Luzna P; Flodrova P; Janovska L; Zapletalova J; Minarik J; Kolar Z; Trtkova KS

Research Article

Ann Hematol Oncol. 2019; 6(7): 1257.

Different Gene Methylation Status of the CDKN2B and/ or PDLIM4 as the Result of Comparative Analysis to the Global DNA Methylation in Unsorted Cell Population of Multiple Myeloma Patients

Luzna P¹, Flodrova P^1,2, Janovska L^3,6, Zapletalova J⁴, Minarik J^2,5, Kolar Z^1,2 and Trtkova KS^1,4*

¹Department of Clinical and Molecular Pathology, Faculty of Medicine and Dentistry, Palacky University Olomouc, Czech Republic

²University Hospital Olomouc, Olomouc, Czech Republic

³Department of Microbiology, Faculty of Medicine and Dentistry, University Palacky Olomouc, Czech Republic

⁴Institute of Molecular and Translational Medicine, Faculty of Medicine and Dentistry, Palacky University Olomouc, Czech Republic

⁵Department of Hemato-oncology, Faculty of Medicine and Dentistry, Palacky University Olomouc, Czech Republic

⁶Department of Chemical Biology and Genetics, Centre of the Region Haná for Biotechnological and Agricultural Research, Faculty of Science, Palacký University, Šlechtitelů 27, CZ-783 71, Olomouc, Czech Republic

*Corresponding author: Katerina Smesny Trtkova, Laboratory of Molecular Pathology, Department of Clinical and Molecular Pathology, Faculty of Medicine and Dentistry, University Palacky Olomouc, Hnevotinska 3, 775 15 Olomouc, Czech Republic

Received: June 22, 2019; Accepted: July 04, 2019;Published: July 11, 2019

Abstract

Background: Multiple Myeloma (MM) is a hemato-oncological disease characterized by clonal expansion of malignant plasma cells in the Bone Marrow (BM). Apart from genetic changes, such as point mutations, deletions or translocations, it is well known, that in pathogenesis of MM are also involved epigenetic changes such as DNA methylation. Methylation of both CDKN2B gene, representing an inhibitor of cyclin dependent kinases, and PDLIM4 gene, one of potential tumor suppressor genes engaged in MM evolution, were evaluated in newly diagnosed multiple myeloma patients.

Methods: The quantification of the global DNA methylation at 5´-CCGG- 3´sequence using LU minometric Methylation Assay (LUMA) and the colorimetric quantification of the global DNA methylation were performed. Bisulfite-treated DNA in 13 CpGs of a promoter, and 16 CpGs of the first exon of the CDKN2B gene, 9 CpGs of the PDLIM4 gene promoter were analyzed by pyrosequencing.

Results: Studied CDKN2B gene regions revealed CpGs methylation in the range 2.8 - 6%, whereas PDLIM4 gene promoter showed increased level of methylated CpGs in the range 13.1 - 27%. We found a strong positive correlation between the global DNA hypomethylation (LUMA) and CDKN2B expression (r = 0.766, P < 0.01), and strong negative correlation between global DNA hypermethylation (LUMA) and PDLIM4 promoter methylation level (r = -0.994, P< 0.01). Our data indicate functional unmethylated CDKN2B gene, in contrast to methylated tumor-suppressor PDLIM4 gene in newly diagnosed multiple myeloma patients.

Conclusion: In unsorted bone marrow cells of newly diagnosed multiple myeloma patients, the CpG methylation pattern of the studied CDKN2B and PDLIM4 genes varies depending on overall DNA methylation level. Their different methylation status determined in both global DNA hypomethylated and hypermethylatied groups of patients could be related to a followed progression of the multiple myeloma disease. On the base of statistical analysis, the PDLIM4 gene show significantly increased methylation state with negative correlation to the detected DNA methylation level. These methylation changes of the PDLIM4 gene can contribute to pathogenesis of myeloma and its methylation status acts as a prognostic factor.

Keywords: Methylation; Neoplastic cells; DNA

Abbreviations

AML: Acute Myelogenous Leukemia; BM: Bone Marrow; CDKN2A and CDKN2B: Cyclin Dependent Kinase Of Inhibitors 2A and 2B; CML: Chronic Myelogenous Leukemia; CML: Chronic Myelogenousleukemia; LUMA: Luminometricmethylaion Assay; MGUS: Monoclonagammopathy of Undetermined Significance; MDS: Myelodysplastic Syndrome; MM: Multiple Myeloma; MtL: Methylation Level; NF-κβ: Nuclear Factor - κβ; SMM: Smoldering Multiple Myeloma; PDLIM4: PDZ and LIM Domain Protein 4; TGF-β: Transforming Growth Factor - β.

Background

Monoclonal gammopathies are clonal proliferations of plasma cells, also referred as plasma cell dyscrasias. The most important condition among plasma cell dyscrasias is Multiple Myeloma (MM). Multiple myeloma is characterized by malignant proliferation of clonal Plasma Cells (PC). The overt/active form of the disease is accompanied by various extent of organ impairment. The organ and tissue alteration is usually described with the acronyme “CRAB” - hypercalcaemia, renal insufficiency, anemia and lytic bone lesions. Smoldering Multiple Myeloma (SMM) is an intermediate asymptomatic malignant plasma cell disorder, with a risk of progression to symptomatic MM of ~10 % per year, for the first 5 years [1,2]. Multiple myeloma is a multistep transformation process evolving from its premalignant state - Monoclonal Gammopathy of Undetermined Significance (MGUS), with transformation rate of ~1% per year. Molecular biology of MM is complex and involves important and not yet fully understood relationship between neoplastic cells and microenvironment of the bone marrow, which is known to assert one of the most important processes in the disease´s evolution [3]. Very important role in MM pathogenesis belongs to different cytogenetic abnormalities - hyperdiploid or hypodiploid genome, translocations with very high incidence of IgH gene involvement, deletions and amplifications. Some of these changes are associated with poor prognosis in MM patients. Apart from genetic abnormalities, epigenetic modifications are known to participate in pathogenesis of monoclonal gammopathies.

DNA methylation belongs to a group of epigenetic changes together with histone posttranslational modifications and changes in microRNA. Epigenetic modifications deal with affecting of gene expression in other levels than in changes in DNA sequence. The key role of DNA methylation during the physiological processes as embryogenesis, ontogenesis and aging is well known. Tumor development and progression caused by DNA methylation changes have been already described as well [4-6]. Aberrant DNA methylation seems to be an important event in multiple myeloma pathogenesis. DNA methylation patterns have been found to change as multiple myeloma progresses [7], and differential methylation at certain gene loci has an association with adverse outcomes [8,9].

In this study, we used two methodological procedures – colorimetric and Luminometric Methylation (LUMA) assays to detect the global DNA methylation with the followed CpG methylation pattern analysis of CDKN2B and PDLIM4 genes detected by bisulfite pyrosequencing. In unsorted bone marrow cells of newly diagnosed multiple myeloma patients, the CpG methylation pattern of the studied CDKN2B and PDLIM4 genes varies depending on overall DNA methylation level. On the base of statistical analysis, the PDLIM4 gene show significantly increased methylation state with negative correlation to the detected DNA methylation level. These methylation changes of the PDLIM4 gene can contribute to pathogenesis of myeloma or could act as a prognostic factor.

Methods

Patient samples

Unsorted cell population in the bone marrow aspirate of patients with symptomatic multiple myeloma were studied. The diagnosis of MM followed International Myeloma Working Group criteria. The study was approved by the Ethics Committee of University Hospital Olomouc, and samples were collected after informed consent.

ELISA analysis of global DNA methylation

Genomic DNA was obtained from unsorted cell population in the bone-marrow aspirates. The extraction was performed using QIAamp^® DNA Blood Kit (Qiagen, Hilden, Germany) according to manufacturer´s protocol and the 260/280 nm absorbance ratio was used to assess the purity of DNA. Isolated DNA (100 ng) was used for a colorimetric detection of the global methylation status (% 5mC) by MethylFlash™ Methylated DNA Quantification Kit (Epigentek, Farmingdale, NY, USA).

Luminometric Methylation Assay (LUMA)

Quantification of global DNA methylation was examined using Luminometric Methylation Assay [10,11]. Briefly, genomic DNA was isolated by the Wizard^® Genomic DNA Purification kit (Promega, Madison, WI, USA).Extracted genomic DNA (~400 ng) was treated with sodium bisulfite using EpiTect^® Bisulfite kit according to the manufacturer’s instructions (Qiagen, Hilden, Germany). DNA after bisulfite modification was cleaved using two restriction enzymes, HpaII (methylation sensitive) and MspI (methylation insensitive) in two separate reactions in the presence of EcoRI to standardize for DNA amounts. The LUMA assay is a global rather than a gene-specific assay and from this reason, it is performed without a reference genome [10]. Digested DNA were then used as templates for pyrosequencing with the Pyromark Q96 ID instrument and Pyromark Gold Q96 reagents (Qiagen, Hilden, Germany). The luminometric signals produced by either the sequential incorporation of C and G nucleotides (reflecting the number of 5´-CCGG- 3´ sequence digested by HpaII or MspI) or the sequential incorporation of A and T nucleotides (reflecting the number of 5´-AATT- 3´ sequence digested by EcoRI), were then quantified using Pyromark Q96 ID software 1.0. The global methylation percentage per sample was then calculated as follows [12]: (Average signal obtained with HpaII after EcoRI normalization / Average signal obtained with MspI after EcoRI normalization) • 100.

Bisulfite methylation analysis by pyrosequencing

DNA methylation of PDLIM4 promoter and the CDKN2B promoter and first exon gene regions was determined by pyrosequencing method [13-15]. Bisulfite treatment of extracted genomic DNA was done as describe above. Primers for PCR and following pyrosequencing reaction were designed using PyroMark Assay Design SW 2.0 (Qiagen, Hilden, Germany) (Table 1). For PCR reaction 1 μl of bisulfite treated DNA was added in a 25-μl PCR reaction mixture containing 1 x PyroMark PCR Master Mix (Qiagen, Hilden, Germany), 1 x CoralLoad Concentrate (Qiagen, Hilden, Germany), 1.5 mM MgCl2, 0.2 μM forward primer and 0.2 μM biotinylated reverse primers. For Hot Start Taq Polymerase activation, the PCR reaction mixture was initially denaturated at 95°C for 15 min followed by 45 cycles of denaturation at 94°C for 30 sec, annealing at 56°C for 30 sec, elongation at 72°C for 30 sec, and the final extension at 72°C for an additional 10 min after the last cycle. The final biotinylated PCR product was immobilized on Streptavidin Sepharose^® HP (GE Healthcare, Pittsburg, PA), precipitated with 70% ethanol, passed through denaturation step and then washing step using PyroMark Q96 Vacuum Workstation (Qiagen, Hilden, Germany). The amplicons were transferred to each well of the PyroMark Q96 plate containing 40 μl of 0.4 μM sequencing primer diluted in annealing buffer (Qiagen, Hilden, Germany). Control unmethylated and methylated DNA (Qiagen, Hilden, Germany) after bilusulfite treatment were part of a set of analyzed MM patient samples. Pyrosequencing analysis was performed according to the PyroMark CpG Sofware 1.0.11 (Qiagen, Hilden, Germany). Methylation was quantified in terms of the methylation level (MtL) as the average percentage of cytosines methylated per CpG: MtL (%) = (S% methylated cytosines) / No. of CpGs analyzed).

Table 1: Primer sequences corresponding to bisulfite-treated DNA used for pyrosequencing.




  
    Gene 
    PCR Primer Forward 
    PCR Primer Reverse 
    Pyrosequencing
      Primer 
    CpGs 
  
  
    RILpromot    (216 bp)
    5´GGGTTTATG- 
      AGGAGGTATT- 
      TGAGTTG-3´
    5´-biotin-ACACC- 
      CCCACTCAACT- 
      CTC-3´
    5´-TGTAGATA9 
      GTTGGGTTTG- 
      G-3´
    �
  
  
    P15promot    (223bp)
    5´-AGGAGTTG-    AGGGTAGTGG- 
      T-3´
    5´-biotin-TCCCC- 
      ACCCCCTTAAA- 
      CT-3´
    5´-GGATATTT13 
      AGAGAGTAG- 
      TGTAGTTA-3´
     
  
  
    P15exon 
      (138bp) 
       
    5´-AGGAGGGG-    TAGTGAGGAT- 
      3´
    5´-biotin-ACTTT- 
      TCCTAACCCTC- 
      AAAAACCAAC-3´
    5´-GGGGTAG-    TGAGGATT-3´
    16



Table 1: Primer sequences corresponding to bisulfite-treated DNA used for pyrosequencing.

Quantitative real-time PCR (qRT-PCR)

Total RNA was isolated from unsorted population of bone marrow samples using TRI Reagemt^® BD (Molecular Research Center, Inc., OH, USA). Input of 100 ng of total RNA was reverse transcribed using Transcriptor First Strand cDNA Synthesis kit (Roche, Basel, Switzerland). qRT-PCR with TaqMan probes and Xceed qPCR Probe Mix (Institute of Applied Biotechnologies, Prague, Czech Republic) was performed using Light cycler^®480 System (Roche, Basel, Switzerland). Expression of CDKN2B (Hs00793225_m1) and PDLIM4 (Hs00896441_m1) genes was normalized to expression of endogenous housekeeping control GAPDH (Hs01041237_g1) gene. All probes were provided by Thermo Fisher Scientific (Waltham, MA, USA) and commercially available human BM total RNA (Takara Bio USA, Inc., Mountain View, CA, USA) was used as the calibrator for 2-^ΔΔCt quantification approach.

Statistical analysis

Both methylation and gene expression data were quantilenormalized to make the distributions the same across patient samples. A Pearson correlation coefficient (r) was calculated for each global DNA methylation and gene expression probe pair, for global DNA methylation and gene methylation, and for gene methylation and gene expression. A negative correlation was defined when the directionality of changes for gene expression and DNA methylation were in the opposite direction (e.g. presence of methylation and low expression, or viceversa). A positive correlation occurred when the directionality of changes was the same between DNA methylation and gene expression (e.g. presence of methylation and positive expression, or viceversa).

In the methylation parameter, for measure of central tendency, median as well as mean (MtL) were included in calculations. The set of 42 MM patient samples were divided on the basis of global methylation status determined: hypomethylated, moderately methylated or hypermethylated. This approach is referred as “the discretization approach” because samples are discretized into one of the three groups [16]. Thus, samples with methylation level < μ – δ (mean plus one standard deviation) were classified into the hypomethylated group; samples with methylation level > μ + δ were classified as hypermethylated; and all other remaining samples vere considered as moderately methylated samples. Followed, the global DNA methylation was determined in MM samples using both colorimetric detection (ELISA) and Luminometric Methylation Assay (LUMA) as well. Based on the global methylation status (% 5mC) detected by LUMA assay, the hypomethylated group (68,28 < μ - ρ), the moderately methylated and the hypermethylated group (75,724 > μ - ρ) of the same set of MM patient samples were formed (Supplement I). Furthermore, on the basis of values obtained by ELISA method, we classified the hypomethylated group (0,572 < μ - ρ), the moderately methylated group, and the hypermethylated group (1,428 > μ - ρ) of MM patient samples (Supplement II). Statistical significance of difference among tested groups of patients was analyzed using post- hoc test Mann-Whitney U test with Bonferroni´s correction of significance. All tests were made at a significance level of 0,05. Data were statistical analyzed with IBM SPSS Statistics, v.22.

Results

Global DNA Methylation Versus Gene Expression

Generally, at both monitored files detecting global DNA methylation LUMA and ELISA assays, a level of the PDLIM4 expression was increased in comparison to the CDKN2B gene expression (Table 2). Furthermore, the highest level of PDLIM4 expression was detected in MM patients with the lowest level of global DNA methylation - the hypomethylated group (Figure 1, Table 2). On the other hand, a high level of the global DNA methylation, in the hypermethylated group was accompanied by a high CDKN2B expression (mean value 3,85) (Figure 2, Table 2).

Table 2: The set of 42 MM samples was divided on the basis of the global DNA methylation parameter determined by both LUMA and ELISA assays, and the level of both CDKN2B and PDLIM4 gene expressions. *Values were not included in the correlation analysis calculation.




  
    Hypomethylated    group 
    Global methylation 
  
  
    <μ-s 
    0.57157 
    ELISA LUMA (%) 
  
  
    1/M56 
    0.405
    62.73
  
  
    2/M57 
    0.293
    29.55
  
  
    3/M58 
    0.08
    75.71
  
  
    4/M59 
    0.379
    68.49
  
  
    5/M64 
    0.273
    75.44
  
  
    7/M50 
    0.539
    74.96
  
  
    6/M73 
    0.172
    44.26
  
  
    8/M86
    0.546
    65.65



Table 2: The set of 42 MM samples was divided on the basis of the global DNA
methylation parameter determined by both LUMA and ELISA assays, and the
level of both CDKN2B and PDLIM4 gene expressions. *Values were not included
in the correlation analysis calculation.

Following correlation analysis of the set of 42 MM samples (Table 2) was performed between the global DNA methylation parameter determined by both LUMA and ELISA assays, and the level of both CDKN2B and PDLIM4 gene expressions. In the hypomethylated group of MM patients (LUMA), a strong positive correlation (P< 0,01; r = 0,766) between the level of global DNA methylation and the CDKN2B gene expression was demonstrated (Table 3A). Positive correlation of the PDLIM4 expression (r = 0,587) with the moderately methylated group based on the ELISA assay was at significant P-value < 0,01 (Table 3B).

Table 3: Correlation analysis between global DNA methylation determined by LUMA (A) and/or ELISA (B) and CDKN2B and PDLIM4 gene expressions in the whole set of MM patients.




  
    A 
     
    CDKN2B
      Gene
      Expression 
    PDLIM4
      Gene
      Expression 
  
  
    LUMA (%) 
      Gl.    DNA
      Methylation
    Pearson    (r) p-Value
    0,786    < 0,01
    0,445    < 0,01
  
  
    Hypomethyl.gr.
      Gl.    DNA
      Methylation
    Pearson    (r) p-Value
    0,766    < 0,01
    0,709
      0,0326
  
  
    Moderately    m.
      Gl.    DNA
      Methylation
    Pearson    (r) p-Value
    0,430    0,0223
    0,447� 0,010
  
  
    Hypermethyl.gr.
      Gl.    DNA
      Methylation
    Pearson    (r) p-Value
    0,592
      0,4084
    0,707
      0,1818
  
  
    
      
        B
         
        CDKN2B
          Gene
          Expression
        PDLIM4
          Gene
          Expression
      
      
        ELISA 
          Gl.      DNA
          Methylation
        Pearson      (r) p-Value
        5
        0,556      < 0,01
      
      
        Hypomethyl.gr.
          Gl.      DNA
          Methylation
        Pearson      (r) p-Value
        0,601      0,1151
        0,528      0,1783
      
      
        Moderately      m.
          Gl.      DNA
          Methylation
        Pearson      (r) p-Value
        0,472 0,0129
        0,587 < 0,01
      
      
        Hypermethyl.gr.
          Gl.      DNA
          Methylation
        Pearson      (r) p-Value
        0,776 0,224
        0,643� 0,1685



Table 3: Correlation analysis between global DNA methylation determined by
LUMA
(A) and/or ELISA (B) and CDKN2B and PDLIM4 gene expressions in the whole
set of MM patients.

Global DNA Methylation Versus Gene Methylation

In search for causes of aberrant methylation of CDKN2B gene, a 223-bp promoter region with 13 CpGs, and a 138-bp region of the first exon of the gene with 16 CpGs were analyzed by pyrosequncing. For the assessment of the PDLIM4 gene, the 216-bp region containing 9 CpGs localized in the promoter and 3´-UTR region was used. As described by [17], there are 2 alleles in the population, long and short, with a polymorphic area near a CGG repeat sequence adjacent to the transcription start site, where the long allele is created by an insertion of a 12-bp fragment (5´-CGGCGGCGGCTC- 3´) and substitution of T to G 3 bases upstream of the insertion site in the short allele (Figure 1). Despite the awareness of suspicion, the short allele being methylated more than the long one, we analyzed the 216-bp region where both alleles are shared.

To determine the CDKN2B gene methylation level (MtL), i.e. the promoter and the first exon, we calculated the mean of all measured values, and the median calculated for each sample from all 13 CpG promoter and 16CpG first exon values, for each of the three groups of MM patients: hypomethylated, moderately and hypermethylated (Supplement II). We found significantly higher MtL mean (21,4) of PDLIM4 promoter CpGs methylation in the hypermethylated group than in the hypomethylated group (MtL mean = 17,7; P = 0,026) as well as the moderately methylated group (MtL mean = 15,8; P = 0,023) (Table 7, Table 8, Figure 4).

For correlation analysis between the global DNA methylation and gene expression, the methylation level (mean and median) of both CDKN2B and PDLIM4 gene promoters were calculated.

In the three selected groups of MM patients, we found the methylation level of the CDKN2B promoter in the range 2,8 - 6% (patient mean MtL) and/or 0 - 7% (median). The patient MtL of the PDLIM4 promoter was in the range 13,1 - 27% and/or 12 - 26 % (median). (Supplement I, Supplement II). In the whole set of 42 samples (Table 2), data obtained from both methods for detection of the global DNA methylation status showed positive correlations with the CDKN2B gene promoter methylation and negative correlations with the PDLIM4 gene promoter methylation. We have found very strong negative correlation in the hypermethylated group (P< 0,01; r = -0,994) between the global DNA methylation detected by LUMA assay and the PDLIM4 promoter region methylation, in which median of 9 CpGs was used for calculation (Table 4A, Supplement I). In the hypermethylated group of MM samples examined by ELISA kit (Table 4B), strong negative correlations for average MtL (P = 0,0105; r = -0,915) and for median (P = 0,0162, r = -0,894) were found, these both were significant at P< 0,05. The negative correlations, were detected in the moderately methylated group (LUMA) and PDLIM4 expression (r = -0,414 for average MtL; r = -0,88 for median) (Table 4A, Table 4B). Strong negative correlation between global DNA methylation (LUMA) and PDLIM4 gene promoter methylation were found in the hypomethylated group of MM patients for average MtL (P = 0,035, r = -0,843) and for median of 9 CpGs of the PDLIM4 promoter (P = 0,0157, r = -0,896).

Table 4: Correlation analysis between global DNA methylation determined by LUMA (A) and/or ELISA (B) and CDKN2B and PDLIM4 gene expressions in the whole set of MM patients (Table 2) and different methylated group of MM patients.




  
    A 
     
    CDKN2B
      Gene
      Expression
    PDLIM4
      Gene
      Expression
  
  
    LUMA (%) 
      Gl.    DNA
      Methylation
    Pearson    (r) p-Value
    0,786    < 0,01
    0,445    < 0,01
  
  
    Hypomethyl.gr.
      Gl.    DNA
      Methylation
    Pearson    (r) p-Value
    0,766    < 0,01
    0,709
      0,0326
  
  
    Moderately    m.
      Gl.    DNA
      Methylation
    Pearson    (r) p-Value
    0,430    0,0223
    0,447� 0,010
  
  
    Hypermethyl.gr.
      Gl.    DNA
      Methylation
    Pearson    (r) p-Value
    0,592
      0,4084
    0,707
      0,1818
  
  
    
      
        B
         
        CDKN2B
          Gene
          Expression
        PDLIM4
          Gene
          Expression
      
      
        ELISA 
          Gl.      DNA
          Methylation
        Pearson      (r) p-Value
        0,471      < 0,01
        0,556      < 0,01
      
      
        Hypomethyl.gr.
          Gl.      DNA
          Methylation
        Pearson      (r) p-Value
        0,601      0,1151
        0,528      0,1783
      
      
        Moderately      m.
          Gl.      DNA
          Methylation
        Pearson      (r) p-Value
        0,472�� 0,0129
        0,587      < 0,01
      
      
        Hypermethyl.gr.
          Gl.      DNA
          Methylation
        Pearson      (r) p-Value
        0,776�� 0,224
        0,643� 0,1685



Table 4: Correlation analysis between global DNA methylation determined by
LUMA (A) and/or ELISA (B) and CDKN2B and PDLIM4 gene expressions in
the whole set of MM patients (Table 2) and different methylated group of MM
patients.

Gene expression Versus gene methylation

To identify CDKN2B and PDLIM4 gene promoter methylation to correlate to their gene expression, we performed Pearson correlation analysis in two normalized cohorts according to the method of global DNA methylation detection (Suplement I, Supplement II). We found that both ELISA and LUMA assays detecting global DNA methylation level in the moderately methylated group, show strong negative correlation between PDLIM4 gene expression and PDLIM4 promoter methylation at significant level < 0,01 (Table 5, Table 6). In both groups of MM patients, i.e. with low or high level of global DNA methylation (the hypomethylated group as well as the hypermethylated group), the effect of methylation changes on PDLIM4 expression was not significant. We found no correlation between the CDKN2B gene expression and the CDKN2B promoter methylation, suggesting there is a limited effect of DNA methylation on the CDKN2B expression or, that this gene expression is probably not affected by DNA methylation at all.

Table 5: Correlation analysis between CDKN2B gene expression and CDKN2B promoter methylation (A), PDLIM4 gene expression and PDLIM4 promoter methylation (B) in the whole set of MM patients (Table 2) and different methylated group of MM patients divided on the basis of the global DNA mehtylation detected by LUMA assay).




  
    A
     
    CDKN2B
      Promoter
      Methylation
      (% MtL)
    CDKN2B
      Promoter
      Methylation
      (Median)
  
  
    LUMA (%) CDKN2B
      Gene    expres.
    Pearson    (r) p-Value
    0,276    0,3402
    -0,345� 0,227
  
  
    Hypomethyl.gr.
      CDKN2B
      Gene    expr.
    Pearson    (r) p-Value
    0,451
      0,5494
    -0,398
      0,6019
  
  
    Moderately    m.
      CDKN2B
      Gene    expr.
    Pearson    (r) p-Value
    0,229
      0,621
    -0,426    0,34
  
  
    Hypermethyl.gr.
      CDKN2B
      Gene    expr.
    Pearson    (r) p-Value
    0,806
      0,4032
    -0,5416� 0,6353
  
  
    
      
        B  
         
        PDLIM4 
          Promoter 
          Methylation 
          (% MtL)
        PDLIM4 
          Promoter 
          Methylation 
          (Median)
      
      
        LUMA (%) 
          PDLIM4
          Gene      expr.
        Pearson      (r) p-Value
        -0,612      < 0,01
        -0,659      < 0,01
      
      
        Hypomethyl.gr.
          PDLIM4
          Gene      expr.
        Pearson      (r) p-Value
        -0,605      0,395
        -0,742      0,258
      
      
        Moderately      m.
          PDLIM4
          Gene      expr.
        Pearson      (r) p-Value
        -0,654<      0,01
        -0,686      < 0,01
      
      
        Hypermethyl.gr.
          PDLIM4
          Gene      expr.
        Pearson      (r) p-Value
        -0,561�� 0,3252
        -0,5677      0,3189



Table 5: Correlation analysis between CDKN2B gene expression and CDKN2B
promoter methylation (A), PDLIM4 gene expression and PDLIM4 promoter
methylation (B) in the whole set of MM patients (Table 2) and different methylated
group of MM patients divided on the basis of the global DNA mehtylation detected
by LUMA assay).

Table 6: Correlation analysis between CDKN2B gene expression and CDKN2B promoter methylation (A), PDLIM4 gene expression and PDLIM4 promoter methylation (B) in the whole set of MM patients (Table 2) and different methylated group of MM patients divided on the basis of the global DNA mehtylation detected by ELISA assay.




  
    A 
     
    CDKN2B� 
      Promoter 
      Methylation 
      (% MtL)
    CDKN2B 
      Promoter 
      Methylation 
      (Median)
  
  
    ELISA 
      CDKN2B
      Gene    expr.
    Pearson    (r) p-Value
    0,276    0,3621
    -0,345� 0,227
  
  
    Hypomethyl.gr.
      CDKN2B
      Gene    expr.
    Pearson    (r) p-Value
    N.A.
    N.A.
  
  
    Moderately    m.
      CDKN2B
      Gene    expr.
    Pearson    (r) p-Value
    0,154
      0,692
    -0,432
      0,2456
  
  
    Hypermethyl.gr.
      CDKN2B
      Gene    expr.
    Pearson    (r) p-Value
    0,89
      0,3011
    -0,3871� 0,747
  
  
    
      
        B
         
        PDLIM4
          Promoter
          Methylation
          (% MtL)
        PDLIM4
          Promoter
          Methylation
          (Median)
      
      
        ELISA 
          PDLIM4
          Gene      expr.
        Pearson      (r) p-Value
        -0,612      < 0,01
        -0,659      < 0,01
      
      
        Hypomethyl.gr.
          PDLIM4
          Gene      expr.
        Pearson      (r) p-Value
        -0,543      0,457
        -0,523      0,477
      
      
        Moderately      m.
          PDLIM4
          Gene      expr.
        Pearson      (r) p-Value
        -0,68      < 0,01
        -0,714      < 0,01
      
      
        Hypermethyl.gr.
          PDLIM4
          Gene      expr.
        Pearson      (r) p-Value
        -0,538�� 0,2709
        -0,5964      0,2118



Table 6: Correlation analysis between CDKN2B gene expression and CDKN2B
promoter methylation (A), PDLIM4 gene expression and PDLIM4 promoter
methylation (B) in the whole set of MM patients (Table 2) and different methylated
group of MM patients divided on the basis of the global DNA mehtylation detected
by ELISA assay.

Table 7: Mean of all measured values were obtained from all CpGs of the CDKN2B gene tested regions - promoter (13 CpGs) and first exon (16 CpGs).




  
    CDKN2B (CpGs):
      promoter + first    exon
    Hypermethylated group (mean MtL)
    Moderately methylated group (mean MtL)
    Hypomethylated group (mean MtL)
  
  
    Mean
    3,9
    3,9
    4,7
  
  
    SD
    3,3
    4,3
    3,3
  
  
    Median
    4,0
    4,0
    5,0
  
  
    Minimum
    0,0
    0,0
    0,0
  
  
    Maximum
    16
    20
    16



Table 7: Mean of all measured values were obtained from all CpGs of the CDKN2B gene tested regions - promoter (13 CpGs) and first exon (16 CpGs).

Table 8: Mean of all measured values were obtained from all CpGs of the PDLIM4 gene promoter (9 CpGs).




  
    PDLIM4 (CpGs):
      promoter
    Hypermethylated group (mean MtL)
    Moderately methylated group (mean MtL)
    Hypomethylated group (mean MtL)
  
  
    Mean
    21,4
    15,8
    17,7
  
  
    SD
    10,1
    7,4
    8,9
  
  
    Median
    19,0
    15,0
    15,0
  
  
    Minimum
    8,0
    7,0
    6,0
  
  
    Maximum
    49
    42
    53



Table 8: Mean of all measured values were obtained from all CpGs of the PDLIM4 gene promoter (9 CpGs).

Discussion

In our study, DNA methylation and gene expression analysis from 42 newly diagnosed multiple myeloma patients were performed. The global DNA methylation status was determined by two methods - Luminomentric methylation assay and ELISA colorimetric assay, and for methylation analysis, CDKN2B and PDLIM4 genes were selected for quantification of DNA methylation changes of unsorted cell population in bone marrow aspirates. We found that in newly diagnosed patients, the level of global DNA methylation reflects the CDKN2B gene expression, but not an expression of the PDLIM4 gene, whose expression is negatively affected the PDLIM4 promoter methylation (Figure 3).

CpG islands are genomic regions with a high frequency of CpG sites, which are often located near the transcripton start site of genes. DNA methylation changes, including hypermethylation, in gene promoter CpG islands are responsible for altering the expression of many different tumor suppressor genes and oncogenes in cancer diseases [18,19]. On the other hand, global DNA hypomethylation, which occurs predominantly outside CpG islands, leads to a genomic instability. Moreover, global hypomethylation supports the transition towards an increasingly aggressive phenotype: normal plasma cells / MGUS / asymptomatic myeloma / symptomatic myeloma [7,20], whereas distinct patterns of hypermethylation are presented within sub-groups of symptomatic myeloma [20].

Methylation patterns in cells relevant to the pathogenesis of myeloma plasma cells (Blymphoma cells, normal plasma cells, MGUS and MM cells) are capable of distinguishing nonmalignant and malignant cells [7]. The reason is due to the difference in hypomethylation of global DNA, which is accompanied by genespecific hypermethylation [21]. For this, the PDLIM4 gene, a potential tumor suppressor gene in MM progression, and the CDKN2B gene, representing an inhibitor of cyclin dependent kinases - p15^INK4B and p16^INK4A, were chosen for our analysis. The p15^INK4B and p16^INK4A located on chromosome 9p21 encoded by CDKN2B and CDKN2A genes, belong to the INK4 family proteins. Frequencies of CDKN2B and CDKN2A promoter methylation in multiple myeloma samples are significantly higher in comparison to normal samples [22] as well as frequent methylations of these genes were detected [23-25]. Different prognostic values of p15^INK4B and p16^INK4A methylation in multiple myeloma course suggest a distinct influence on the outcome of MM, and as described Guillerm et al. (2003), survival is impaired in patients with methylatedp16^INK4A, but is not different between patients with methylated and unmethylated CDKN2B gene. Methylation of the CDKN2A gene is significantly associated with short survival [26], while expression of the CDKN2B gene is induced by transforming growth factor - β (TGF-β) [27]. The TGF-β acts as a potent tumor suppressor through regulation of variety of physiological processes, including adhesion of myeloma cells to stromal cells, growth inhibition and stimulation of apoptosis [28-30]. With this being related, the possible inactivation of CDKN2B by methylation could suggest a mechanism of escape from regulatory signals provided by the bone marrow microenvironment [31]. However, the role of p15^INK4B methylation in the development and progression of the multiple myeloma disease remains unknown and other analysis are needed.

In MM patients, interactions between malignant plasma cells derived from multiple myeloma stem cells and bone marrow tumor microenvironment occur directly via cell receptors and adhesion molecules. Some of the adhesion molecules, such as cadherins, facilitate formation of multicellular structures in the bone marrow by anchoring to the actincytosleleton. Our selected RIL (product of PDLIM4 gene) is an actin associated nuclear protein containing both PDZ and LIM domains, which belongs to a large family of LIM proteins [32]. PDZ-LIM proteins, including PDLIM4 may act as adaptors between cytoskeleton and intracellular signaling components, thereby regulating diverse cell signaling pathways [33], In the process of adhesion of myeloma cells to stromal cells, an important role is also played by extracellular factor, such as cytokines and growth factors, which have the ability to activate multiple signalling pathways (such as TGF-β, NF-κβ and Notch), and increase the expression of cell cycle regulatory proteins (D-type cyclins) and Bcl-2 family anti-apoptotic proteins in both stromal and myeloma cells [28-30].Transforming growth factor - β induced growth arrest is attenuated in association with aberrant activation of Nuclear Factor - κβ (NF-κβ) [34], which is in activated cells degraded by PDLIM2 [31]. Among PDZ-LIM proteins, PDLIM1, PDLIM3 and PDLIM4 are most closely related to PDLIM2, since they all contain one PDZ domain and one LIM domain [35], but their role in this signaling pathway has not yet been described. Furthermore, [33] found that PDLIM1 is also highly expressed in dendritic cells, while PDLIM4 is exclusively expressed in CD4+T cells, but not in dendritic cells.

The LIM protein RIL, PDLIM4 gene is located on 5q31.1, a region frequently deleted in the malignant cells of patients with Myelodysplastic Syndrome (MDS) and Acute Myelogenous Leukemia (AML) [36,37], and appears to be a candidate for tumorsuppressor gene resides in this area. Reversion Induced LIM (RIL) frequent in abnormal methylation of a Src suppressor gene PDLIM4 was associated with shortened survival independently of Chronic Myelogenous Leukemia (CML) stage [38]. [39] show that RIL methylation is associated with loss of expression, and that RIL reexpression leads to suppression of cell growth and sensitizes cells to apoptosis. In addition, its methylation status could be a marker of adverse prognosis independent of chromosome 5 and 7 deletions [39,40]. Furthermore, [40] describe a polymorphism in the RIL promoter that creates an Sp1/Sp3 binding site on a 12-bp polymorphic sequence around its transcription start site that creates a long allele. Methylation analysis showed, the short allele had 2,1-3,1-fold higher methylation than the long allele, suggesting that the polymorphic Sp1 site protects against time-dependent silencing.

In our study, we found the significant positive correlation between the groups of both hypomethylated and moderately methylated multiple myeloma patients, and the CDKN2B gene expression. The CDKN2B expression corresponds to a level of the global DNA methylation and is the highest in hypermethylated patients (Figure 1). On the other hand, the level of PDLIM4 expression is the highest in the hypomethylated group of patients (Figure 2). The revealed low PDLIM4 expression in hypermethylated patients seems to be a consequence of the determined increased promoter methylation. This findings indicate a different gene methylation status between p15^INK4B and PDLIM4, which was supported of the strong negative correlation between the PDLIM4 promoter methylation and global DNA hypermethylation. Furthermore, patients with global DNA hypermethylation have got significantly higher MtL mean values of PDLIM4 promoter CpGs than hypomethylated or moderately methylated groups (Table 8). Obviously, the PDLIM4 promoter methylation is present in MM patients and their methylation status may affect a reduction of the PDLIM4 expression. This is supported by the significant correlation between PDLIM4 expression and PDLIM4 promoter methylation.

Our findings show that newly diagnosed hypermethylated patients, as well as the hypomethylated group of patients, show expression variabilites of CDKN2B and PDLIM4 genes in comparison to patients in moderately methylated global DNA methylation state. If we assume that the global DNA hypomethylation is a marker of cancer diseases [21], then the global DNA hypermethylation as the potential initial stage of the myeloma disease progression should be accompanied by slightly changed expression of genes of antitumor defense - including CDKN2B gene. Still not clear whether the CDKN2B gene promoter contains methylation changes leading to the gene silencing. There is a lack of systematic analysis of p15^INK4B epigenetic modification such as promoter or body gene methylations. Although a number of reports have determined that the inactivation of CDKN2B gene is mainly induced by hypermethylation in MM, the reported rates of CDKN2B hypermethylation in MM are remarkably diverse. As described [42] in the set of 72 MM patients, in both tested regions of the CDKN2B gene, i.e. the promoter and the first exon, there were no significant differences between the methylation levels - mean MtL ranged from 3% to 6% in the CDKN2B gene promoter and from 3% to 12% in the first exon. Therefore, detailed pyrosequencing methylation analysis of promoter and first exon the CDKN2B gene regions were performed, but for our comparison analysis we used methylation data from the promoter region of both studied genes.

Although we present a pilot study and data, which should be further extended and complemented by comparison with clinical data of tested patients, in our analyzed the set of newly diagnosed multiple myeloma patients, a level of the global DNA methylation reflects the CDKN2B gene expression, while the PDLIM4 gene expression is negatively affected by the PDLIM4 promoter methylation. Our results indicate that increased methylation in PDLIM4 promoter has an effect on the PDLIM4 gene decreased expression as is generally described in cancer proliferations [18,19]. Moreover, we have found no correlations between both used methods - Luminomentric methylation assay and ELISA colorimetric assay detecting the global DNA hypermethylation, and the CDKN2B gene expression, suggesting a possibility that this gene expression is not affected by DNA methylation. The CDKN2B gene hypomethylated status corresponding to the CDKN2B gene low expression could be associated with impaired prognosis in newly diagnosed hypermethylated MM patients rather than in patients with hypomethylated and moderatetely methylated global DNA methylation states. Finally, our findings indicate that both global DNA hypermethylation and hypomethylation states of newly diagnosed patients could be related to a followed progression of the multiple myeloma disease.

Acknowledgement

We thank the Hemato-Oncology Department of the University Hospital Olomouc to providing samples of selected multiple myeloma patients. We thank Dr. Karin Jöhrer from Tyrolean Cancer Research Institute in Innsbruck for reading the manuscript.

Funding

This study was supported in part by NV18-03-0050 from the Ministry of Health of the Czech Republic and LF_2018_001 from Palacky University Olomouc.

References

Download PDF

Citation: Luzna P, Flodrova P, Janovska L, Zapletalova J, Minarik J, Kolar Z, et al. Different Gene Methylation Status of the CDKN2B and/or PDLIM4 as the Result of Comparative Analysis to the Global DNA Methylation in Unsorted Cell Population of Multiple Myeloma Patients. Ann Hematol Oncol. 2019; 6(7): 1257.

Instruction for Authors

Submit Your Article