Tumor Necrosis Factor α Polymorphism Associated with Increased Susceptibility to Development of Adult T-Cell Leukemia/Lymphoma in Human T-lymphotropic Virus Type 1 Carriers

INTRODUCTION

ATL 3 is a distinct clinicopathological entity, i.e., peripheral T-lymphotropic malignancy caused by human T-lymphotropic virus type 1 (HTLV-1) (1 , 2) . The incidence of ATL is rather low (2–5%) among HTLV-1 carriers, and those who develop the disease usually have a 40- to 60-year latency period from the time of infection to progression to ATL (3) . Progression to ATL may require at least five events, based on a statistical analysis of development of ATL according to age (4) . The tumor suppressor genes p53, Rb, p15^INK4A, and p16^INK4B, and an apoptosis-signaling cell surface receptor, Fas, are often abnormal in aggressive ATL (5, 6, 7, 8) . Sometimes, these genetic abnormalities have been associated with the transition from indolent (chronic or smoldering) ATL to aggressive (acute or lymphoma) ATL (5 , 7) . Factors reportedly associated with the development of ATL include infection early in life, increase in age, male gender, past history of infective dermatitis, smoking of tobacco, serum antibodies against HTLV-1 and selective HLA subtypes (9, 10, 11, 12, 13, 14) . Only a few reports have delved into possible genetic susceptibilities to development of ATL, and no genetic factors promoting the progression to ATL have been identified (14 , 15) .

Polymorphisms of some genes have been associated with their altered function and have been linked to the pathogenesis of various diseases including malignancy (16 , 17) . We investigated among HTLV-1 carriers the relationship between susceptibility to ATL and several polymorphisms: “decreased-detoxifying” polymorphisms in the GSTM1, GSTT1, and CYP1A1 genes, the “proapoptotic” polymorphism in BCL2, and five “high-production” polymorphisms in TNF-α (17, 18, 19) .

MATERIALS AND METHODS

Subjects

HTLV-1 carriers (n = 80) were identified in the general population in an endemic region for HTLV-1, the Nagasaki district of Japan, after a mass medical and laboratory screening for individuals with anti-HTLV-1 serum antibodies, as described previously (20) . Serum samples were examined for anti-HTLV-1 antibodies by particle agglutination assay and ELISA (21 , 22) . Samples positive for both assays were considered to be positive for anti-HTLV-1 antibody.

ATL patients (n = 71) were diagnosed at the Department of Hematology, Nagasaki University School of Medicine, which is one of the major hospitals for ATL in the Nagasaki district of Japan, by clinical features, abnormal cells having the immunophenotype of mature T cells, presence of anti-HTLV-1 serum antibodies, and monoclonal integration of HTLV-1 proviral DNA in the tumor cells as detected by Southern blotting, as described previously (23) . The subgroups of ATL were defined based on the Lymphoma Study Group classification (24) . All subjects were Japanese.

Genotyping Analysis

After informed consent, blood samples were obtained at the onset of ATL and at the time of mass examination for HTLV-1 carriers. High molecular weight DNA was extracted by the standard phenol-chloroform method. We analyzed DNA using PCR-based genotyping assays for five genes (see below). Negative and positive control samples were included in each amplification series.

GSTM1.

The polymorphic deletion of the GSTM1 gene was genotyped using the multiplex PCR approach as described previously (25) . The PCR primers were: P1, 5′-CGCCATCTTGTGCTACATTGCCCG; P2, 5′-ATCTTCTCCTCTTCTGTCTC; and P3, 5′-TTCTGGATTGTAGCAGATCA. P1 and P3 amplified a 230-bp product that is specific to GSTM1, whereas P1 and P2 annealed to GSTM1 and GSTM4 genes yielding a 157-bp fragment that serves as an internal control. PCR was performed in 20 μl containing 50 ng of genomic DNA, 0.5 μm of each primer, 200 μm of each dNTP, 10 mm Tris-HCl (pH 8.3), 50 mm KCl, 1.5 mm MgCl₂, and 0.5 units of Taq DNA polymerase (Qiagen, Chatsworth, CA). After denaturation for 4 min at 94°C, the PCR was performed for 35 cycles of 30 s at 94°C, 1 min at 58°C, and 1 min at 72°C. The last elongation step was extended to 7 min. GSTM1 alleles, identified by a 230-bp fragment, or its complete deletion (null genotype), was analyzed by electrophoresis on a 1% agarose gel. The absence of amplifiable GSTM1 (in the presence of the GSTM4 coamplifiable control) indicates a null genotype.

GSTT1.

The polymorphic deletion of the GSTT1 gene was determined by a modification of the PCR protocol described previously (25) . The amplification reaction was performed in 20 μl containing 50 ng of genomic DNA, 0.5 μm of each primer, 200 μm of each dNTP, 10 mm Tris-HCl (pH 8.3), 50 mm KCl, 2.0 mm MgCl₂, and 0.5 units of Taq DNA polymerase (Qiagen). The primers used to amplify GSTT1 were: sense, 5′-GCCCTGGCTAGTTGCTGAAG and antisense, 5′-GCATCTGATTTGGGGACCACA. A 268-bp fragment in exon 1 of the β-globin gene was coamplified with a sense primer, 5′-CAACTTCATCCACGTTCACC and an antisense primer, 5′-GAAGAGCCAAGGACAGGTAC as a control. The PCR was performed for 35 cycles of 15 s at 94°C, 30 s at 58°C, and 45 s at 72°C. The last elongation step was extended to 7 min. The presence of one or two GSTT1 alleles, identified by a 112-bp fragment, or its complete deletion (null genotype) was shown by electrophoresis on a 1% agarose gel. GSTT1 genotypes were scored only if the PCR signal corresponding to the β-globin internal control was evident.

CYP1A1.

CYP1A1 polymorphism T6235C was characterized by analysis of a PCR-RFLP as described previously (25) . A DNA fragment of 899 bp was amplified in 20 μl containing 50 ng of genomic DNA, 0.5 μm of sense primer, 5′-GGCTGAGCAATCTGACCCTA and antisense primer, 5′-TAGGAGTCTTGTCTCATGCCT, 200 μm of each dNTP, 10 mm Tris-HCl (pH 8.3), 50 mm KCl, 2.5 mm MgCl₂, and 0.5 units of Taq DNA polymerase (Qiagen). The PCR was performed for 35 cycles of 30 s at 94°C, 1 min at 63°C, and 1 min at 72°C. The PCR product (5–10 μl) was digested with MspI (3 units for 1 h at 37°C), resulting in smaller fragments (693 and 206 bp) in the case of the polymorphism, and was subjected to electrophoresis on a 1% agarose gel.

BCL2.

PCR was performed to amplify DNA segments within the portion of the BCL2 open reading frame corresponding to amino acids 1–162 using two primer sets (26 , 27) : sense (−40) AGAGGTGCCGTTGGCCCCCGTTGC/antisense (221) GTCTGCAGCGGCGAGGTCCT and sense (202) AGGACCTCGCCGCTGCAGAC/antisense (487) TGACGCTCTCCACACACATGAC. This portion included all somatic mutations and polymorphisms in BCL2. The amplification reaction was performed in 20 μl containing 100 ng of genomic DNA, 0.5 μm of each primer, 200 μm of each dNTP, 1 μCi of [³²P]dCTP, 10 mm Tris-HCl (pH 8.3), 50 mm KCl, 1.5 mm MgCl₂, and 1 unit of Taq DNA polymerase (Qiagen). The PCR was performed for 30 cycles of 1 min at 94°C, 2 min at 58°C, and 2 min at 72°C. The last elongation step was extended to 7 min. SSCP analysis was done using mutation detection enhancement gel (FMC BioProducts, Chicago, IL), as previously reported (28) . Aberrantly migrating bands were excised from the gel and reamplified. Purified PCR products were sequenced by the Applied Biosystems PRIZM dye terminator cycle sequencing reaction (Perkin-Elmer, Foster City, CA).

TNF-α.

TNF-α polymorphism was characterized by a modification of SSOP dot blot hybridization as described previously (19) . A 1042-bp DNA fragment of the 5′-flanking promoter/enhancer region of the TNF-α gene at position −1107 to −66 was amplified with a sense primer, 5′-GCTTGTGTGTGTGTGTCTGG, and an antisense primer, 5′-GGACACACAAGCATCAAGG. PCR was performed in 20 μl containing 100 ng of genomic DNA, 1 μm of each primer, 300 μm of each dNTP, 10 mm Tris-HCl (pH 8.3), 50 mm KCl, 1.5 mm MgCl₂, and 0.5 units of Taq DNA polymerase (Qiagen). After denaturation for 1 min at 94°C, the PCR was performed for 35 cycles of 1 min at 94°C,annealing at 60°C for 2 min, and extension at 72°C for 3 min. Two microliters of the PCR products were blotted onto a Hybond-N (Amersham, Arlington Heights, IL) and hybridized with ³²P-end-labeled SSOPs in a hybridization buffer (3 m TMAC; Sigma, St. Louis, MO), 50 mm Tris-HCl (pH 8.0), 2 mm EDTA, 0.1% SDS, and 5× Denhardt’s solution) at 54°C for 2 h. After the incubation, the filter was washed with TMAC solution (same as the hybridization buffer but without 5× Denhardt’s solution) at 58°C for 5–10 min, three times. The following SSOPs were used: for the −1031T analysis, 5′-TGAGAAGATGAAGGAAAA; for the −1031C, 5′-CTGAGAAGACGAAGGAAA; for the −863C, 5′-ATGGGGACCCCCCCTTAA; for the −863A, 5′-ATGGGGACCCCCACTTAA; for the −857C, 5′-CTTAACGAAGACAGGGCC; for the −857T, 5′-TTAATGAAGACAGGGCCA; for the 308G, 5′-AGGGGCATGGGGACGGG; for the −308A, 5′-AGGGGCATGAGGACGGG; for the −238G, 5′-CGGAATCGGAGCAGGGAG; and for the −238A, 5′-CGGAATCAGAGCAGGGAG.

Statistical Analysis

The statistical significance of the difference between the groups was calculated using the Fisher exact test (two-tailed) or the Mann-Whitney U test. Crude odds ratios were calculated and were given within a 95% confidence interval. All analyses were performed using the Prizm package (Statistical Software, Los Angeles, CA).

RESULTS

Genetic polymorphisms of TNF-α, GSTM1, GSTT1, CYP1A1, and BCL2 were determined in 71 individuals with ATL and 80 healthy HTLV-1 carriers of a similar ethnic background (Table 1) ⇓ . The HTLV-1 carriers were significantly older than the ATL patients. The male:female ratio in patients was significantly higher than in the carriers.

The frequency of the variant alleles of the TNF-α gene as well as the distribution of GSTM1, GSTT1, CYP1A1, and BCL2 genotypes in the individuals with ATL and the HTLV-1 carriers are demonstrated in Tables 2 ⇓ and 3 ⇓ , respectively. The most important finding was that we discovered the frequency of the TNF-α-857T allele was significantly increased in those individuals with ATL compared with the HTLV-1 carriers (odds ratio, 2.34; 95% confidence interval, 1.2–4.7, P = 0.017; Table 2 ⇓ ). No significant differences occurred between the two cohorts in their frequency of the other four TNF-α polymorphisms, although the odds ratios (ATL versus carrier) of all of the TNF-α polymorphisms were higher than 1.0.

Using SSCP and sequence analyses, we detected previously reported polymorphism (Ala⁴³Thr) of the BCL2 gene (18) , but no somatic mutation of this gene was found. Analysis of polymorphisms for GSTM1, GSTT1, CYP1A1, and BCL2 showed no significant difference between those individuals who had ATL or those who were HTLV-1 carriers (Table 3) ⇓ .

We subdivided the 71 ATL cases in this series into those with acute type (39 cases), lymphoma type (6 cases), chronic type (24 cases), and smoldering type (2 cases). No differences in the frequency of polymorphisms in the five genes between the aggressive forms (acute or chronic types) and the indolent forms (chronic or smoldering types) of ATL were observed (data not shown).

DISCUSSION

Although epidemiological factors have been associated with the development of ATL among HTLV-1 carriers (9, 10, 11, 12, 13, 14) , few reports explored the molecular epidemiology associated with the development of ATL (14 , 15) . To our knowledge, this is one of the first reports showing a genetic polymorphism associated with the susceptibility to development of ATL among HTLV-1 carriers.

TNF-α is a major cytokine involved in the promotion of inflammatory responses, and it plays a critical role in the pathogenesis of various inflammatory, autoimmune, and malignant diseases (29) . High-production polymorphisms in the promoter region of the TNF-α gene at position −308 or −238 have been associated with increased severity of infectious diseases, autoimmune diseases, and non-Hodgkin’s lymphomas, suggesting that genetic variations within the TNF-α locus may be functionally relevant in vivo (30, 31, 32) . Recently, three new high-production polymorphisms located in the 5′-flanking promoter/enhancer region of the TNF-α gene at positions −1031, −863, and −857 have been identified in a relatively large proportion of healthy Japanese individuals (19) . In the present study, we found that the frequency of the TNF-α-857T allele, which is associated with high transcriptional activity of the TNF-α gene, was enriched in individuals with ATL compared with healthy carriers. None of the other high producer TNF-α polymorphisms proved to be significant indicators of the risk of ATL, although odds ratios (ATL versus carrier) of all of the TNF-α polymorphisms were higher than 1.0. A similar study performed with individuals who had either HU, ATL, or were HTLV-1 carriers or healthy controls showed that the frequency of the TNF-α polymorphisms at −1031 and −863 was significantly higher in individuals with HU than in the control individuals (15) . They found no significant difference in the frequency of any of the five polymorphisms among ATL patients, HTLV-1 carriers, and controls, although the odds ratio (ATL versus carrier) of polymorphisms at positions −1031, −863, and −857 was higher than 1.0. However in that study, HTLV-1 carriers were selected from individuals visiting a hospital, and no information was provided concerning their ages (15) . Taken together, genetic polymorphisms leading to increased TNF-α production apparently enhance susceptibility to not only ATL, but also to another HTLV-1 related disease, HU.

We found no difference in the frequency of any of the five TNF-α polymorphisms when comparing those individuals with aggressive versus indolent types of ATL. In contrast, a high frequency of several TNF-α polymorphisms has been associated with increased plasma levels of TNF-α in a subset of non-Hodgkin’s lymphoma patients who had an adverse course (32) . We and others have shown that genetic alterations associated with the late stage of multistep leukemogenesis in ATL include p53 mutation, Rb, and p15^INK4A/p16^INK4B deletion and Fas mutations, all of which are somatic changes (5, 6, 7, 8) . The present study suggests that genetic polymorphisms such as in the TNF-α gene may enhance the early progression to the development of leukemia in HTLV-1 carriers.

We and others previously reported that HTLV-1 infected T-cell clones from HTLV-1 carriers as well as from ATL patients produced significant amounts of various cytokines, including TNF-α (33 , 34) . Aberrant production of these various cytokines was thought to be a result of the transactivating abilities of the Tax protein of HTLV-1 (35) . Several studies suggested that TNF-α might promote the growth of lymphoid cells (32 , 36) . Genetic polymorphism leading to increased TNF-α production may enhance susceptibility to ATL among HTLV-1 carriers. Nevertheless, several studies regarding the influence of TNF-α polymorphisms on the clinical course of either autoimmune or infectious diseases have suggested that a single polymorphism of the TNF-α gene probably cannot account for the variability of the disease course (37 , 38) . Furthermore, the HLA complex or other genes in linkage disequilibrium with HLA might play an important role in leukemogenesis because the TNF-α gene lies within the class III region of the MHC complex, and the −857T allele is in significant linkage disequilibrium with HLA-B54, -B35, -B59, and -DRB1*0405 (19) . However, the HLA alleles reportedly associated with the development of ATL in Japanese indviduals do not include these alleles (14 , 39) .

The average age at onset of ATL in this study is similar to that previously reported from a Japanese multicentric analysis (10) . We found that HTLV-1 carriers were significantly older than ATL patients, probably reflecting the high average age of the individuals who were participants in the mass examination. The male:female ratio of our individuals with ATL compared to the carriers was significantly higher, probably because a higher incidence of HTLV-1 infection is found in females and a high frequency of progression to ATL occurs in males, as previously reported (3 , 10 , 40, 41, 42) .

Several polymorphisms of human drug-metabolizing enzymes, including GSTM1, GSTT1, and CYP1A1 which were analyzed in this study, influence an individual’s susceptibility for several malignancies (17) . For example, lung cancer has been found to be more frequent in smokers with GSTM1 and CYP1A1 polymorphisms (43) . Although smoking of tobacco was found to be a cofactor for development of ATL in HTLV-1 carriers by a case control study (12) , our analysis of polymorphisms for GSTM1, GSTT1, and CYP1A1 showed no significant difference between ATL patients and healthy carriers.

BCL2 is an important antiapoptotic protein, which is highly expressed in follicular center lymphoma; these cells have the chromosomal translocation t(14:18) placing the BCL2 gene in juxtaposition with the immunoglobulin heavy chain locus (44) . Furthermore, many cancers without the t(14:18) including ATL have abundant levels of BCL2 protein (44 , 45) . BCL2-transgenic mice who overexpress the gene frequently develop T-cell lymphoma (46) . Mutations in the coding region of BCL2 have been detected in cell lines derived from lymphoid malignancies without the t(14:18) and in samples of aggressive lymphomas having t(14:18) (26 , 27 , 47) . By SSCP analysis, we found no somatic mutations of the coding region of BCL2 in ATL. We did, however, detect a polymorphism (Ala⁴³Thr) with proapoptotic function, which had been recently reported to be found fairly frequently in the Japanese population (18) . However, we found that the frequency of this polymorphism was similar in individuals with ATL and in healthy carriers.

In summary, the frequency of the TNF-α-857T allele, reported to be associated with high transcriptional activity of the promoter/enhancer region of the TNF-α gene, was more frequent in individuals with ATL compared with healthy HTLV-1 carriers. This germ line change may contribute to the rare progression from infection with HTLV-1 to ATL.