Article Text
PDF
Abstract
Aims DNA methylation has its distribution influenced by DNA demethylation processes with the catalytic conversion of 5-methylcytosine (5mC) into 5-hydroxymethylcytosine (5hmC). Myelodysplastic syndrome (MDS) has been associated with epigenetic dysregulation of genes related to DNA repair system, chronic immune response and cell cycle.
Methods We evaluated the tissue DNA methylation/hydroxymethylation in bone marrow trephine biopsies of 73 patients with MDS, trying to correlate with the mRNA expression of 21 genes (POLH, POLL, REV3L, POLN, POLQ, POLI, POLK, IRF-1, IRF-2, IRF-3, IRF-4, IRF-5, IRF6, IRF-7, IRF-8,IRF-9, MAD2, CDC20, AURKA, AURKB and TPX2).
Results The M-score (5mC) was significantly higher in patients with chromosomal abnormalities than patients with normal karyotype (95% CI –27.127779 to –2.368020; p=0.022). We observed a higher 5mC/5hmC ratio in patients classified as high-risk subtypes compared with low-risk subtypes (95% CI –72.922115 to –1.855662; p=0.040) as well as patients with hypercellular bone marrow compared with patients with normocellular/hypocellular bone marrow (95% CI –69.189259 to –0.511828; p=0.047) and with the presence of dyserythropoiesis (95% CI 17.077703 to 51.331388; p=0.001). DNA pols with translesion activity are significantly influenced by methylation. As 5mC immunoexpression increases, the expressions of POLH (r=−0.816; r2 =0.665; p=0.000), POLQ (r=−0.790; r2=0.624; p=0.001), PCNA (r=−0.635; r2=0.403; p=0.020), POLK (r=−0.633; r2=0.400; p=0.036 and REV1 (r=−0.578; r2=0.334; p=0.049) decrease.
Conclusions Our results confirm that there is an imbalance in the DNA methylation in MDS, influencing the development of chromosomal abnormalities which may be associated with the low expression of DNA polymerases with translesion synthesis polymerases activity.
- myelodysplastic syndromes
- DNA
- medical oncology
- immunohistochemistry
Data availability statement
All data relevant to the study are included in the article or uploaded as supplemental information.
Statistics from Altmetric.com
Introduction
DNA methylation is characterised by a covalent modification in the cytosine nitrogenous base, in which a methyl group is inserted through a methyl donor called s-adenosylmethionine in carbon 5' of CpG dinucleotide by the action of DNA methyltransferases, originating 5-methylcytosine (5mC).1 2 DNA methylation has its distribution influenced by DNA demethylation processes, which can occur passively by dilution after DNA replication during cell division or through an active process initiated by the enzymes called Ten-Eleven-Translocation 1–3. These enzymes are responsible for the catalytic conversion of 5mC into 5-hydroxymethylcytosine (5hmC).3 4
Myelodysplastic syndrome (MDS), the most common cancer of bone marrow in the elderly, has been associated with epigenetic dysregulation.5 Abnormal methylation of promoter site is very common in MDS and the number of genes involved is increased in the advanced stages of the disease,6–8 such as MDS with excess blasts (MDS-EB)9 and cases classified as high risk according to the Revised International Prognostic Score System (IPSS-R).10 The pathogenesis of MDS has been linked to dysregulation of DNA repair system, chronic immune response, apoptosis, and cell cycle,6–8 ultimately leading to chromosomal abnormalities and mutations in up to 95% of cases.11
We evaluated the tissue DNA methylation/hydroxymethylation (5-mC and 5hmC) in bone marrow trephine biopsies of patients with MDS, trying to correlate with the mRNA expression of 21 genes related to DNA repair system (DNA polymerases with translesion activity—POLH, POLL, REV3L, POLN, POLQ, POLI, POLK), chronic immune response (interferon regulatory factors—IRF-1, IRF-2, IRF-3, IRF-4, IRF-5, IRF6, IRF-7, IRF-8 and IRF-9), and mitotic spindle and checkpoint (MAD2, CDC20, AURKA, AURKB and TPX2). We detected a very significant negative correlation between bone marrow methylation and expression of polymerases with translesion activity (DNA Pols), suggesting a possible link to the genesis of chromosomal abnormalities in myelodysplastic syndrome.
Patients and methods
Patients
We evaluated 73 bone marrow trephine biopsies (for tissue DNA methylation and hydroxymethylation) and 30 bone marrow samples (for gene expression of 21 genes), at diagnosis, of 73 unique patients with MDS from Drug Research and Development Center, Federal University of Ceara, Brazil (NPDM-UFC). All cases were diagnosed according to 2016 WHO classification9 and risk stratification was performed based on the IPSS-R.10 All cases were treatment-naive patients. Bone marrow trephine biopsies of 10 bone marrow donors were used as healthy control.
Tissue microarray (TMA) preparation
For immunoexpression, we used the TMA technique. TMA is an organised collection of tissue samples arranged in the form of a matrix where each sample is large enough to be representative and conveniently small to allow the non-predatory use of the block. In each block, the appropriate location for the best visualisation of bone marrow tissue was chosen by reading H&E slides for each case. Then, the blocks were drilled (1 mm in diameter, 4 mm in height) and transferred to the receptor pore corresponding to the prepared TMA block. The blocks were heated at 42°C for 3 min and sectioned to 4 µm thick.
Immunohistochemistry (IHC)—tissue DNA methylation (the M-score: 5-methylcytosine%) and tissue DNA demethylation (5-methylcytosine/5-hydroxymethylcytosine%)
IHC was performed on a 4 µm paraffin-embedded BM biopsy section on poly-L-lysine coated slides. Sections were dewaxed in xylene and rehydrated by passing through graded alcohol. Antigen retrieval was done with 0.01 M citrate buffer (pH 6). The sections were treated with peroxidase blocking for 5 min and then incubated with a serum blocking agent for 15 min to reduce the non-specific staining. The sections were incubated with primary antibody (anti-5mC: Abcam—Ab214727ug50; anti-5hmC: Abcam—Ab214728ug50) in a moist chamber at room temperature overnight and then incubated with biotinylated secondary antibody for 30 min. The antibody was visualised with 3,3’-diaminobenzidine and counterstained with Mayer’s hematoxylin. The negative internal controls of the reactions were obtained in reactions without the primary antibodies and positive controls with sections of tissues indicated according to the manufacturer’s instructions for each antibody.
Immunoreactivity for 5mC and 5hmC was evaluated by three independent observers (RFP, FDRD and GMC; figure 1). For each patient, the cell counts were repeated three times, and the mean values were used to calculate a final methylation of tissue 5mC% (M-score of methylation) and the ratio of 5mC to 5hmC (5mC/5hmC%) commonly reported as a marker of tissue DNA demethylation.12 Interobserver and intraobserver variability were evaluated, and concordance was >80%. Any discrepancy of more than 10% in each case among the three independent observers was re-evaluated.
Representative images of 5hmC and 5mC expression in bone marrow tissue. (A) Low expression of 5-hydroxymethylcytosine (5hmC) (1%). (B) High expression of 5hmC (60%). (C) Low expression of 5-methylcytosine (5mC) (M-score 1%). (D) High expression of 5mC (M-score 80%). The ratio of 5mC to 5hmC (5mC/5hmC%) was considered the marker of tissue DNA demethylation.
Cytogenetic analysis—chromosomal abnormalities
The bone marrow cells were cultivated as previously reported for chromosome analysis.13 14 Briefly, the bone marrow cells were separated into two short-term cultures with 7 mL RPMI 1640 media (Gibco) (pH 7.0), 3 mL fetal bovine serum (Invitrogen), and incubated for 24 hours at 37°C. Colcemid (50 µL) was added to the culture for 30 min, followed by 0.075 M KCl at room temperature for 20 min and Carnoy’s fixative for 5 min, four times. Slides were prepared and submitted to G-banding and whenever possible at least 20 metaphases were analysed. Conventional G-banded karyotype analyses were performed from bone marrow cells of 73 patients with MDS. The metaphases were analysed using CytoVision Automated Karyotyping System (Applied Imaging, San Jose, California, USA). Results were reported according to International System Human Cytogenetic Nomenclature (ISCN) 2016.
Total RNA/DNA extraction and cDNA synthesis
Thirty bone marrow samples were obtained by aspiration at the same time of bone marrow trephine biopsy (at diagnosis) and used to evaluate the expression of genes from the DNA repair system, chronic immune response, and cell cycle regulation. Total RNA from isolated mononuclear cells were performed with TRizol Reagent (Invitrogen, Carlsbad, California, USA), and cDNA synthesis was generated from total RNA with the High Capacity cDNA Reverse Transcription kit (Applied Biosystems, San Jose, California, USA), according to the manufacturer’s protocol. cDNA samples were stored at −20°C until further use.
mRNA expression analysis by quantitative real-time PCR
Quantitative real-time PCR reactions were based on TaqMan methodology (Applied Biosystems, Carlsbad, California, USA) and performed on 7500 Fast System (Applied Biosystems, Carlsbad, California, USA). Predeveloped TaqMan gene expression assays (Assays-on-Demand, Applied Biosystems, Carlsbad, California, USA) for genes related to mitotic spindle and checkpoint (AURKA: Hs01582072_m1, AURKB: Hs00945858_g1, TPX2: Hs00201616_m1, MAD2: Hs01554513_g1, and CDC20: Hs00426680_mH), DNA polymerase with translesion (TLS) activity (POLH: Hs00197814_m1, POLL: Hs00203191_m1, REV3L: Hs00161301_m1, POLN: Hs00394916_m1, POLQ: Hs00981375_m1, POLI: Hs00969214_m1, POLK: Hs00211965_m1), and interferon regulatory factors (IRF1: Hs00971980_m1, IRF2: Hs00180006_m1, IRF3: Hs01547283_m1, IRF4: Hs01056533_m1, IRF5: Hs00158114_m1, IRF6: Hs01062178_m1, IRF7: Hs01014809_g1, IRF8: Hs01128710_m1, IRF9: Hs00196051_m1) as well as the TaqMan Universal Master Mix II (Applied Biosystems, Carlsbad, California, USA) were used to quantify mRNA expression. To normalise input cDNA, reference endogenous genes were chosen according to a previously described method.14 Each sample was performed in duplicate and the expression ratios were calculated using the 2−∆Cq method15 from the Cq values provided by the 7500 real-time PCR System software (Applied Biosystems, Foster City, California, USA).
Statistical analysis
The M-score of methylation (5mC%) and the 5mC/5hmC ratio of DNA demethylation were associated with clinical and prognostic features in 73 patients. These scores we also evaluated regarding the expression of DNA polymerase with translesion (TLS) activity, interferon regulatory factors and mitotic spindle and checkpoint genes—total—21 genes). Normality was evaluated by the Shapiro-Wilk test. Outliers were removed. The Student’s t-test or one-way analysis of variance with Tukey/Games Howell posthoc test were used to analyse the influence of relative expression regarding variables: age, WHO classification,9 low risk versus high risk, bone marrow cellularity, dysplasias (number of dysplasias, dyserythropoiesis, dysgranulopoiesis, dysmegakaryopoiesis), IPSS-R variables10 (blast count, haemoglobin count, absolute neutrophil count, platelets), number of cytopenias, karyotype (normal vs abnormal, aneuploidy, number of alterations, classification of IPSS-R), transfusion dependence, death and transformation into AML. The variance homogeneity was evaluated by Levene’s test.
The M-score of methylation and the ratio 5mC/5hmC of DNA demethylation were correlated with the mRNA expression of DNA polymerases with TLS activity, IRFs genes and mitotic spindle/checkpoint genes. Pearson’s correlation test or Spearman’s correlation (if normal distribution or not, respectively) was used for obtaining the r and the r-square (r2). Statistical analysis was performed using the SPSS V.24.0 (SPSS) and GraphPad Prism V.8 (GraphPad Prism software, La Jolla, California, USA) software. Probability level (p value)<0.05 was assumed.
Results
Patients characteristics
The clinical and laboratory features of patients are presented in table 1. The patient’s mean age was 67 years (range 28–91) while for controls (bone marrow donors) was 66 years (range 48–82 years) with no significant difference (p>0.05). During follow-up, only one case evolved into AML and there were six deaths. The most relevant results are shown in graphical abstract (figure 2).
Summary of clinical and prognostic characteristics of patients with myelodysplastic syndrome
Graphical abstract. Levels of 5-methylcytosine (5mC) M-score was high in patients with abnormal karyotype. The ratio of 5mC to 5-hydroxymethylcytosine (5hmC; 5mC/5hmC%) was high in patients with dyserythropoeieses, bone marrow hypercellular and advanced myelodysplastic syndrome (MDS) forms. Correlations between DNA polymerases with TLS activity mRNA expression and immunoexpression of 5mC. Correlations between interferon regulatory (IRF6) and ratio of 5mC to 5hmC (5mC/5hmC%).
5-Methylcytosine (5mC) levels (M-score) regarding karyotype. Expression of 5mC (M-score) regarding karyotype. Student’s t-test with Levene’s test for homogeneity of variances; 21 cases with normal karyotype; 17 cases with chromosomal abnormalities.
Regarding the IPSS-R,10 the majority of cases were low risk (53.3%). Cytogenetic analyses were performed for all patients with 30 (41.1%) normal karyotype, 22 (30.13%) showing chromosomal abnormalities and 21 (28.77%) presented no metaphases. Regarding chromosomal abnormalities, 36,35% were classified as good/very good, 36.35% as intermediate and 27.3% as high/very high according to IPSS-R10 (table 1).
Tissue Methylation Score (M-score) of 5mC
Tissue Methylation Score M-score (5mC) is higher in patients with chromosomal abnormalities than patients with normal karyotype
The M-score (5mC) was significantly higher in patients with chromosomal abnormalities than patients with normal karyotype (95% CI –27.127779 to –2.368020; p=0.022) (figures 2 and 3). Of utmost importance, the majority of chromosomal abnormalities among patients with MDS were complex karyotype, abnormalities of chromosome 7, abnormalities of chromosome 17, and markers (chromosomes that cannot be identified by microscopy—according to ISCN 2016), all results commonly associated with genomic instability and reported as unfavourable/intermediate prognosis in MDS10 (table 1). These results reinforce the link between the increase of DNA methylation and genomic instability. All data regarding M-score analysis are presented in online supplemental file table S1.
Supplemental material
Tissue Methylation Score M-score (5mC) and other clinical and laboratory features
We found no significant associations between the M-score of 5mC with variables of gender (p=0.081), cytopenias according to IPSS-R (p=0.988), blast percentage (p=0.119), bone marrow cellularity (p=0.344), number of dysplasia (p=0.237), ring sideroblast (p=0.918), transfusion dependence (p=0.178), 2016 WHO classification (p=0.989) and IPSS-R score (p=0.168).
Ratio of 5mC to 5hmC (5mC/5hmC%; DNA demethylation)
The ratio of 5mC to 5hmC, a potential target for DNA demethylation,12 regarding clinical and prognostic variables was also evaluated (all data are presented in online supplemental file table S1).
MDS presents a high 5mC/5hmC ratio (low demethylation)
Comparing patients with MDS to healthy controls, we observed an increased 5mC/5hmC ratio in patients with MDS (95% CI 1.873 to 66.460; p=0.039), suggesting low DNA demethylation (figure 4A). As expected, this result reinforces problems of DNA demethylation in MDS.
5-Methylcytosine (5mC) into 5-hydroxymethylcytosine (5hmC) ratio regarding clinical and laboratory features. Student’s t-test with Levene’s test for homogeneity of variances. (A) 5mC/5hmC ratio regarding MDS versus control. (B) 5mC/5hmC ratio regarding age. (C) 5mC/5hmC ratio regarding risk (2016 WHO). (D) 5mC/5hmC ratio regarding bone marrow cellularity. (E) 5mC/5hmC ratio regarding dyserythropoiesis.
Elderly patients present a high 5mC/5hmC ratio (low demethylation)
According to IPSS-R,10 elderly patients (>60 years) show worse prognosis than patients with less than 60 years old. We detected a higher 5mC/5hmC ratio in patients older than 60 years compared with patients younger than 60 years (95% CI –57.420660 to –5.912673; p=0.017; figure 4B), suggesting low demethylation in the elderly.
High-risk WHO (MDS-EB 1 and 2) subtypes present a high 5mC/5hmC ratio (low demethylation)
According to WHO 2016, we divided patients into low risk (MDS-SLD, MDS-MLD and MDS-RS) and high risk (MDS-EB 1 and 2).9 We observed a higher 5mC/5hmC ratio in patients classified as high-risk subtypes compared with low-risk subtypes (95% CI –72.922115 to –1.855662; p=0.040; figure 4C). We did not detect any significant difference (p=0.230) between IPSS-R categories by ANOVA (although the mean range of high-risk cases was higher than low-risk cases) probably due to the low number of cases in each group and the absence of metaphases in 28.7% of cases which precluded its classification (cytogenetics is fundamental to IPSS-R determination of risk).
Patients with dyserythropoiesis and hypercellular bone marrow present a high 5mC/5hmC ratio (low demethylation)
Regarding bone marrow cellularity, the 5mC/5hmC ratio was higher in patients with hypercellular bone marrow compared with patients with normocellular and hypocellular bone marrow (95% CI –69.189259 to –0.511828; p=0.047; figure 4D). Analysing the presence of dysplasia among all bone marrow lineages, we detected an increased 5mC/5hmC ratio in patients with dyserythropoiesis (95% CI 17.077703 to 51.331388; p=0.001) compared with the patients without dyserythropoiesis (figure 4E).
Correlations between the 5mC/5hmC ratios and expressions of genes related to DNA translesion, chronic immune response, mitotic spindle and checkpoint genes
Regarding the 21 genes evaluated, we detected very significant high negative correlations between genes related to DNA translesion with M-score of 5mC, demonstrating a very similar behaviour. Regarding chronic immune response (interferon regulatory genes), mitotic spindle and checkpoint genes (cell cycle), there were very inconsistent associations.
DNA pols with TLS activity are significantly influenced by methylation (immunoexpression of 5mC-M-score)
As 5mC immunoexpression increases, the expressions of POLH, POLQ, PCNA, POLK and REV1 decrease. We observed strong negative correlations between the M-score of 5mC and the expressions of genes POLH (r=−0.816; r2=0,665 p=0.000), POLQ (r=−0.790; r2=0.624; p=0.001), PCNA (r=−0.635; r2=0.403; p=0.020), and POLK (r=−0.633; r2=0.400; p=0.036) and REV1 (r=−0.578; r2=0.334; p=0.049; additional files online supplemental figure S1 and table S2).
Interferon regulatory genes expression are not significantly influenced by methylation
We found a strong negative correlation between the 5mC/5hmC ratio and IRF6 mRNA expression (r=−0.612; r2=0.374; p=0.020).
Discussion
We evaluated DNA methylation and hydroxymethylation in bone marrow trephine biopsies of patients with MDS searching for associations between clinical and laboratory variables of patients at diagnosis with the mRNA expression of 21 genes related to the pathogenesis of MDS. We detected two very important results: (1) hypermethylation of bone marrow is associated with chromosomal abnormalities and low expression of genes related to DNA repair; (2) there are more significant associations when using the 5mC/5hmC ratio regarding the prognosis of MDS.
Hypermethylation, as here demonstrated by high 5mC (the M-Score), increases genomic instability by diminishing expression of tumour suppressor genes responsible for maintaining chromosome alignment, the correct progression of metaphase and the DNA repair. Among these patients here reported with chromosomal abnormalities, the majority showed alterations classified as intermediate, poor and very poor prognosis such as −7, del(7q) and complex karyotype, according to IPSS-R.10 Römermann et al 16 also reported previously the association between global methylation and chromosomal abnormalities of unfavourable prognosis in MDS using pyrosequencing technology. But our great difference is to show, to the best of our knowledge, for the first time, that this association in MDS is linked to a lower expression of DNA repair genes.
Our cells are continually exposed to DNA-damaging agents and the DNA repair processes rapidly target the damaged DNA for repair, but some lesions persist and block genome duplication. Trying to avoid the dangerous consequence of a stalled replication fork, our cells use specific polymerases to traverse the damage, the TLSs. There are 15 DNA polymerases with TLS activity in human cells.17 18 We detected a very significant negative correlations between the expressions of POLH, POLQ, PCNA, POLK, and REV-1 with 5mC levels. As 5mC immunoexpression increases, the expressions of POLH, POLQ, PCNA, POLK, and REV1 decrease. These genes with TLS activity are very important to maintain genomic stability and its downregulation due to methylation may predispose to chromosomal abnormalities. Using the r2 score, we detected increased bone marrow methylation was able to reduce 65% of POLH, 62% of POLQ, 40% of PCNA, 40% of POLK and 33% of REV1 expressions. DNA pols with TLS activity are very important during DNA repair because are the final chance of correction before a lesion becomes a chromosome abnormality.17 18
As chromosomal abnormalities are very important markers of prognostic and overall survival in MDS,10 our results suggest new pathways of pathogenesis and possible new targets to be evaluated in MDS. These results also reinforce the important link between hypermethylation and genomic instability in patients with cancer.
Epigenetic events play an important role in cancer and have been extensively reported in MDS pathogenesis.7 Using a very cheap methodology, we detected the immunoexpression of 5mC and 5hmC in bone marrow sections using TMA preparations simultaneously, reporting the M-score (5mC) of methylation and the score of demethylation (5mC/5hmC ratio). This methodology gave us the possibility to evaluate a large cohort of patients in a parallel analysis with standardisation. The most significant results were associated with the demethylation score (5mC/5hmC ratio) which showed low demethylation in elderly, high-risk MDs (MDS-EB), dyserythropoiesis and hypercellular bone marrow, all markers of unfavourable prognosis. But which marker (M-Score of 5mC or 5mC/5hmCratio) should we choose for evaluating methylation/demethylation status in bone marrow tissue of patients with MDS? The distribution of the demethylated cytosine (5hmC is tissue dependent (including liver and placenta)19 and are generally less than those of 5mC. The ratio of 5mC/5hmC has been reported as more constant across many genomic features,20 making it possible more reliable in real time regarding the intrinsic link between methylation and demethylation.
This study presents limitations. It was not possible to analyse all the bone marrow trephines by TMA due to the poor quality of some samples. The IHC has inherent limitations to the technique, such as the presence of nonspecific cell labelling, and difficulty in reading samples with hypocellular bone marrow.21 Although we evaluated 21 genes, we studied its expressions in only 30 patients, difficulting a complete match with the 73 bone marrow biopsies
Our results confirm there is an imbalance in the DNA methylation in MDS, influencing the development of chromosomal abnormalities which may be associated with the low expression of DNA polymerases with TLS activity.
Take home messages
There is an imbalance in the global DNA methylation in MDS, influencing the development of chromosomal abnormalities and creating possible new targets for pathogenesis and treatment.
Negative correlations between POLH, POLQ, PCNA, POLK, and REV-1 with 5mC levels are related to genomic instability and its downregulation due to methylation may predispose to chromosomal abnormalities in MDS.
These results intensify the necessity of study the effects of global methylation in MDS.
Data availability statement
All data relevant to the study are included in the article or uploaded as supplemental information.
Ethics statements
Patient consent for publication
Ethics approval
The study was approved by the local scientific ethical committee of Federal University of Ceara and all participants provided written informed consent.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
GMC and DPB are joint first authors.
Handling editor Mary Frances McMullin.
GMC and DPB contributed equally.
Contributors GMC, APNNA and RFP designed the study. GMC, DSdP and AMS constructed the tissue microarray and IHC. TMA scoring was done by GMC, DPB, FDRF and RFP. GMC and DPB performed all the statistical analyses under supervision from HLR-J and RFP. SMMM, CLMF and RTGdO contributed to the acquisition and interpretation of the data. The manuscript was written by DPB, GMC and RFP. All authors read and approved the final manuscript.
Funding This study has been supported by CAPES, CNPq (PRONEX) and FUNCAP.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.