Acute Promyelocytic Leukemia (APL) is characterized by the typical chromosomal translocation t(15;17)(q22;q21) leading to the fusion product PML-RARA, which blocks granulocytic differentiation in the promyelocyte stage. Several experimental in vitro and in vivo studies have demonstrated that PML-RARA is necessary but not sufficient for the generation of APL. This circumstance has motivated the search for additional leukemogenic and cooperating molecular lesions.
We have analyzed 101 APL patient bone marrow samples with high density Genome-Wide Human SNP 6.0 arrays, which interrogate >900.000 SNPs and >900.000 non-polymorphic copy number markers throughout the genome (Affymetrix, Santa Clara, CA, USA) in search for copy number alterations (CNAs) potentially relevant in the pathogenesis of APL. Genomic DNA from samples at initial diagnosis of 94 patients was analyzed. Furthermore, DNA from 11 samples at relapse was available, whereby 4 of these relapse samples also had paired DNA from initial diagnosis. Data analysis was carried out with the CNAG 3.3 software using anonymous references. For exclusion of copy number polymorphisms, all detected CNAs were compared with the databases of known copy number polymorphisms in the UCSC genome browser. For data validation, putatively acquired CNAs and regions of copy number neutral loss of heterozygosity (CNLOH) were confirmed by hybridization of DNA from paired normal samples when the patients were in remission, by quantitative real time PCR of genomic DNA and by direct sequencing of informative SNPs.
The high density SNP array analysis detected a total of 120 heterozygous deletions, 97 duplications or amplifications and 7 regions of telomeric CNLOH leading to an average of 2.3 CNAs per sample (range 0–30). The most common numerical and large structural aberrations were found on chromosome (chr.) 8 with either trisomy 8 (n=11) or duplication of regions on chr. 8q (n=10) followed by heterozygous deletions of chr. 7q (n=5) and chr. 16q (n=5). Furthermore, unbalanced translocations of chr. 15 and 17 involving PML and RARalpha were detected in five cases leading to duplication of the PML-RARA fusion or deletion of genomic regions flanking either PML or RARalpha. Recurrent microlesions (<1Mbp) were found in several regions as heterozygous deletions on chr. 1q31.3 containing the micro RNAs MIR181B1 and MIR181A1 (n=5), on chr. 2q32.3 containing serine/threonine kinase 17b (STK17B) (n=5) or chr. 3p24.3 containing ankyrin repeat domain 28 (ANKRD28) (n=5). One recurrent region of telomeric CNLOH was found on chr. 19q in two samples. Of note, besides the few regions of telomeric CNLOH a large number of intrachromosomal CNLOH regions (n=265) was identified, with recurrent regions on chr. 6p21.1 (n=10) or chr. 5q23.3-5q31.1 (n=6) containing genes relevant in hematopoiesis such as IL3, CSF2 or DNA damage repair such as RAD50. Although these CNLOH regions were not somatically acquired they may possibly harbor genetic predispositions for disease.
We describe a detailed high density SNP array genomic profiling of bone marrow DNA from patients with APL, which has led to the identification of several new cryptic recurrent genomic lesions. These genomic alterations point to candidate genes, which could be cooperating factors in addition to PML-RARA. Therefore, our data helps to provide a better understanding of the molecular mechanisms underlying the development of APL.
Kohlmann:MLL Munich Leukemia Laboratory: Employment. Lengfelder:Cephalon: Research Funding.
Asterisk with author names denotes non-ASH members.