Expression of O⁶-methylguanine-DNA methyltransferase causes lomustine resistance in canine lymphoma cells

Abstract

The DNA repair protein O⁶-methylguanine-DNA methyltransferase (MGMT) causes resistance to nitrosoureas in various human cancers. In this study, we analyzed the correlation between canine lymphomas and MGMT in vitro. Two of five canine lymphoma cell lines required higher concentrations of lomustine to inhibit cell growth by 50%, but their sensitivity to the drug increased when they were cultured with an MGMT inhibitor. Fluorometric oligonucleotide assay and real-time polymerase chain reaction of these cell lines revealed MGMT activity and high MGMT mRNA expression, respectively. We analyzed the methylation status of the CpG islands of the canine MGMT gene by the bisulfite-sequencing method. Unlike human cells, the canine lymphoma cell lines did not show significant correlation between methylation status and MGMT suppression levels. Our results suggest that in canine lymphoma MGMT activity may influence sensitivity to nitrosoureas; thus, inhibition of MGMT activity would benefit nitrosourea-resistant patients. Additional studies are necessary to elucidate the mechanism of regulation of MGMT expression.

Introduction

Lymphoma, the most common hematopoietic malignant disease in dogs, is usually treated with the combination of cyclophosphamide, doxorubicin, vincristine, and prednisone (CHOP) as first-line therapy (1,2). Although the rate and duration of remission with this protocol are more than 80% and more than 9 mo, respectively, chemotherapeutic resistance develops eventually in most animal patients, one cause being increased expression of the gene encoding P-glycoprotein (3,4).

Lomustine [1-(2-chloroethyl)-3-cyclohexyl-1-nitrosourea], also called CCNU, is an alkylating nitrosourea compound with a molecular weight of 233.7 Da that cannot act as a substrate for P-glycoprotein. Lomustine induces alkylation and cross-linking of DNA at the O⁶-position of guanine and thus inhibits DNA synthesis (5). Although lomustine is an effective rescue agent for relapsed canine lymphoma (6,7), the remission often lasts for less than 100 d, which indicates that the tumor cells may acquire resistance to lomustine.

The DNA repair protein O⁶-methylguanine-DNA methyltransferase (MGMT), which is expressed in human tumor cells, plays a significant role in the development of resistance to nitrosoureas like lomustine and carmustine (BCNU) (8). This protein reverses the formation of adducts at the O⁶-position of guanine by transferring the alkyl adduct to a cysteine residue in itself (9), thereby averting the formation of lethal DNA cross-links. Absence of MGMT activity has been detected in approximately 30% of brain tumors (10,11) and may be associated with enhanced sensitivity to the action of alkylating agents (12–14). The inactivation of MGMT is most frequently caused by epigenetic changes, specifically alterations in the methylation status of the promoter region. Methylation of the MGMT promoter region, which suppresses the expression of MGMT messenger RNA (mRNA), is a strong predictor of the cell response to carmustine and overall survival in humans with brain tumors (15). Moreover, hypermethylation of the MGMT promoter is associated with a favorable prognosis in humans with diffuse large B-cell lymphoma treated with multidrug regimens, including the alkylating agent cyclophosphamide (16). However, the role of MGMT expression in canine lymphoma remains understudied.

To determine MGMT expression in canine lymphoma cells, we conducted a cytotoxicity assay with lomustine and an MGMT inhibitor, O⁶-benzylguanine (BG), on canine lymphoma cell lines, detecting MGMT activity by a fluorometric oligonucleotide assay. We also examined hypermethylation of the MGMT promoter region by a bisulfite-sequencing method.

Materials and methods

Cells and chemicals

Five canine lymphoma cell lines, GL-1 (B-cell type) (17), CL-1, Ema, Nody-1, and UL-1 (T-cell type) (18–20) as well as immortalized human T-lymphocytes (Jurkat cells) were used. All the cell lines were maintained in complete medium [RPMI-1640 containing 10% fetal bovine serum (FBS), penicillin (100 U/mL), streptomycin (100 μg/mL), and 2-mercaptoethanol (55 μM)] and grown at 37°C in an atmosphere containing 5% CO₂. Peripheral blood mononuclear cells (PBMCs) from 3 healthy dogs were used as the control cells. Briefly, heparinized whole blood was centrifuged and the buffy coat suspended in phosphate-buffered saline (PBS: 2.7 mM KCl, 0.14 M NaCl, 1.5 mM KH₂PO₄, and 8.1 mM Na₂HPO₄·12H₂O). The PBMCs were isolated by gradient centrifugation with the use of Lymphoprep (Fresenius Kabi Norge, Oslo, Norway) and further purified by being overlaid on FBS and then centrifuged to remove the platelets.

Lomustine (Bristol-Myers Squibb, New York, New York, USA) and BG (Sigma-Aldrich, St. Louis, Missouri, USA) were dissolved in 200-proof ethanol at concentrations of 100 and 16.7 mM, respectively, and stored at −20°C until required. In addition, 5-aza-2′-deoxycytidine (Sigma-Aldrich), which causes DNA demethylation, was dissolved in distilled water at a concentration of 43.8 mM, diluted to 20 μM in the complete medium, and stored at −20°C until required.

Cytotoxicity assay

Treatment conditions for BG and lomustine were determined from previously described protocols (21,22). The BG-treated (BG+) and BG-untreated (BG−) groups were pretreated with BG (final concentration 80 μM) and ethanol (equal amount), respectively, for 2 h at 37°C in 5% CO₂. The canine lymphoma cells were washed with PBS and incubated for 1 h with various concentrations of lomustine. The cells were washed with PBS, and the medium was replaced with fresh medium without the drugs. The cells were then grown for 5 d, and an MTT-based assay was done with use of the Cell Growth Determination Kit (Sigma-Aldrich). Each experiment was done in triplicate and independently repeated 3 times. The concentrations of lomustine that inhibited cell growth by 50% (IC₅₀) were calculated from drug survival curves.

Fluorometric oligonucleotide assay

The MGMT activity was measured by a fluorometric oligonucleotide assay described by Kreklau et al (23) with minor modifications. Briefly, exponentially growing cultured cells or PBMCs were suspended in an assay buffer (50 mM Tris, pH 8.0, 1 mM dithiothreitol, 1 mM EDTA, and 5% glycerol), pulse-sonicated in ice-cold water 25 times (for 10 s each time) by means of a BioRuptor UCD-200T (Thermo Fisher Scientific, Waltham, Massachusetts, USA), and then centrifuged at 14 000 × g at 4°C for 30 min. The protein concentration of the supernatants was quantified with the Bradford protein assay (Thermo Fisher Scientific). The protein samples were snap-frozen in liquid nitrogen and stored at −80°C until analyzed.

To obtain a double-stranded oligonucleotide of 45 base pairs (bp) containing a single O⁶-methylated guanine residue nested in a PvuII restriction site and a fluorometric 5′-hexachloro-fluorescein phosphoramidite (HEX), we purchased 2 custom oligonucleotides from Sigma-Aldrich: Sense 45 (5′-GCAGTCCAGCTTCAG^mCTGCACGTCATCCTGTGCAGTCGTCTCGAC-3′) and Antisense 45 (5′-HEX-GTCGAGACGACTGCACAGGATGACGTGCAGCTGAAGCTGGACTGC-3′). The oligonucleotides were diluted to 2 μM with TE Buffer (50 mM Tris, pH 8.0, and 1 mM EDTA), mixed in a 1:1 molar ratio, denatured at 95°C for 10 min, and annealed at room temperature for 30 min. The resulting double-stranded oligonucleotide (25 nmol) was incubated with the extracted proteins (600 μg) at 37°C for 2 h, purified by phenol–chloroform extraction and ethanol precipitation, digested with PvuII, and subjected to electrophoresis on a 20% denaturing polyacrylamide gel. The HEX-labeled oligo was detected with a green light-emitting diode light (excitation wavelength 530 nm; MeCan Imaging, Saitama, Japan) and a digital camera with FUJI ultraviolet-absorbing filter SC54 (Fujifilm, Tokyo, Japan), and the fluorescence intensity was analyzed with Multi Gauge version 3.0 (Fujifilm).

Treatment with 5-aza-2′-deoxycytidine

The conditions of 5-aza-2′-deoxycytidine treatment (incubation time and concentration) were determined from previously described protocols (24,25). The 5 lymphoma cell lines and PBMCs were treated with 5-aza-2′-deoxycytidine (final concentration 0.5 μM) for 3 d at 37°C in 5% CO₂. The culture medium was replaced with fresh medium containing 5-aza-2′-deoxycytidine every day. After 3 d of culture the cells were washed with PBS and stored at −80°C until required.

Real-time polymerase chain reaction (PCR)

Total RNA was isolated from the 5 lymphoma cell lines and PBMCs with the use of TRI reagent (Molecular Research Center, Cincinnati, Ohio, USA), according to the manufacturer’s instructions, and resolved with 30 μL of water treated with diethylpyrocarbonate. Samples of 2 μg were treated with DNase I by means of the TURBO DNA-free kit (Applied Biosystems, Carlsbad, California, USA) and transcribed into cDNA with Superscript II (Invitrogen Life Technologies, Carlsbad, California, USA) according to the manufacturers’ instructions. An oligo-dT primer was used to prime the first-strand synthesis for each reaction. The single-stranded cDNA was subjected to real-time PCR amplification with a QuantiTect SYBR Green PCR Kit (Qiagen, Valencia, California, USA) according to the manufacturer’s protocol. The primers used for assaying the expression of MGMT and ribosomal protein L32 (RPL32), from a housekeeping gene (26), were designed on the basis of canine nucleotide sequences (Table I). Each assay was done in duplicate. Predenaturation at 95°C for 15 min was followed by 45 cycles of PCR amplification consisting of denaturation at 94°C for 15 s, annealing at 58°C for 30 s, and extension at 72°C for 30 s. The StepOne PCR system (PerkinElmer, Waltham, Massachusetts, USA) was used for both the PCR and fluorescence intensity detection, and the data were analyzed with StepOne software, version 2.0. Briefly, the PCR cycle number at the threshold was represented as Ct, and the difference between Cts for the target and internal control (ΔCt) was calculated.

Table I.

Primers used in this study

Target	Forward primer	Reverse primer	Ta	Use
Canine MGMT	5′-TGCTCGGGAGGATGGACAAG-3′	5′-AGTCACTCAGATGTTTACTCCC-3′		RT-PCR
MGMT region 1	5′-GGGGATAGATTTTGGAAAATG-3′	5′-CAAAATCAACCCACTACTCCATC-3′	55	BS
MGMT region 2-1	5′-AGGGAAGTGGAGGTAGTAGG-3′	5′-CCCCTAACCCCTAATCCTAT-3′	53	BS
MGMT region 2-2	5′-ATAGGATTAGGGGTTAGGGG-3′	5′-ACCAACTTCCTTCACCAACAAA-3′	54	BS
MGMT region 3-1	5′-GTTTTTGTTGAGGAGGAGTA-3′	5′-CCTACTACCACCACAATAAC-3′	50	BS
MGMT region 3-2	5′-GGTTATTGTGGTGGTAGTAGG-3′	5′-CCCAAAACAACAATCTTACC-3′	52	BS
MGMT region 4	5′-GGGTATTGGGGGAAGGTGTT-3′	5′-CCAAAACCATCTCTCCAATCCC-3′	61	BS
Canine MGMT	5′-CCTGGCTGGATGCCTATTTCC-3′	5′-CTGCTGGTAGGAAACCGTGT-3′		RT-PCR
Canine RPL32	5′-TGGTTACAGGAGCAACAAGAAA-3′	5′-GCACATCAGCAGCACTTCA-3′		RT-PCR
Canine PRL13A	5′-GCCGGAAGGTTGTAGTCGT-3′	5′-GGAGGAAGGCCAGGTAATTC-3′		RT-PCR

Ta — annealing temperature (°C); MGMT — O⁶-methylguanine-DNA methyltransferase; RT-PCR — real-time polymerase chain reaction; BS — bisulfite sequencing; RP — ribosomal protein.

Bisulfite sequencing

Total DNA was isolated from the 5 lymphoma cell lines and PBMCs with the QIAamp DNA Mini Kit (Qiagen), and 5 μg was digested with BamHI for 1 h. After phenol–chloroform extraction and ethanol precipitation, 1 μg of the DNA sample was subjected to bisulfite modification with the CpGenome DNA Modification Kit (Chemicon, Temecula, California, USA) according to the manufacturer’s instructions and then stored at −80°C until required. The presence of CpG islands [regions of unmethylated DNA with a high frequency of dinucleotides of C (cytosine) followed immediately by G (guanine), or C plus G (CpG)] around exon 1 of the canine MGMT gene, in which methylation is associated with the silencing of MGMT expression in human tumor cells (27), was determined by means of Methyl Primer Express software, version 1.0 (http://www.appliedbiosystems.com/absite/us/en/home/support/software-community/free-ab-software.html), with the following parameters: a CpG content greater than 55% and a ratio of observed CpG to expected CpG greater than 0.65 (28). From the results, 6 sets of PCR primers were prepared (Table I), and PCR amplification was done as follows: predenaturation at 95°C for 15 min, followed by 40 cycles of denaturation at 94°C for 30 s, annealing at 50°C to 61°C (Table I) for 30 s, and extension at 72°C for 60 s, and then a final extension at 72°C for 7 min, with the bisulfite-modified DNA as the template. For direct sequencing, the gel-purified PCR product was sequenced with the same primers and a BigDye Terminator, version 3.1, Cycle Sequencing Kit (PerkinElmer) and analyzed with an ABI377 automated DNA sequencer at the Yamaguchi University Center for Gene Research. Alternatively, the amplified DNA was cloned into a pCR2.1 plasmid vector by means of the TOPO TA Cloning Kit (Invitrogen), and 5 independent clones that were obtained were subjected to sequence analysis (direct sequencing).

Statistical analyses

The cytotoxicity of lomustine with and without BG was analyzed by 2-way analysis of variance and Bonferroni post-hoc tests. The association between the IC₅₀ of lomustine and the real-time PCR values for the mRNA expression of MGMT was assessed by calculating Spearman’s rank correlation coefficient. Analyses were done with GraphPad Prism 5 software (GraphPad Software, La Jolla, California, USA). A P-value of 0.05 or less was considered to be statistically significant.

Results

To examine cell susceptibility to lomustine with and without BG, the cytotoxicity assay was conducted on 5 canine lymphoma cell lines as well as Jurkat cells, which are known to have MGMT activity (29). Figure 1 and Table II show the rates of cell survival at various concentrations of lomustine and the mean IC₅₀ values, respectively. All the cell lines showed a dose-dependent decrease in cell survival. Whereas pretreatment with BG resulted in a significant decrease in the viability of the Jurkat, CL-1, and Nody-1 cells, no difference in viability with or without BG was observed in the Ema, GL-1, and UL-1 cell lines. Moreover, although the IC₅₀ values were decreased by BG pretreatment, they were relatively high for the Jurkat, CL-1, and Nody-1 cells compared with the Ema, GL-1, and UL-1 cells.

Figure 1.

Mean percent viability (± standard deviation) of lymphoma cell lines pretreated with (dotted lines) or without (solid lines) O⁶-benzylguanine (BG), an inhibitor of the DNA repair protein O⁶-methylguanine-DNA methyltransferase (MGMT), and then incubated with various concentrations of lomustine. The experiment was done in triplicate, and the data are representative of 3 independent experiments.

Table II.

Mean concentrations of lomustine required to inhibit cell growth by 50% in each cell line when not pretreated with O⁶-benzylguanine (BG−) and when pretreated (BG+)

	Mean concentration (± standard deviation), μM
	Jurkat	CL-1	Ema	GL-1	Nody-1	UL-1
BG (−)	45.0 (± 5.1)	> 50.0	14.0 (± 1.7)	17.0 (± 0.7)	36.5 (± 3.5)	23.5 (± 0.6)
BG (+)	18.5 (± 0.9)	11.5 (± 4.0)	13.5 (± 2.3)	16.0 (± 0.5)	7.5 (± 4.7)	19.0 (± 0.7)

The difference in MGMT activity between the cell lines was examined by the fluorometric oligonucleotide assay (23), in which a HEX-labeled oligonucleotide was incubated with cellular extracts from the 5 canine lymphoma cell lines and Jurkat cells. If the extract contained MGMT activity, an O⁶-methylated guanine residue nested within a PvuII restriction site could be demethylated, and the oligonucleotide could be digested by PvuII. Figure 2 represents the assay results and shows a PvuII-digested band in the CL-1 and Nody-1 cells as well as the Jurkat cells, indicating that these cells contain MGMT activity.

Figure 2.

Results of fluorometric oligonucleotide assay for MGMT activity. A double-stranded oligonucleotide of 45 base pairs (bp) containing a single O⁶-methylated guanine residue nested in a PvuII restriction site and a fluorometric 5′-hexachloro-fluorescein phosphoramidite were incubated with extracts from the 5 canine lymphoma cell lines and Jurkat cells, treated with the PvuII restriction enzyme, and then subjected to electrophoresis. If the extract contained MGMT activity, an O⁶-methylated guanine residue nested within the PvuII restriction site could be demethylated, and the oligonucleotide could be digested by PvuII. NC — uncleaved oligonucleotide; C — cleaved oligonucleotide.

To examine the expression of MGMT mRNA in the canine lymphoma cells, real-time PCR analysis was done on the 5 lymphoma cell lines and PBMCs collected from healthy dogs. The CL-1, Nody-1, and UL-1 cells, as well as the PBMCs, expressed MGMT in the absence of the methylation inhibitor. In contrast, no MGMT mRNA was detected in the Ema and GL-1 cell lines (Figure 3). Since methylation of the MGMT promoter region suppresses MGMT gene expression in some human cancer cells, the canine cells were pre-treated with 5-aza-2′-deoxycytidine to inhibit DNA methylation. As shown in Figure 3, after pretreatment with the methylation inhibitor the Ema and GL-1 cells gained expression of MGMT mRNA to some extent, whereas the other cell lines and the PBMCs did not show any change in MGMT mRNA expression. This suggests that MGMT mRNA expression is suppressed, at least in part, by DNA methylation in Ema and GL-1 cells but not in the other cell lines or in PBMCs.

Figure 3.

A — Expression of mRNA by the MGMT gene in comparison with the ribosomal protein L32 (RPL32) gene with (black bars) or without (white bars) pretreatment with the methylation inhibitor 5-aza-2′-deoxycytidine (final concentration 0.5 μM) as determined by real-time polymerase chain reaction (PCR). The PCR cycle number at the threshold was represented as Ct, and the difference between Cts for the target and internal control (ΔCt) was calculated. B — Results of agarose gel electrophoresis of the PCR products. PBMC — peripheral blood mono-nuclear cell.

Hypermethylation of the CpG islands surrounding the human MGMT exon 1 can suppress the expression of MGMT mRNA (30). To investigate if such suppression also occurs in canine lymphoma cells, bisulfite sequencing was done. Four CpG islands were identified in sequences from approximately 10 000 bp upstream to 7000 bp downstream of the canine MGMT exon 1 (Figure 4). We first examined region 1, which contains 24 CpGs, through bisulfite modification of DNA isolated from the canine lymphoma cells, PCR amplification, purification of the PCR products, and direct sequencing. As shown in Figure 5A, the CpGs in the Ema cells that did not express MGMT activity and mRNA were not methylated. In contrast, the CpGs in all the other cell lines were mostly methylated. To confirm these results, we cloned the amplified DNA into a plasmid vector and sequenced 5 independent clones (Figures 5B to 5D). Almost all the CpGs in the Ema cell line were found to be unmethylated by this method. Approximately 70% of the CpGs in the Nody-1 cells were methylated in more than 60% of the clones.

Figure 4.

Scheme of CpG islands [regions of unmethylated DNA with a high frequency of dinucleotides of C (cytosine) followed immediately by G (guanine), or C plus G (CpG)] surrounding the canine MGMT gene exon 1 and the regions analyzed in this study, which were approximately 10 000 bp upstream through 7000 bp downstream of the exon 1. The islands were determined by means of Methyl Primer Express software version 1.0 (http://www.appliedbiosystems.com/absite/us/en/home/support/software-community/free-ab-software.html) with the following parameters: a CpG content greater than 55% and a ratio of observed CpG to expected CpG greater than 0.65. A — CpG islands upstream of the canine MGMT exon 1. B — Nucleotide numbers of the analyzed regions, based on the sequence of Canis lupus familiaris chromosome 28 (GenBank accession number NC_006610.2).

Figure 5.

A — Percent methylation of 24 CpGs of region 1 obtained through bisulfite modification of DNA isolated from the canine lymphoma cells, PCR amplification, purification of the PCR products, and direct sequencing. Black bars — methylated cytosine; white bar — unmethylated cytosine; striped bars — undetermined. B to D — Amplified DNA from Ema and Nody-1 cell lines and PBMCs, respectively, was inserted into the pCR2.1 plasmid vector and transformed into bacteria. Five independent clones were analyzed and directly sequenced.

We next examined regions 2-1, 2-2, 3-1, and 3-2, which were found to contain 29, 30, 20, and 27 CpGs, respectively, by the direct bisulfite-sequencing method. The CpGs in the Ema cell line were unmethylated, while those in Nody-1 and the canine PBMC were mostly methylated (Figures 6A to 6C). Since we were unable to determine the methylation status of region 4 via the direct bisulfite-sequencing method, we analyzed it by the subcloning bisulfite-sequencing method (Figures 6D to 6F). Almost all the CpGs in the Ema cells were found to be unmethylated, while approximately half of the CpGs in Nody-1 and almost all the CpGs in the canine PBMCs were found to be methylated.

Figure 6.

Percent methylation in regions 2-1, 2-2, 3-1, 3-2, and 4. Methods and bar identification as for Figure 5. A to C — As determined by the bisulfite-sequencing method. D to F — As determined with at least 5 clones per cell line.

Discussion

Tomiyasu et al (31) have reported that expression of MGMT mRNA is not associated with the chemosensitivity observed in cases of canine B-cell lymphoma treated with the CHOP protocol. However, these patients were never treated with nitrosoureas. Lomustine is effective in treating relapses of canine lymphoma (6,7). Therefore, in this study we examined whether MGMT affects the sensitivity of canine lymphomas to lomustine.

As shown in Figure 1 and Table II, BG inactivation of MGMT activity significantly increased the toxicity of lomustine to CL-1 and Nody-1 cells. This increment was also observed in Jurkat cells that express MGMT (29). In addition, the fluorometric oligonucleotide assay revealed that CL-1 and Nody-1 cell lines have the same MGMT activity as Jurkat cells (Figure 2). Furthermore, MGMT gene expression was also observed in these cell lines with real-time PCR (Figure 3). These results strongly demonstrate that CL-1 and Nody-1 cell lines with MGMT activity are resistant to lomustine toxicity. In contrast, the Ema, GL-1, and UL-1 cell lines did not show a similar resistance to the cytotoxicity caused by lomustine, although real-time PCR detected weak MGMT expression in the UL-1 cells. These results indicate that MGMT strongly influences the sensitivity of canine lymphoma cell lines to lomustine. Furthermore, the IC₅₀ data (Table II) indicated that canine lymphoma cell lines such as CL-1 and Nody-1 that have MGMT activity are resistant to lomustine, whereas cell lines like Ema, GL-1, and UL-1 that have no MGMT activity are sensitive to lomustine.

Although it is unclear why cells do or do not have MGMT activity, we hypothesize that normal lymphocytes have MGMT activity (Figure 3) (32) and that MGMT may act as a tumor-suppressor gene by preventing DNA mutations. When MGMT expression is inhibited by an unknown mechanism, DNA may not be repaired, and this could lead to mutations in oncogenes and/or tumor-suppressor genes, resulting in tumorigenesis. If the suppression of MGMT occurs in the early stages of tumor development, gene mutations that are essential for tumorigenesis may remain unrepaired, promoting neoplasticity. Cancer development in lymphoma cells that have MGMT activity may result from mechanisms other than the inactivation of MGMT. Although MGMT may not be suppressed, accumulation of various gene mutations may establish cancer cells that constitutively express MGMT.

In this study, the CL-1, Nody-1, and Jurkat cells showed significantly lower IC₅₀ values than the Ema, GL-1, and UL-1 cells after BG treatment (data not shown). This finding indicates that mechanisms other than MGMT activity contribute to drug resistance in the latter group. In addition to MGMT activity, defective mismatch repair and overexpression of the negative cell-cycle regulator WAF1/Cip1 have been reported as causes of resistance to alkylating agents (33,34). A combination of these factors may result in complicated refractory situations with anticancer agents.

In human gliomas and colorectal cancer, methylation of MGMT exon 1 is strongly associated with the loss of MGMT mRNA expression (35). Therefore, we compared the expression level of MGMT mRNA with and without treatment with a methylation inhibitor to investigate whether DNA methylation affects MGMT expression in canine cells. As shown in Figure 3, treatment with the methylation inhibitor induced MGMT mRNA expression in Ema and GL-1 cells that did not originally express the MGMT gene. In contrast, the methylation inhibitor did not show any effect on MGMT mRNA expression in CL-1, Nody-1, and UL-1 cells or in PBMCs from healthy dogs. The increase in MGMT gene expression observed in this study indicates that MGMT gene expression is inhibited by DNA methylation in Ema and GL-1 cells.

To investigate the methylation status of the promoter region of the canine MGMT gene, we analyzed region 1, which corresponds to the promoter region in the human MGMT gene (Figure 4). Indeed, it has been demonstrated that methylation of this region contributes to the suppression of MGMT in human cancers (30). Intriguingly, the CpGs of region 1 of the Ema cells were completely unmethylated, and almost all the CpGs of the other 4 cell lines were methylated, as shown by the direct bisulfite-sequencing analysis (Figure 5A). To confirm these results, we performed a subcloning method followed by bisulfite sequencing on the Ema and Nody-1 cells, which were without and with MGMT gene expression and activity, respectively. The Ema cells showed 100% lack of methylation in almost all the CpGs, whereas the Nody-1 cells showed more than 60% methylation in 70% of the CpGs, and the PBMCs from healthy dogs showed 100% methylation in most of the CpGs (Figures 5B to 5D). Taken together, the results of the subcloning-mediated bisulfite sequencing reflect the results of the direct bisulfite-sequencing analysis in the Ema cells. Interestingly, the Nody-1 cells had many clones that were a mix of both methylated and unmethylated cytosines in the same CpGs, as seen by the subcloning-mediated bisulfite sequencing, and many of these regions showed a double wave of thymine and cytosine by the direct bisulfite-sequencing method (data not shown). Therefore, the CpG regions defined as “undetermined” in the direct bisulfite-sequencing analysis may have had a mixture of methylated and unmethylated cytosines. Indeed, the human cell lines expressing MGMT have been shown to have almost 100% demethylation of the CpGs in the promoter region, whereas variable methylation status of clones has been observed in the human cell lines that show no MGMT expression in the promoter region (30). These results indicate that if unmethylated cytosines are observed by direct bisulfite sequencing, most cells may not have CpG methylation. In contrast, if methylation is observed, the cells could contain a mix of methylated and unmethylated cytosines. Collectively, the results of the direct bisulfite sequencing may well correlate with the subcloning-mediated bisulfite sequencing. Although expression of the MGMT gene is suppressed by CpG methylation of the promoter region in human cells, Ema cells that do not express MGMT show complete CpG demethylation. Since MGMT gene expression is increased by the methylation inhibitor in Ema cells, region 1 is unlikely to correlate with the regulation of MGMT gene expression in canine cells.

To identify the CpG region whose methylation is fundamental to expression of the MGMT gene, we further analyzed the regions upstream of the canine MGMT gene. However, we could not identify the region (Figure 6). We can speculate that CpG methylation or demethylation in these regions depends on the cells rather than their association with MGMT expression; the CpGs in the Ema cells tend not to be methylated, whereas those in the Nody-1 cells and healthy canine PBMCs are methylated.

Although we targeted and analyzed the tentative promoter regions of the canine MGMT gene, including the upstream 10 000 bp and downstream 7000 bp of exon 1, we were unable to identify the regions in which methylation is essential to suppress expression of the MGMT gene. We hypothesize that the canine MGMT promoter may include regions other than the ones analyzed in this study and the ones found in the human MGMT promoter. Alternatively, expression of the canine MGMT gene may be regulated by a mechanism other than direct methylation of the MGMT promoter region. For example, promoter regions of other transcription factors that promote MGMT expression could be methylated, resulting in the suppression of MGMT. Additional studies are necessary to elucidate this mechanism.

This study has demonstrated that MGMT is strongly correlated with the resistance of canine lymphoma cell lines to the nitrosourea lomustine, as it is with human cells. Furthermore, we have shown that there are 2 types of lymphoma cells: those that possess MGMT activity and show low sensitivity to lomustine, and others that do not possess MGMT activity and show high sensitivity to lomustine. In future, investigating expression of the MGMT gene and its activity in clinical samples may prove beneficial. If our in-vitro results could be translated into in vivo results, they might help to predict the effectiveness of nitrosoureas and to overcome the resistance to these agents in lymphoma cases. The analysis of the effect of the methylation inhibitor suggested that MGMT gene expression may be suppressed by DNA methylation in canine lymphomas. It will be important to identify such regulatory regions and develop methylation-specific PCR strategies for clinical application in terms of predicting the effectiveness of nitrosoureas.