Array-CGH and Gene Expression Profiling Based Molecular Characterization of Myeloid Leukemia Cell Lines.

Myeloid leukemias are characterized by cytogenetic and molecular-genetic aberrations resulting in altered gene expression. Nevertheless, so far still little is known regarding the underlying mechanisms of leukemogenesis. To model and investigate the different aspects of leukemia pathogenesis a widely accepted approach is to use immortalized leukemia cell lines. While these provide powerful tools for the investigation of gene function, the dissection of signal transduction pathways and the analysis of drug effects, to our knowledge only few studies have addressed the question whether hematopoietic cell lines represent reliable model systems.

In order to improve the molecular characterization of these model systems we therefore analyzed 18 myeloid leukemia cell lines using DNA microarray technology. To determine the secondary aberrations acquired in addition to their characteristic primary cytogenetic aberrations during numerous passages in vitro, we first analyzed all cell lines by array-CGH (comparative genomic hybridization). Using a BAC/PAC platform with an average resolution of ~ 1.5 Mb across the entire genome we identified recurrent losses and gains, as well as high level amplifications like e.g. an amplification in 4q12 (Kasumi1) and in 8q24 (HL60). The parallel analysis of gene expression using a whole genome cDNA microarray platform helped to further delineate potential candidate genes in the affected regions (e.g. overexpression of KIT in the 4q12 and MYC in the 8q24 amplicon). Furthermore, unsupervised hierarchical cluster analysis revealed distinct gene signatures pointing towards dysregulated transcriptional networks. Comparison of our findings with acute myeloid leukemia patient data (Bullinger et al. 2004) showed the signatures underlying characteristic cytogenetic aberrations like e.g. t(15;17) were conserved in cell lines. Interestingly, these signatures were also quite robust as they displayed a highly significant correlation with published cell line data profiled on a different DNA microarray platform in a different laboratory.

Thus, our analyses demonstrate that cell lines exhibit conserved signatures correlating with the primary balanced cytogenetic aberrations, and that most cell lines even when grown and analyzed under different conditions provide highly robust signatures. Therefore, our refined molecular characterization of myeloid cell lines supports the utility of cell lines as powerful model systems and provides additional insights into the molecular mechanisms of leukemogenesis.