Gene Expression Profiles of Hydroxyurea Tolerant Cell Lines Identify Genetic Mechanisms That Influence Hydroxyurea Maximum Tolerated Dose

Abstract 1005

Hydroxyurea has emerged over the past decade as an effective therapeutic agent for patients with sickle cell anemia (SCA). However, drug dosing and hematological responses can be highly variable; both %HbF response and maximum tolerated dose (MTD) vary widely among patients with SCA who receive hydroxyurea treatment. To obtain further insight into the cellular and molecular pathways, as well as genetic factors that might influence the hydroxyurea MTD, K562 erythroleukemia cells were exposed to hydroxyurea in vitro, to create cell lines that were highly drug tolerant to doses ranging from 250μM to 1500μM. Cell lines had dose-response curves that exhibited clear drug tolerance; naïve K562 showed 50% proliferation in the presence of 250μM hydroxyurea, while tolerant cell lines showed >90% proliferation at the same dose as measured by the BrdU Cell Proliferation Assay. In addition, the tolerant lines showed normal and equivalent progression through cell cycle by flow cytometry cell cycle analysis. After 15 weeks of continuous exposure, cells were harvested and mRNA microarray expression profiles were analyzed for naïve K562 (no hydroxyurea exposure) and cell lines tolerant to 500, 1000, or 1500μM hydroxyurea. Gene expression was measured on Affymetrix U133 Plus 2.0 chips. Differential expression between sample groups was determined using ANOVA, and p-values were corrected for multiple testing using the Benjamin-Hochberg false discovery rate (FDR) method to identify genetic profiles and genes consistently increased or decreased compared to naïve K562 cells. Using a threshold of 2-fold change compared to untreated cells and a false discovery rate <5%, a total of 864 genes were significantly altered in hydroxyurea tolerant cells, including 337 genes whose expression consistently correlated with increasing hydroxyurea dose (Pearson correlation p<.001). The PANTHER classification system was used to group genes into categories based on molecular functions. Of the genes that correlated significantly with increasing hydroxyurea dosing (n=337), there were 181 up-regulated genes and 156 down-regulated genes that had molecular functions including catalytic activity, binding, transcription regulator activity and transporter activity. Genes with transporter activity included SLC6A19, ATP6VOD1, ABCG2, ATP6V1B2 and KCNN4. Other genes of interest based on function included RRM2, PLS3, KCNAB2, UBE2A and SRI. Real-time quantitative reverse transcription (RT)-PCR then quantified the expression of 20 candidate genes to verify the accuracy of the microarray expression data. The next steps will include correlation of these findings with clinical data, specifically early reticulocyte mRNA expression and hydroxyurea MTD values obtained from children with SCA enrolled in the prospective Hydroxyurea Study of Long-term Effects (HUSTLE, NCT00305175). These data document that continuous in vitro exposure of K562 cells to hydroxyurea leads to tolerant cell lines that feature substantial changes in gene expression. Altered expression of certain genes present in erythroid cells including RRM2 and membrane transporters represent compensatory changes in response to hydroxyurea exposure, and may help explain the variability in hydroxyurea MTD observed among patients with SCA.