Interleukin 6, tumour necrosis factor alpha and lymphotoxin alpha polymorphisms in monoclonal gammopathy of uncertain significance and multiple myeloma.

Abstract

We read with interest the paper by Zheng et al (2000) and would like to comment on their findings concerning the interleukin 6 (IL-6) −174 G/C polymorphism and the tumour necrosis factor α (TNF-α) −308 restriction fragment length polymorphism (RFLP). We have analysed a series of 261 DNA samples from Caucasian patients with plasma cell dyscrasias, of which 198 had multiple myeloma (MM) and 63 had monoclonal gammopathy of uncertain significance (MGUS). We determined the genotypes of these samples for the IL-6 −174 polymorphism and for the TNFα−308, and lymphotoxin α (LT-α) +252 RFLPs and compared them with a series of 250 age- and sex-matched controls. Patient and control details have been described fully elsewhere (Davies et al, 2000; Rollinson et al, 2000). Genotyping for the IL-6 polymorphism was achieved using TaqMan allelic discrimination technology. Primers flanking the −174 polymorphic region were: 5′-CAA TGA CGA CCT AAG CTG CAC TT-3′ (forward) and 5′-GGG CTG ATT GGA AAC CTT ATT AAG AT-3′ (reverse), and probes complementary to the polymorphic region were: ACG TCC TTT AGC ATG/C GCA AGA CAC AAC TAG (PE BioSystems). TaqMan positive control DNA sample genotypes were confirmed by sequencing. Thermal cycling conditions were: 50°C for 2 min, 95°C for 10 min, followed by 40 cycles of 95°C for 15 s and 66°C for 1 min. The TNF-α and LT-α RFLP methods are described elsewhere (Davies et al, 2000). The TNF-α gene, located on chromosome 6p, is known to be in linkage disequilibrium with the gene for lymphotoxin alpha (LT-α). Like TNF-α, LT-α is a proinflammatory cytokine and inducer of IL-6. A G→A RFLP at +252 of the first intron of the LT-α gene results in the generation of two alleles, LT-10.5 and LT-5.5 (Wilson et al, 1993), and the LT-5.5 allele has been shown to be associated with a high producer phenotype. It is therefore important when looking at disease associations with these genes to consider the TNF-α and LT-α polymorphisms together as haplotypes, as both the TNF-2 and LT-5.5 alleles are associated with higher cytokine levels (Warzocha et al, 1998). In our large series of cases and controls, we found no associations of this IL-6 polymorphism with MM and MGUS, in agreement with Zheng et al (2000). However, the observed frequencies of the C allele in our MM, MGUS and control populations (0·4, 0·36 and 0·43 respectively) are closer to those stated previously (Fishman et al, 1998), which may highlight a difference in allele distribution between the populations studied. Zheng et al (2000) report no significant association of the TNF-α−308 RFLP with MM/MGUS; however, as described above, they failed to take into account variation at the LT-α locus. When this is taken into account (Davies et al, 2000), we found a significant association of high producer haplotypes with MM. Double heterozygotes TNF-1/2 LT-10.5/5.5 were present in 36% of MM cases but in only 18% of the controls, suggesting that this haplotype is associated with a significantly increased risk of MM (odds ratio 2·05, 95% confidence intervals 1·26–3·35). A similar distribution of alleles was seen in the MGUS cases, implying a possible role for these polymorphisms in a general predisposition to plasma cell neoplasms rather than in progression of MGUS to MM. To conclude, we would like to confirm the findings of Zheng et al (2000) in a larger series of patients with regard to the IL-6 −174 polymorphism, but to add that the TNF-α−308 RFLP, when considered together with the LT-α+252 RFLP, does appear to be associated with an increased risk of plasma cell disease.