Mechanistic Paradigm for Combination Signal Transduction Inhibitor Therapy in Zap70+ CLL Involves Bicompartmental Functions of EZH2.

Abstract

We have characterized a tonically active B cell receptor signaling module of zap70+ve CLL cells that contributes to a high-risk disease phenotype associated with rapid tumor doubling time and cellular bcl-2 overexpression. In CLL cells, syk directs tyrosine phosphorylation of certain adaptors including cbl and crkL and this is also accompanied by tyrosine phosphorylation of the interdomain-B loop of zap70 to affect its own interaction with adaptors and vav1. Vav1 along with zap70 allows JNK activation and the formation of heightened c-jun/AP-1 involved in CLL cell bcl-2 transcription. However, CLL pathogenesis is equally dependent upon ezh2/PRC2-mediated 5′-methylation of CpG islands in several tumor suppressor genes such as PTPRO (a cell-intrinsic syk inhibitor) and DAPK1. In addition, ezh2 functions not only in the nucleus for gene repression, but in cytoplasm for stabilization/perpetuation of a zap70-vav1 signal complex upstream of JNK1 in lymphocytes (IH Su et al., Cell 121:425, 2005). We investigated the possible participation of ezh2 in events relevant to the outcome of syk inhibition as a primary therapy for CLL. We used R406/R788, a potent, selective syk inhibitor now in clinical trial (Rigel Pharmaceuticals). We found that R406 treatment strongly inhibited CLL viable cell recovery from culture, and this was associated with inhibition of syk (PY525/catalytic domain) phosphorylation as well as loss of zap70 PY319 (interdomain) phosphorylation with an IC50 of 1 uM R406, associated with downregulation of phospho-c-jun (S73) and of bcl-2. Failure of R406 to inhibit lyn autophosphorylation suggests strongly that syk is the upstream kinase for zap70 in this regulatory pathway. A survey of ezh2 expression showed content of ezh2 correlated with p(319)-zap70 intensity in different CLL cells. R406 caused release of syk substrate cbl from the JNK1 signalosome of R406-treated cells, but ezh2 remained tightly bound to vav1-zap70-JNK1 as demonstrated by immunoprecipitation. To overcome the bi-compartmental functions of ezh2, we used HDAC inhibitor, SBHA (suberoyl hydroxamic acid); a class of compound reported to deplete ezh2 from leukemia cells (W Fiskus et al. Mol Canc. Ther. 5: 3096, 2006). In 3 different experiments, SBHA alone, 1 uM, had little inhibitory activity for CLL cell viable cell recovery, however, when combined with R406 (also at 1uM) SBHA caused massive cell death (>80%) only attainable with R406 alone at 2.5X dose. Mechanistic aspects of the reversal of tumor suppressor gene repression and of inactivation of the JNK1 signalosome by SBHA are under scrutiny to fully explain its cooperativity with R406, focused especially on the duality of roles performed by ezh2, and the depletion of ezh2 affected by HDAC inhibition.