Comprehensive Genomic Copy Number and Sequence Analysis of 28 Chromosome 5q31.2 Candidate Genes in De Novo MDS.

Deletions involving the long arm of chromosome 5 are among the most common recurring cytogenetic abnormalities detectable in human myelodysplastic syndromes (MDS). The minimally deleted segment has been mapped by several groups to a 2.3 megabase interval at 5q31.2. This region contains 21 annotated genes, one pseudogene, 4 small RNAs, and 2 predicted genes (SPOCK1, KLHL3, HNRPA0, NPY6R, MYOT, PKD2L2, C5orf5, WNT8A, NME5, BRD8, KIF20A, CDC23, GFRA3, CDC25C, FAM53C, JMJD1B, REEP2, EGR1, ETF1, HSPA9B, SNORD63, LOC391836, LOC729429, CTNNA1, LRRTM2, SIL1, MATR3, SNORA74A). Extensive study of many of these genes by other investigators has raised the possibility that gene(s) in this interval contribute to MDS pathogenesis by haploinsufficiency. To address this hypothesis, we selected samples from 46 de novo MDS patients that had adequate amounts of paired tumor (bone marrow) and germline (skin) DNA available. The 46 patients include all FAB subtypes, with a range of IPSS scores from 0–3 (median = 1), and a median blast count of 5%. The majority of these samples (n=37) do not have cytogenetically apparent 5q31.2 deletions. We first asked whether these samples might contain deletions in this interval below the limit of detection by cytogenetics. We used a custom oligomer based array comparative genomic hybridization (aCGH) platform developed by NimbleGen Systems, Inc. that has ∼385,000 probes spanning human chromosome 5, providing an average probe spacing of 500 base pairs. Analysis of aCGH data for 20 patients demonstrated that cytogenetically apparent 5q31.2 deletions could be detected by this platform, but no cytogenetically “silent” interstitial deletions were seen. We next asked whether genes in this interval might instead be targeted by point mutations. To define the sensitivity of our sequence-based studies, we used standard agarose gel electrophoresis and the Agilent LabChip to detect FLT3 internal tandem duplications (ITD) in 3 of 90 MDS samples. We chose one sample with an ITD and performed 6 serial dilutions of MDS DNA with wild-type genomic DNA and performed PCR amplification followed by DNA sequencing of the PCR products. We could detect the FLT3 ITD when 12% of the alleles harbored the mutation, or when ∼20–25% of cells contain a heterozygous mutation. From these results, we are confident that our strategy can detect mutant alleles in unpurified MDS bone marrow samples, but may miss mutations occurring in a rare cell. We then designed and validated primers for 415 amplicons covering the coding region and proximal introns of all 28 genes in the 5q31.2 interval and produced 7.2 megabases of sequence for these genes in the 46 patient samples. No somatic changes were identified. Twelve novel non-synonymous SNPs were discovered. The allele and genotype frequencies of 49 known SNPs in these genes were similar to the frequencies in race-matched HapMap individuals, apart from an enrichment in a non-synonymous CDC25C SNP in the MDS cohort (rs3734166, Odds Ratio=1.87, 95% confidence interval=1.06–3.32, p=0.036). The over-representation of this SNP and the other rare novel SNPs in the MDS population suggest that they may mark susceptibility alleles for MDS. Taken together, these data suggest that small deletions and/or point mutations of genes in 5q31.2 are not common events in the pathogenesis of MDS, and that larger deletions leading to haploinsufficiency of several genes in the interval appear to be the dominant mechanism.