To obtain a comprehensive genomic profile of presenting multiple myeloma cases we performed high resolution single nucleotide polymorphism (SNP) mapping array analysis to examine changes in DNA copy number and loss of heterozygosity in 115 cases with matched peripheral blood controls. Identification of minimally deleted regions (MDRs), and the integration of expression array data, reduces the complexity of identifying key genes involved in the pathogenesis of myeloma in otherwise large genomic regions potentially containing hundreds of genes. Samples were purified by CD138 selection, classified by fluorescent in situ hybridization (FISH) results and included hyperdiploid (HRD) samples (n=48; 13 with translocations); t(4;14), n=15; t(6;14), n=1; t(11;14), n=19; t(14;16), n=4; t(14;20), n=2; no HRD or translocation, n=9; unknown translocation with IgH, n=5; and samples with no FISH data, n=12. DNA was used on the Affymetrix Human Mapping 500K array set and RNA was hybridized to the U133 Plus 2.0 expression array. The resulting data were analyzed using dChip, CNAG, and GeneSpring. The chromosomes with the most common deletional events were 1p (29.5%), 6q (33%), 8q (27%), 12 (21%), 13q (59%), 14 (36.5%), 16q (40%), 17p (7%), 20 (17%) and 22 (33%). In addition, copy-neutral LOH was also present on chromosomes 1p (5/115), 8 (1/115), 12 (1/115), 13 (2/115), 17 (4/115), 20 (1/115), 22 (3/115), and most prevalently on 1q (9/115) and 16q (10/115 samples), but not present at all on chromosomes 6 or 14. Using both regions of monosomy and copy-neutral LOH we were able to identify MDRs on these chromosomes which may harbor genes that are down-regulated in the tumor. On chromosome 1 there were 3 regions of interest; an MDR at 1p31.1 covering 4.2 Mb, and two regions containing homozygous deletions at the CDKN2C/FAF1 loci and at the FAM46C locus. Expression array data failed to detect a change in expression of the genes present in the MDR at 1p31.1 or at the FAM46C locus, however FAF1 was under-expressed in deleted samples. On 6q two common regions of deletion were apparent. The first, at 6q22.1, involved 23 samples and centered on the TSPYL4 locus but included NT5DC1, TSPYL1 and DSE, which were homozygously deleted in one sample. The second region, present in 32 samples, was located at 6q25.2–q26 and spanned 6.4 Mb containing 34 genes. Using array expression data we found that 14 of these genes are down-regulated in the deleted samples including IGF2R, TFB1M, WTAP and SOD2, which has been implicated in myeloma previously. On chromosome 13 an 18.75 Mb MDR was apparent in a region surrounding the RB1 locus amongst 57 other genes. However, one sample had a 235 kb homozygous deletion containing only RB1, P2RY5 and RCBTB2. Other homozygous deletions were detected on 13q, but they did not fall within the MDR. By integrating expression data from the same samples we were able to elucidate that RB1 had the largest down-regulation in deleted samples, closely followed by RCBTB2. Chromosome 14 contained a 3.3 Mb MDR at 14q24.1–24.2 with 21 genes in 27 samples. By integrating expression array data into the mapping data the list of genes whose expression is affected by the deletion is 7, with the most down-regulated being SPRS5, MED6, and WDR22. Additional common deletional events at 14q32.33 were observed and are due to translocation events involving the IGH locus. Homozygous deletions of TRAF3 and AMN were also noted. 16q had no single MDR, but 3 commonly deleted regions surrounding the CYLD, WWOX and ATBF1 loci were observed. All three loci were homozygously deleted in at least one sample, and WWOX and CYLD are known to be down-regulated by deletions on this chromosome. On 17p there was no MDR, but two samples had homozygous deletions neither of which affected TP53. As may be expected, TP53 expression was not down-regulated in deleted samples but the expression of several other genes was affected. Deletions on chromosomes 20 and 22 predominantly involved the whole chromosome, with the exception of the IGL locus at 22q11.22 which was more frequently deleted. Using this large dataset and high resolution mapping arrays we have been able to identify minimally deleted regions on chromosomes to narrow down the number of target genes which may be inactivated in myeloma. Furthermore, using expression array data we have been able to identify key target genes to study.
Disclosures: No relevant conflicts of interest to declare.
Author notes
Corresponding author