Introduction
Both genetic and epigenetic aberrations play vital roles in tumorigenesis and the development of cancer. DNA methylation is an important epigenetic modification, and aberrant changes in DNA methylation are prominent features in the pathogenesis and development of various diseases, including cancer. Aberrant methylation within the promoter regions of tumor-suppressor genes has been correlated with their silencing and with uncontrolled proliferation of tumor cells (1,2).
There are four types of inhibitor of DNA binding (ID) proteins in mammals, belonging to the helix-ring-helix protein family, and their coding regions are located on chromosome 6p21-22. Animal experiments and in vitro cellular assays have shown that the ID4 protein inhibits the proliferation of tumor cells and promotes cell apoptosis (3). In addition, aberrant promoter methylation of the ID4 gene has been correlated with the pathogenesis and development of hematological malignancies (4). The tight junction protein 1 (ZO-1) gene belongs to the membrane guanylate kinase family, and it encodes a tight junction protein involved in many signaling pathways that participate in the regulation of cell proliferation and differentiation (5,6). The aberrant hypermethylation of the promoter region of ZO-1 was reported in murine leukemia cell lines, and the concomitant suppression of the expression of the ZO-1 gene was also observed. The ZO-1 gene may be a novel oncogene in hematological malignancies that was identified using restriction landmark genomic scanning (RLGS) (7). In summary, the alteration of the methylation pattern on the promoter regions of the ID4 and ZO-1 genes is a distinctive feature of hematological malignancies when compared to normal cells (8).
Lymphoma is a hematological malignancy which is diagnosed based mainly on pathological examination. However, the diagnosis is not easy, partially because the accuracy of the pathological examination relies on the experience of the clinical practitioner and is dependent on subjective impressions, particularly when the malignant atypical features of the tumor cells are not apparent, or when the distinction between benign and malignant hyperplasia is not clear (9,10). Although auxiliary examinations are currently available, such as immunohistochemical tests, gene rearrangement analysis, and T cell receptor (TCR) detection, the diversified pathologic morphologies of lymphoma tissues often result in difficulties with diagnosis. Therefore, there is a need for more sensitive and specific molecular diagnostic methods for the diagnosis, treatment, and prognostic prediction of lymphoma. The highly sensitive methylation-specific PCR (MSP) method has been previously applied to detect methylation of the promoter regions of the ID4 and ZO-1 genes in the diagnosis of breast cancer (11). One objective of the current study was to identify the promoter methylation status of the ID4 and ZO-1 genes in the lymph node and bone marrow of lymphoma patients.
The MSP (12) technique was utilized in the present study to detect aberrant alterations in the methylation status of the ID4 and ZO-1 genes and to evaluate the feasibility of utilizing these changes as epigenetic markers for lymphoma diagnosis, or for detection of minimal residual lymphoma after treatment. A comparison of the methylation statuses of ID4 and ZO-1 was performed between lymphoma patients and a control cohort comprised of patients without lymphoma. Meanwhile, the feasibility of using the ID4 and ZO-1 methylation status as a molecular indicator for lymphoma diagnosis and prognosis was also evaluated.
Materials and methods
Patients and tissue samples
Paraffin-embedded lymphoma tissue samples were obtained from 92 lymphoma patients treated at the PLA General Hospital (Beijing, China) between May 2006 and October 2009, in whom the diagnosis was confirmed by pathologic morphological analysis and immunohistochemical studies. Paraffin-embedded lymph node samples were also obtained from 10 patients with either chronic cholecystitis or reactive lymphadenitis, as the control set. The bone marrow specimens were obtained before treatment from 90 lymphoma patients admitted to our hospital between February 2008 and February 2010, and their clinical information is summarized in Table I. The therapeutic strategy was determined according to National Comprehensive Cancer Network (NCCN) lymphoma treatment guidelines, or hematopoietic stem cell transplantation was performed if possible. The control samples of bone marrow tissue were from 8 donors who had donated bone marrow for hematopoietic stem cell transplantation. Written informed consent was provided by the patients and their families. The deadline for the last follow-up for all enrolled patients was October 31, 2011.
The HL-60 and K562 cell lines were used as positive and negative controls, respectively, for MSP of the ID4 and ZO-1 genes, and the cells were maintained in the Department of Hematology, PLA General Hospital. Deionized water was used as the blank control for MSP.
PCR primers
The methylation-specific primers for the ID4 gene were ID4-MF forward, TCGGAGTTTTCGTTTTCGTT and ID4-MF reverse, CGATACTACTCACAACCGCG; ID4-UMF forward, TGTTTGGAGTTTTTGTTTTTGTT and ID4-UMR reverse, CCCAATACTACTCACAACCACAC. The methylation-specific primers for ZO-1 were ZO-1-MF forward, AAAATAAATAGGAAGATTCGTACGG and ZO-1-MR reverse, GAAACTAAACGAAACGAAACGAA; ZO-1-UMF forward, GGATAAAAATAAATAGGAAGATT TGTATG and ZO-1-UMR reverse, AACAAAACTAAACA AAACAAAACAA.
Genomic DNA extraction and base modification
DNA extraction from paraffin-embedded tissues was performed using a commercial kit (Qiagen, USA), and the genomic DNA Purification kit (Promega, USA) was utilized for DNA extraction from the bone marrow. The concentration of extracted DNA was determined using spectrophotometry. Two micrograms of DNA was sheared by passing through 26G syringe needles repeatedly after dilution with water. Then, 5 μl freshly prepared NaOH solution (3 mol/l) was added to the sheared DNA, and the mixture was incubated at 37°C for 25 min, followed by the addition of 30 μl freshly prepared hydroquinone solution (10 mmol/l) and 520 μl sodium bisulfite (3 mol/l) and subsequent incubation at 55°C for 16 h. The DNA was purified from the above mixture using a DNA gel extraction kit (Axygen Biosciences, USA) and stored at −20°C.
MSP reactions
The total volume of the reaction was 25 μl, and the mixture consisted of 4 μg DNA, 2.5 μl 10X Herman’s buffer, 1.25 μl of 25 mM dNTPs, 0.15 μl Hot Start Taq polymerase, and 3 μl of each upstream and downstream methylation-specific or unmethylated DNA-specific PCR primers. The PCR reaction settings were as follows: pre-denaturation at 94°C for 10 min, followed by 35 cycles of 94°C for 60 sec, 56°C for 40 sec, and 72°C for 50 sec, and a final elongation for 10 min at 72°C. The HL-60 and K-562 cells were used as positive and negative controls, respectively, for each trial, and deionized water was used as the blank control. The products of PCR were separated with 2% agarose DNA electrophoresis, the resolved DNA was stained with ethidium bromide and images were recorded with an ultraviolet transmission analyzer.
The PCR products were recovered and purified from the agarose gel with the DNA gel extraction kit (Qiagen).
Plasmid construction and DNA sequencing
The PCR fragments recovered from the agarose gel were inserted into the pGEM-T Easy vector by T4 DNA ligase-based cloning, and subsequently transformed into competent E. coli DH5α. Positive colonies were obtained by antibiotic resistance screening and blue-white colony screening. One microliter of a bacterial culture grown overnight was used as the template for PCR. The PCR-positive colonies were sequenced by Invitrogen (Beijing, China).
Readout
For the ID4 gene, complete methylation was indicated by the presence of a 186-bp PCR fragment alone upon amplification with the methylation-specific primers; completely unmethylated DNA was indicated by the presence of a 191-bp PCR fragment alone upon amplification with the unmethylated DNA-specific primers. Partial methylation was indicated by the presence of both the 186- and 191-bp PCR fragments after amplification. Similarly, for the ZO-1 gene, the presence of a 177-bp or a 179-bp PCR fragment alone indicated complete methylation or a complete lack of methylation, respectively, while the presence of both 177- and 179-bp fragments after amplification was indicative of partial methylation. Since the bone marrow of the patients consisted of both lymphoma cells and normal cells, partial methylation and complete methylation were classified as a methylation-positive readout.
Statistical analysis
Quatitative data are expressed as means ± SD, and differences between data sets were analyzed using the Student’s t-test. Categorical data were described as absolute frequencies and analyzed by the Pearson’s Chi-square test. Cumulative event curves were plotted using the Kaplan-Meier survival method and differences between curves were tested for statistical significance using the log-rank method. P-values <0.05 were considered statistically significant. All data were analyzed using the SPSS version 12.0 statistical software (SPSS Inc., Chicago, IL, USA).
Results
Methylation status of genes amplified from the paraffin-embedded lymphoma and bone marrow tissues
The PCR results showed that 64 out of 92 paraffin-embedded lymphoma tissues were ID4-positive and 57 were ZO-1-positive, respectively, either in the methylated or unmethylated state, and that 51 were both ID4- and ZO-1-positive. Eight out of 10 paraffin-embedded lymphoma tissues from either chronic cholecystitis or reactive lymphadenitis patients were both positive for completely unmethylated ID4 and ZO-1 genes. All bone marrow tissues of lymphoma patients were ID4- and ZO-1-positive, for either methylated or unmethylated DNA.
There were 41 and 43 samples positive for ID4 and ZO-1 gene methylation out of the 51 paraffin-embedded lymphoma tissues, and the positivity rates were 80.4 and 84.3%, respectively. Forty-seven samples were positive for either ID4 or ZO-1 methylation, and the coverage rate for both genes was 92.2%. The ID4 and ZO-1 genes were completely unmethylated in the 8 paraffin-embedded lymph node tissues from either chronic cholecystitis or reactive lymphadenitis patients.
There were 23 and 13 samples positive for ID4 and ZO-1 gene methylation, respectively, out of 90 bone marrow samples from lymphoma patients, and the positivity rates were 25.56 and 14.44%, respectively. The ID4 and ZO-1 genes of the 8 samples from bone marrow donors were completely unmethylated (Tables II and III, Fig. 1).
Survival curve
The patients with positive MSP results for ID4 or ZO-1 isolated from the bone marrow had a lower survival rate compared to the patients with unmethylated ID4 or ZO-1 genes. A significant difference was observed in the survival rate between the patients with and without ID4 methylation in the bone marrow (P<0.05). On the contrary, the methylation status of ID4 and ZO-1 of DNA isolated from the lymphoma tissue did not show any correlation with the survival rate (Fig. 2).
DNA sequencing results
The CpG methylation sites of individual bacterial clones are shown in Fig. 3. The total methylation rates of ID4 and ZO-1 from 10 randomly picked colonies were 22.08 and 28.4%, respectively (Fig. 3). On the contrary, the total methylation rates of ID4 and ZO-1 were 5.0 and 5.1%, respectively, in the paraffin-embedded tissues of subjects without lymphoma, indicating that these genes are overwhelmingly hypomethylated in subjects without lymphoma. The total methylation rates of ID4 and ZO-1 of 10 colonies from the paraffin-embedded tissues of lymphoma patients were 69.79 and 95.0%, respectively, which were significantly higher than the rates in the control subjects without lymphoma.
Discussion
Genetic and epigenetic aberrations are the two major mechanisms of tumorigenesis. The epigenetic changes are mainly reflected by alterations in the DNA methylation patterns. For instance, the hypermethylation of CpG islands in the promoter region of tumor-suppressor genes and the hypomethylation status of the entire genome have both been found in tumors (13). Usually, one or several tumor-suppressor genes are thought to be responsible for tumorigenesis in certain types of malignancies, and the epigenetic features of the promoter regions of each tumor-suppressor gene are unique (14–17). The pathogenesis and development of lymphoma are closely associated with the methylation status of various tumor-suppressor genes (18–20).
Successful methylation specific PCR was performed in 51 paraffin-embedded lymphoma samples out of the total 92, and 41 and 43 samples were positive for ID4 and ZO-1 methylation, respectively, with a positivity rate of 80.4 and 84.3%, respectively. Forty-seven samples were positive for either ID4 or ZO-1 gene methylation, and the coverage rate for both genes was 92.2%. Importantly, the ID4 and ZO-1 genes of 8 paraffin-embedded lymphoma tissues from either chronic cholecystitis or reactive lymphadenitis patients were completely unmethylated. Therefore, the methylation of ID4 and ZO-1 promoters in lymphoma patients may be an epigenetic feature of lymphoma, and the combined detection of the methylation of ID4 and ZO-1 may be a potential highly specific, highly sensitive, auxiliary diagnostic approach for lymphoma.
There are a few factors that may have influenced the detection rate of MSP in the present study. First of all, the DNA isolated from the paraffin-embedded samples may hve been degraded or damaged because of prior chemical and physical treatments, such as formalin and temperature changes (21,22). Second, the amount of DNA isolated from the paraffin-embedded samples was relatively low, and the final concentration was only 50 μg/μl. The quality of DNA was also problematic, since the ratio of absorption at 260 and 280 nm was usually above 1.9, indicating contamination with RNA or protein. Moreover, the severe degradation of DNA along with the prolonged storage of paraffin-embedded samples may be another reason for the low detection rate of DNA methylation.
Of the 90 bone marrow samples from the lymphoma patients, 23 and 13 samples were positive for ID4 and ZO-1 gene methylation, respectively, and the positivity rate was 25.56 and 14.44%, respectively. Interestingly, the ID4 and ZO-1 genes of the 8 bone marrow donors were completely unmethylated. The patients with positive ID4 or ZO-1 gene methylation in bone marrow had lower survival rates compared to the patients with unmethylated ID4 or ZO-1 genes. In addition, a significant difference was observed in the survival rate between the patients with and without ID4 gene methylation in the bone marrow, indicating that the methylation status of the ID4 gene in the bone marrow may be associated with a more favorable prognosis in lymphoma patients. Further studies in a larger cohort of patients are needed to evaluate the generalizability and clinical utility of this observation.
In the present study, we found that the MSP-derived methylation detection rate for the ID4 and ZO-1 genes was higher in samples from paraffin-embedded lymphoma tissues than in those from fresh bone marrow. This may be due to the large amount of lymphoma cells in the paraffin-embedded tissues, and may be also due to the fact that the bone marrow is not involved during the early stages of lymphoma. We also found in the present study that the methylation status of ID4 and ZO-1 in non-Hodgkin lymphoma patients was higher than that in Hodgkin lymphoma patients, in DNA samples derived either from the paraffin-embedded tissue or from fresh bone marrow (Tables II and III). Meanwhile, the stages, the cellular type of lymphoma, and the International Prognostic Index were not correlated with the methylation status of the ID4 and ZO-1 genes (Tables II and III). Moreover, the survival rate of the patients with methylated ID4 or ZO-1 in samples derived from the paraffin-embedded tissues was not significantly different from that of patients with unmethylated ID4 and ZO-1 genes, indicating that ID4 and ZO-1 methylation in the bone marrow is of particular significance in lymphoma.
In order to analyze the CpG island methylation status in the promoter regions of the ID4 and ZO-1 genes, bisulfite sequencing was performed for individual clones of ID4 and ZO-1 amplified from the bone marrow of healthy individuals and the lymphoma patients without prior treatment. Similar to the MSP results, the bisulfite sequencing results showed that the total methylation rates of ID4 and ZO-1 were 22.08 and 28.4%, respectively. On the contrary, the total methylation rates of ID4 and ZO-1 were 5.0 and 5.1%, respectively, in paraffin-embedded tissues of subjects without lymphoma, indicating that these genes are predominantly in a hypomethylated state in healthy subjects. The total methylation rates of ID4 and ZO-1 genes of 10 colonies derived from the paraffin-embedded tissue samples of lymphoma patients were 69.79 and 95.0%, respectively, which were significantly higher than the rates in the control subjects without lymphoma. The CpG island methylation status in the promoter regions of the ID4 and ZO-1 genes was also low in the DNA isolated from bone marrow of the subjects without lymphoma. Overall, the methylation patterns of the ZO-1 and ID4 genes were altered in the lymphoma patients, indicating that the aberration of methylation patterns of the ZO-1 and ID4 genes may be related to the pathogenesis and development of lymphoma.
In the present study, we also showed that the methylation rates of the ID4 and ZO-1 genes were 80.4 and 84.3%, respectively, in samples derived from 51 paraffin-embedded lymphoma tissues. Control samples only contained completely unmethylated ID4 and ZO-1 DNA, indicating that the methylation of ID4 and ZO-1 is highly specific to patients with lymphoma diagnosis. Therefore, the methylation pattern of the ID4 and ZO-1 genes may be a potential biomarker of lymphoma (5,22), which could be used in the early diagnosis of lymphoma. Moreover, patients with a positive methylation status of ID4 in the bone marrow had significantly lower survival rates than patients whose specimens lacked ID4 methylation. These data suggest that the aberrant methylation status of the ID4 gene isolated from the bone marrow may be one indicator of poor prognosis in lymphoma. Further studies will help establish the clinical value of the methylation status of the ID4 and ZO-1 genes as diagnostic and prognostic markers in lymphoma.