Identification of New Recurrent Multiple Small Interstitial Deletions Affecting Genes Coding for Kinases in Chronic Lymphocytic Leukemia (CLL): a New Pathogenic Mechanism?.

CLL is the most common adult-onset leukemia in the Western world. The most common known genetic lesion is the 13q14.3 deletion targeting MIR15/MIR16. We applied a very high resolution array to identify new genetic lesions in CLL.

266 CLL samples were analyzed with Affymetrix Human Mapping 6.0 arrays, comprising over 1,8 million probes with a median distance of less than 1 Kb. Copy number was inferred using the circulary binary segmentation (CBS) algorithm. Minimal common regions (MCR) were defined using a modified version of the algorithm by Lenz et al. (PNAS 2008), specifically altered to identify very small genomic losses covered by only 2-9 probes and occurring in at least 5% of the cases (mMCRs). mMCRs having 100% overlap with known copy number variations were discarded.

mMCRs occurred in 75 known genes. The most commonly affected genes were CDC73 (cell division 73; 63% of the cases, 3 probes), RREB1 (ras responsive element binding protein 1; 60%, 5 probes), JAK2 (47%, 8 probes), CCDC88A (AKT-phosphorylation enhancer,; 47%, 3 probes), AKT3 (43%, 4 probes). Other affected genes at a lower frequency were PIK3CA (26%), EGFR (25%), XRCC4 (18%), JAK1 (18%), PTPRK (15%), RB1 (14%), ERBB2 (10%), PDGFRA (8%), FHIT (7%). A functional analysis performed with DAVID 2008 (http://david.abcc.ncifcrf.gov/) identified the terms “anti-oncogene” and “tyrosine-protein kinase” and five KEGG (http://www.genome.jp/kegg/) pathways (“prostate cancer”, “non-small cell lung cancer”, “pancreatic cancer”, “endometrial” cancer”) as enriched among the 75 genes with a statistically significant p-value <0.05 after Benjamini multiple test correction. Besides tumor suppressor genes such as RB1 and FHIT, very interestingly, many of the genes appeared to code for kinases and for oncogenes. The mMCRs occurred in intronic regions, and apparently targeted highly conserved regions. These regions might represent regulatory loci and their loss may cause gene activation. Validation of selected genes is on-going.

The application of high resolution arrays on a large series of CLL samples has shown frequent small interstitial deletions targeting a discrete number of genes, highly enriched for transcripts coding for kinases. A potential mechanism of action might be the loss of regulatory regions determining gene activation. Once validated, the current data would provide the basis to explore the rationale for the use of kinase inhibitors in the treatment of CLL.

No relevant conflicts of interest to declare.