as an early leukemogenic event with early indications of specific gene epigenetic changes in CLL samples being described over 30 years ago [14]. Global DNA hypomethylation was reported few years later [15].

Except CLL, abnormal function of basic methylation-related enzymes is widely reported in several hematological malignancies, such as in myelodysplastic syndromes, myeloproliferative neoplasms, acute myeloid leukemia, T-cell acute lymphoblastic leukemia, and diffuse large B-cell lymphoma. The final DNA methylation patterns are suggested to be frequently lineage specific and accompanied by specific secondary mutations [16].

The research in this field led to the study and discovery of a large number of genes whose methylation plays an important role in their regulation, while certain correlations with clinical and laboratory parameters were also identified. The streptothricin acetyltransferase, alpha (SAT-α) gene, whose methylation levels were found to be an independent marker of poor treatment-free survival was an early example [17].

With the appearance and application of newer technics such as whole-genome bisulfite sequencing and DNA methylation arrays, global DNA hypomethylation was confirmed and it was reported that gene body and enhancer regions rather than promoter regions were primarily involved [18, 19] whereas hypermethylation was described to occur mainly in transcribed genomic regions [20] and regions affecting other regulatory mechanisms such as spliceosome [21]. After the CLL genome was discovered to be globally hypomethylated, research for aberrantly methylated oncogene targets revealed that hypomethylation of the B cell lymphoma 2 (BCL2) gene, an important antiapoptotic gene, correlated with higher protein expression in CLL [22]. In the following studies, multiple drug-resistance protein 1 (MDR1) and T-cell leukemia/lymphoma 1 (TCL1) genes were reported to be both hypomethylated and upregulated in CLL [23, 24]. Moreover, the nuclear factor of activated T-cell 1 (NFATc1) gene hypomethylation was also identified and was further shown to be associated with increased mRNA and protein expression, suggesting hypomethylation as a mechanism of constitutive activation of NFATc1 expression in CLL. Through subsequent findings, however, it was discovered that the activation of oncogenes through DNA hypomethylation was a rather infrequent lesion in CLL [25].

To date, several studies have reported gene promoter hypermethylation in CLL patients using both global and single gene approaches. The affected genes include tumor suppressors, transcription factors, genes involved in survival and proliferation, genes with prognostic impact, and microRNA genes. There are several methods that have been used to study gene methylation in chronic lymphocytic leukemia. Each approach has its own advantages and limitations, and the choice of the method depends on the specific research question being addressed. Bisulfite sequencing is considered the gold standard method for analyzing DNA methylation at single-nucleotide resolution. Bisulfite treatment of DNA converts unmethylated cytosine to uracil, while leaving methylated cytosine unchanged. After polymerisation chain reaction (PCR) amplification and sequencing, the pattern of cytosine-to-thymine conversion can be used to determine the methylation status of individual CpG sites. Methylation-specific PCR is a PCR-based method that uses primers specific for either methylated or unmethylated DNA to amplify a region of interest. By comparing the intensity of PCR products amplified using methylated-specific versus unmethylated-specific primers, the methylation status of a given region can be determined. Infinium methylation arrays are microarray-based platforms that can interrogate the DNA methylation status of thousands of CpG sites across the genome. Infinium arrays rely on bisulfite conversion of DNA, followed by hybridization with probes that distinguish between methylated and unmethylated DNA at each CpG site. Methylation-sensitive restriction enzyme digestion relies on the use of restriction enzymes that are sensitive to DNA methylation. By digesting genomic DNA with these enzymes and comparing the resulting fragment patterns with and without prior treatment with a DNA methyltransferase, the methylation status of specific regions can be inferred.

2. The CLL Methylome from a Gene-Specific Perspective

Early in 2004, using genome-wide screening for aberrant promoter methylation in CLL samples, 193 sequences were identified as novel targets for aberrant methylation in CLL. Among them, 173 were homologous to specific genes such as glutamate metabotropic receptor 7 (GRM7), cell division protein kinase 6 (CDK6), T-box transcription factor 3 (TBX3), paired box 5 (PAX5), and protein tyrosine phosphatase nonreceptor type 1 (PTPN1) genes [19]. In another study, where DNA methylation was compared within immunoglobulin heavy-chain gene variable region (IGVH) mutated/unmutated subgroups of CLL, it was reported that in IGVH-unmutated cells, tumor suppressor genes such as von Hippel–Lindau (VHL), Abelson interactor 3 (ABI3), and immunoglobulin superfamily member 4 (IGSF4) were found hypermethylated while genes associated with cell proliferation and migration such as adenosine A3 receptor (ADORA3) and perforin 1 (PRF1) were hypomethylated [26].

Regarding microRNAs, a genome-wide profiling in CLL patients led to the detection of 128 microRNAs carrying aberrantly methylated promoters. Hypermethylated loci included the promoter regions of miR-9-2, miR-124-2, miR-129-2, miR-551b, and miR-708, while miR-21, miR-29a/b-1, miR-34a, miR-155, miR-574, and miR-1204 gene promoters were found hypomethylated [27].

When DNA methylation was studied in paired diagnostic and follow-up samples from IGVH mutated and unmutated CLL patients, genes with prognostic significance including chronic lymphocytic leukemia upregulated 1 (CLLU1), lipoprotein lipase (LPL), zeta chain of T-cell receptor-associated protein kinase 70 (ZAP70) and NOTCH1, the epigenetic regulators histone deacetylases 9 and 4 (HDAC9/4), and DNA methyltransferase 3, beta (DNMT3B), were reported to be aberrantly methylated [14]. Another study used the methylation levels of five specific CpGs to track the cellular origin of CLL, where they seemed to act as enhancers. The varying methylation levels of these biomarkers did not translate into gene expression changes, suggesting that they may not have a functional impact but represented a stable molecular mark [28].

In the context of normal B-cell maturation, CLL DNA methylation was reported to be highly enriched in enhancer and promoter regions, especially in regions of transcriptional elongation and in genes involved in B cell- and lymphocyte-related processes and pathways. Regions targeted for hypomethylation during B cell maturation showed highly significant enrichment for the following six transcription factor families: activator protein 1 (AP-1), early B-cell factor (EBF), runt-related transcription factor (RUNX), octamer-binding transcription factor (OCT), interferon regulatory factor (IRF), and NF-kB [20].

In future, it will be crucial to further develop more efficient and accurate prognostic tools that incorporate clinical, cytogenetic, and molecular data. Therapeutic strategies and agent design are constantly reformed in order to pursue discoveries about the biology of the disease, the pathways involved, and the mechanisms of drug resistance.

To capture the latest discoveries, this review focuses on the present knowledge about single gene/gene promoter methylation in CLL biology while also recording older data that have been published in past papers. Through a review of 126 articles indexed in PubMed, we ended up recording data for 133 genes. Then, we classified them in the following tables according to their associated biological pathways and the related data on CLL studies. It is reasonable for several genes to be associated with more than one pathway. Α characteristic attempt to capture the correlations between most genes and pathways is presented in Figure 1.

In the following tables are presented the genes we identified in our search, with information on their function, as well as the findings of studies regarding CLL. In each table, we introduced in detail, data on the study of methylation or mRNA expression of single genes in CLL patients that showed interesting correlations with clinical and laboratory data.

3. Conclusion

Some studies have focused on the potential clinical impact of DNA methylation in CLL phenotypes. Based on the findings abovementioned, time to treatment, an essential feature of CLL management, was reported to be predicted by the methylation status of SMYD3 gene promoter and UBE2R gene methylation levels [29]. Methylation levels in specific CpG sites located in the gene bodies of REPS1, IL1B, and ATP9B genes also correlated with the time to treatment [29].

Regarding overall survival, hTERT gene promoter hypermethylation was associated with superior overall survival [30], and specific hypermethylated MAFB gene loci correlated with inferior post-treatment survival [66]. Moreover, locus-specific hypomethylation of retrotransposons proximal to DCLK2 and TNFRSF1B was related with higher levels of DCLK2 and TNFRSF1B expression and subsequent inferior survival [88]. When CLL patients were classified upon specific IGVH mutation patterns, promoter methylation of the WNT5A gene showed longer treatment-free survival in a subgroup of the patients [95]. Lower levels of DNA methylation of the ANGPT2 gene were related to a particularly poor prognosis in CLL patients [14, 121].

Findings that could be useful in staging and correlating methylation with pre-existing prognostic factors in CLL include RRM1/2 expression, the promoter methylation status that correlated with lymphadenopathy, 17p gene deletion, Rai stage, and trisomy 12 [39], and the documentation of hypermethylation at specific ITGA4 CpG sites in del13q14+ samples [70]. NFATC1 gene promoter hypomethylation was associated with Binet disease staging [105].

Drug sensitivity and mechanisms of resistance development are another important area of study in CLL. Hypermethylation of specific SLCO3A1 gene loci, a gene which among others is involved in drug uptake mechanisms, has also been associated with inferior post-treatment survival [66]. Another gene, ET-1, is related with drug-resistance potential and its expression is affected upon high methylation levels in its first intron region in patients with CLL [141].

Α last finding that could theoretically have a clinical application in CLL cases is about the SHANK1 gene. A specific CpG locus of its body (hg19) was hypermethylated when compared to control samples. Also, it was even weakly supported that methylation in the same CpG was detectable in blood samples collected years before CLL diagnosis [170]. It is a finding that, if confirmed, is very interesting and could be used in panels for early diagnosis of the disease, certainly in combination with other markers in the context of a population screening program or the surveillance of high-risk patients. Certainly, such methods of investigation have their peculiarities and must be applied after appropriate research and discussion in order to be necessary or beneficial for the patients.

Our findings suggested that certain methylation patterns have been associated with disease progression and survival features and may be used as prognostic markers. In addition, some studies have investigated the use of hypomethylating agents as a part of a therapeutic strategy in CLL.

Regarding findings from previous studies that focused on studying large numbers of genes, data are generally limited but quite intriguing. Kanduri et al. applied high-resolution methylation microarrays (27 578 CpG sites) to CLL samples, which were classified in IGHV-mutated (favorable) and IGHV-unmutated/IGHV3-21 (poor-prognostic) subsets. Results demonstrated significant differences in methylation patterns between these subgroups. In IGHV-unmutated cases, they reported methylation of known or potent tumor suppressor genes (for example, VHL, ABI3, and IGSF4) as well as unmethylated genes involved in cell growth and tumor progression (ADORA3 and PRF1). In contrast, these latter genes were silenced by methylation in IGHV-mutated patients. Moreover, they reported the reinducing of four methylated tumor suppressor genes (including VHL and ABI3) in IGHV-unmutated samples using the methyl inhibitor 5-aza-2′-deoxycytidine [26]. Pei et al. reported a genome-wide DNA methylation analysis with 1,764 gene promoters being identified as differentially methylated in at least one sample when compared with normal B cell samples. Aberrant hypermethylation was discovered in all HOX gene clusters and a significant number of WNT signaling pathway genes. The NFATc1p2 promoter and first intron hypomethylated status correlated with the upregulation of both NFAT protein expressions [25]. Kulis et al. performed a wide analysis of the DNA methylome in normal B cells and CLL samples. They identified widespread hypomethylation targeting mainly the gene body and enhancer regions, suggesting that DNA methylation may be functionally relevant beyond promoter regions. Moreover, they reported that distinct patterns in DNA methylation were recorded between different CLL subtypes [18]. Cahill et al. used high-resolution 450 K arrays to analyze samples from IGHV-mutated or untreated and IGHV-unmutated or treated patients. They identified 2239 differentially methylated CpG regions between IGHV-mutated and -unmutated patients, where the majority of the regions were placed outside annotated CpG islands. Known CLL prognostic genes (i.e., LPL, ZAP70, and NOTCH1), epigenetic regulators (HDAC9/4 and DNMT3B), B-cell signaling, and numerous TGF-β and NF-κB/TNF pathway genes were differentially methylated between the subgroups [14]. Barrow et al. reported 490 differentially methylated regions after exposure to therapy. Among them, 31 were CLL related. Seventeen genes were classified as differentially expressed, following treatment in an independent cohort. Methylation of the HOXA4, MAFB, and SLCO3A1 differentially methylated regions associated with post-treatment patient survival and HOXA4 displayed the strongest association. Reinducing of HOXA4 expression in cell lines and primary CLL cells increased apoptosis following treatment with fludarabine, ibrutinib, and idelalisib [66]. Lastly, Zhang et al. identified 34,797 differentially methylated positions related to CLL. Most of them were hypomethylated and located in gene body sites. They combined these positions with existing DNA methylation and RNA sequencing data and identified regions associated with 1,130 genes whose expression was significantly different in CLL samples [37].

The limitations of this study include possible evidence selection bias because data from statistically significant studies are more likely to be published than those that are not statistically significant. Moreover, in some limited reports, it was not stated whether the methylation study included the gene body, the promoter region, or some other related loci. Lastly, some papers did not reported correlations with patients’ clinical data such as sampling time, staging, or prognostic and treatment features.

To summarize, there are still many mechanisms that need to be investigated in order to define the extent of global aberrant DNA methylation in different prognostic groups, the fundamental role of DNA methylation in sites other than CpG islands, and the interaction of DNA methylation with other regulatory processes in the pathogenesis of CLL. Moreover, with the prospect that there is and will continue to be increasing data on gene methylation, it should be noted that a large number of findings will concern passenger DNA methylation events that need to be identified accordingly.

In conclusion, at some points, the research raises even more questions than answers. Further research is still needed to fully understand the complex interplay between DNA methylation and other epigenetic and genetic alterations in CLL and to develop more effective targeted therapies and prognostic stratification tools for this disease.

Disclosure

Citations are included in the reference list.

Conflicts of Interest

The authors declare that they have no conflicts of interest.

Authors’ Contributions

SC and N-AV had the idea for the article. SC performed the literature search and wrote the first draft of the manuscript. N-AV critically revised the work. All the authors have discussed the results, drafted further work, and approved the version to be published. All the authors have agreed to be responsible for the accuracy and integrity of any part of the manuscript.