Global genomic characterization of acute lymphoblastic leukemia

Abstract

A key goal in cancer research is to identify the total complement of genetic and epigenetic alterations that contribute to tumorigenesis. We are currently witnessing the rapid evolution and convergence of multiple genome-wide platforms that are making this goal a reality. Leading this effort are studies of the molecular lesions that underlie pediatric acute lymphoblastic leukemia (ALL). The recent application of microarray-based analyses of DNA copy number abnormalities (CNAs) in pediatric ALL, complemented by transcriptional profiling, resequencing and epigenetic approaches, has identified a high frequency of common genetic alterations in both B-progenitor and T-lineage ALL. These approaches have identified abnormalities in key pathways, including lymphoid differentiation, cell cycle regulation, tumor suppression, and drug responsiveness. Moreover, the nature and frequency of CNAs differ markedly among ALL genetic subtypes. In this article, we review the key findings from the published data on genome-wide analyses of ALL and highlight some of the technical aspects of data generation and analysis that must be carefully controlled to obtain optimal results.

Figure 1

A. Log2-ratio copy number heatmap of Affymetrix SNP array genome-wide copy number data obtained from 126 pediatric ALL cases. Blue is deletion, red is gain. B, Copy number heatmap of 61 pediatric ALL cases with CNAs involving PAX5 at 9p13.2. Many cases are not identifiable at this whole-chromosome resolution. The region shown in panel C is indicated by an asterisk. C, copy number status at PAX5 for the 61 cases shown in B.

Figure 2

Copy-neutral LOH in ALL. A, 16 pediatric ALL cases with copy neutral LOH involving chromosome 9. LOH is blue. B shows the corresponding copy number heatmap of chromosome 9 for the cases shown in A. The region shown in panel C is indicated by an asterisk. C, Affymetrix 615,000 SNP copy number status at 9p21.3 (the CDKN2A/B locus) for the same cases. Each case has a homozygous deletion that has been reduplicated, and is responsible for the flanking CN-LOH. Several of these deletions are not evident using lower resolution (50K) arrays.

Figure 3

Effect of array resolution on lesion detection in ALL. A shows log ratio copy number data at the IKZF1 (Ikaros) locus in a case with a focal deletion of exons 3-6. Log ratio copy number at each SNP is shown as a vertical line. Only 7 probes cover the region of deletion. The sparse coverage of the gene by the 250k and 50k Affymetrix arrays is shown below the IKZF1 gene symbol, with each SNP on each array represented by a vertical line. Part of IKZF1 intron 2 has no coverage due to a gap in the human genome assembly. B, copy number data for the same case using a Nimblegen 2.1 million feature array CGH platform. Coverage by the Nimblegen and Affymetrix SNP 6.0 arrays are shown (SNP probes are labeled “Affy SNP 6.0”, and copy number probes on the same array is “Affy SNP 6.0 SV”).