Haplotype analysis of chemokine and chemokine receptor polymorphisms

Motivated by the wealth of functional evidence that implicates chemokine signals in tumor development and metastasis, the ultimate goal of this study was to determine if variation in genes that encode these molecules is associated with breast cancer and AIDS-related cancers. To accomplish this, haplotypes of single nucleotide polymorphisms, or SNPs, were constructed in a 77kb cluster of three chemokine genes (MIP-1α, PARC and MPIF-1) on chromosome 17q12, and four chemokine receptor genes ( CCR3, CCR2, CCR5, and CCRL2) spanning 150kb on 3p21. SNPs were genotyped with the TaqMan method (5′ nuclease assay).; Pedigree analysis of 14 SNPs in 40 families revealed a high degree of linkage disequilibrium (LD) in the 3p21 gene region, and all variation in this sample was explained by 11 clearly-defined haplotypes. Haplotypes estimated with the Expectation-Maximization algorithm indicate that this strong LD is maintained globally, even in African populations. Most of the variation in the 3p21 SNPs (79%) is explained by 7 haplotypes that are found at a frequency of greater than 5% in a sample of 11 geographically diverse populations (N = 690). Overall, the 6 SNPs typed in 17q12 exhibit less linkage disequilibrium. Eight haplotypes, each greater than 5% in the total sample (N = 702), comprise 87% of the variation in the chemokine genotype data.; For association analyses, SNPs and haplotypes were tested by contingency tables and statistical permutation methods, in case/control samples of individuals with breast cancer and AIDS-related cancers. Comparisons of African-American breast cancer case (N = 90) and control (N = 110) samples yielded two significant results of note: a frequency difference of a common haplotype on 3p21, and a frequency difference of a nonsynonymous SNP (M106V) in the MPIF-1 chemokine on 17q12. Implementation of genomic control methods may be necessary to resolve possible confounding by substructure in this African-American sample. Contingency table analysis yielded only one significant difference in frequency between 235 Kaposi's sarcoma cases and 273 HIV-l+ controls, 801 (A/G) in the 3′UTR of the SDF-1 chemokine gene. Finally, comparisons of NHL cases and controls confirm a previous observation of the protective effect of the 32bp deletion in CCR5 against non-Hodgkin's lymphoma in HIV-1+ individuals.