Genetic Association of Non Hodgkin’s Lymphoma Susceptibility to Polymorphisms in the CTLA-4 Gene on Chromosome 2q33.

The current knowledge of potential risk factors for lymphomas is limited. There is strong evidence that altered immunological function entails an increased risk of lymphoma. CTLA-4 gene encodes for a receptor that provides a negative signal to the T-cell once an immune response is initiated and complete. It is located in the 2q33 chromosomal region that harbours several genes such as CD28, ICOS, all related to immune activation playing a pivotal role in T-lymphocyte activation. Polymorphisms in these genes have been associated to a number of autoimmune diseases, including blood disorders. We have recently reported that a functional polymorphism in the CTLA-4 gene is associated with Non-Hodgkin’s lymphoma (NHL) and may have a role in genetic susceptibility to the disease.

Since CD28, CTLA-4 and ICOS are closely linked and have related functions, aim of the present study was to extend the genetic analysis to the 2q33 chromosomal region harbouring these genes.

Three polymorphisms of CTLA-4 gene (at positions -308C*T, +49A*G, +642 3′UTR (AT)n), the CD28 (CAA)n, ICOS c1564T*C and D2S72 markers within a 1.3cM region were analyzed in 100 unrelated NHL patients and in 128 healthy controls both originating from the central part of Sardinia. Statistical analysis was performed to test association of each marker with NHL using Chi-squared test and Odds Ratio (OR). Haplotypes were inferred and analyzed using the PHASE 2.1 software. Exact test of Linkage Disequilibrium (LD) was calculated between all pairs of seven markers separately in cases and controls using the Arlequin program.

We found a strong association of the CTLA-4 +49*A and the 3′UTR (AT)82 alleles with NHL with ORs of 2 (CI95% 1.2–3.2) and 1.6 (CI95% 1.1–2.4), respectively. The −308*C−+49*A− (AT)82 was the most represented haplotype in the studied population and resulted associated with NHL (p value = 0.0029; OR = 1.7 [CI95% 1.2–2.5]). No independent association between CD28, ICOS and D2S72 markers was observed. Genetic analysis of the entire haplotype CD28–CTLA-4-D2S72-ICOS resulted in 196 and 234 different best haplotypes in patients and controls, respectively. The most common haplotype was CD28*163-308*C-+49*A-(AT)82-D2S72*156-ICOS*T and it was found to be significantly over-represented in NHLs than in controls (p value = 0.012, OR=1.86 CI95% 1.1–3). Strong LD was observed between the three CTLA-4 gene markers and the two adjacent markers CD28 and D2S72, and less significant LD was found between ICOS and some of the CTLA-4 gene markers. We found genetic linkage between a specific haplotype in the 2q33 region and NHL, but allelic association was detected for CTLA-4 markers and for none of the adjacent functionally related gene markers. Thus, it seems possible to dissect the CTLA-4 as the true “causative” risk gene for NHL susceptibility at least in Sardinian patients.The functional characterization of disease-associated CTLA4 gene variants will be worthy of future investigation to elucidate their role in the pathogenesis of NHL.