Abstract
Abstract
Chromosomal damage is a hallmark of MDS. Inability to detect abnormal karyotype in a portion of MDS patients is consistent with the theory that large lesions may represent the extreme of possible DNA damage. Predisposition to MDS may be due to inherited defects; possibly, genes coding for detoxifying enzymes, DNA repair genes and immunogenetic factors could be involved and the risk can be multifactorial. To date studies dealing with complex pathogenesis of MDS have been based on empiric approaches allowing for very limited insights into the possibly complex genetic traits and DNA changes during evolution of the disease. Gene array technology facilitates detailed genomic analysis. We have applied 50K SNP Affymetrix arrays to the analysis of the MDS genome. The goal of our study was to establish the feasibility of this technology to study 2 separate aspects: i) to detect acquired cryptic chromosomal damage, ii) to investigate germ line allelic variants in MDS that could constitute predisposition factors. Our MDS cohort included 22 patients 15 RA/RS, 7 RAEB; 11 with typical cytogenetic abnormalities and 11 normal karyotypes. Using a threshold value 6 -log10(pval) we confirmed previously known changes in 8/11 patients. The remaining 3 showed multiple scattered lesions throughout affected chromosomes; false negatives were likely due to dilution by non-clonal cells. However, using a high density SNP scan, novel previously cryptic lesions contiguous over various chromosomal portions were found in 5/11 patients with normal metaphase cytogenetics and some patients with previously established defects. Chromosomes 5, 7 and 8 showed a higher number of smaller defects consistent with the pathogenesis of MDS; deletions within 7p14, 7p12.2 and 7q21.3 were found in 4 patients without obvious mononosomy 7. Similarly, deletions of various size within 5q12.3–35.5 were seen in 4/15 patients with low-grade MDS. Analysis of such defects within phenotypically defined subsets of patients may reveal possible consensus lesions or commonly affected genes.
We also investigated the presence of putatively MDS-specific genotypes. Overrepresentation of otherwise rare SNP may suggest existence of cryptic changes involving adjacent portions of the genome. Globally SNP array allowed for classification of 55953 SNP in 22 patients. 28% were heterozygous, while 36% were homozygous for either variant. Complex analysis of such a large number of genotypes may allow for determination of potential associations between genetic makeup and clinical sub-entities. Based on the allelic frequencies of individual SNP in MDS group and controls, we have identified significantly overrepresented SNPs occurring in either homo- or heterozygous form. For example, FOXL1 SNP occurring at a frequency of 2.4% in homo- and heterozygote form in normals was found in 25% of patients, and 10% of patients were homozygous for a Semaphorin SNP present in only 0.015% of controls. Similarly, PARP variant showed an allelic frequency of 2.7 vs. 14% in controls and patients, respectively. In general, our study show the potential value of high density SNP arrays in precise analysis of clonal genomic lesions and/loss heterozygozity as well as complex genotypic profiles that may potentially contribute to inherited predisposition traits.