Abstract
Abstract 1411
Glucocorticoids (GC), such as prednisolone and dexamethasone, are an integral component of the multi-agent treatment of childhood acute lymphoblastic leukemia (ALL). GC-resistance is a significant prognostic indicator of a poor treatment outcome and remains a clinical problem, with the underlying mechanisms still unclear. Mutation or loss of the primary mediator of GC-action, the glucocorticoid receptor (GR), underlies the GC-resistant phenotype in several commonly used leukemic cell lines. However, these events are rare in primary leukemic cells, with relatively few examples in vivo. This suggests that it may be possible to reverse the GC-resistant phenotype pharmacologically. We have used an iTRAQ proteomics approach for hypothesis generation of potential mechanisms for GC-resistance in childhood ALL. To achieve this, we compared a well-characterized GC-sensitive cell line, PreB 697, and a GC-resistant sub-clone (R3F9), both bearing wildtype GR, in a comparative proteomic experiment using 4-channel isobaric tagging for relative and absolute quantification (iTRAQ). A comparison of protein profiles before and after dexamethasone exposure of the two cell lines identified two transcription factors involved in B-cell differentiation, PAX5 and IRF4, to be differentially upregulated in the PreB 697 compared to the R3F9 cell line in response to GC. Experimentally, there was approximately 50% reduction in PAX5 basal protein expression in R3F9 compared to its GC-sensitive parent, a finding which was also evident in four other resistant sub-lines. This was accompanied by a decreased expression of CD19 and CD10, indicative of an increased B-cell maturation state. The reduced PAX5 level in the GC-resistant cell lines was not due to mono-allelic loss or mutation and mRNA levels were not significantly altered, suggestive of a post-transcriptional mechanism for PAX5 protein reduction. Paradoxically, knockdown of PAX5 reversed the GC-resistant phenotype of the R3F9 cell line such that the apoptotic response to dexamethasone was similar to that of the GC-sensitive parent line as measured by Annexin V staining (R3F9: mean 52.22%, SD 12.54%, n=3; PreB 697: mean 67.23%, SD 9.96%, n=3) and cell viability assays. This chemosensitization after PAX5 knockdown was specific to GC, with no difference in cell viability observed in either cell line after exposure to daunorubicin, vincristine or L-asparaginase when compared to negative siRNA or mock controls. This increase in GC-sensitivity was coupled with a significant upregulation of GR and its transcriptional target, GILZ. We also showed an enhanced GC response after PAX5 knockdown in two out of eight primary, diagnostic pre-B lineage ALL patient samples. Thus, in this ALL cell line model, quantitative proteomic analysis revealed increased maturation as a recurrent mechanism underlying GC-resistance and identifies PAX5 as a possible therapeutic target to fully re-sensitise GC-response in childhood ALL.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.