Abstract
Cyclic AMP (cAMP) potentiates glucocorticoid (GC) induced apoptosis in lymphoid cells but the mechanisms associated with these effects are unclear. We previously showed that cAMP inhibits PI3K/AKT activity and that in diffuse large B-cell lymphomas (DLBCL) overexpression of the phosphodiesterase 4B (PDE4B) blocks these effects. Recently, high AKT/mTOR activity was implicated in the GC resistance found in acute lymphoid leukemia (ALL) and rapamycin shown to modulate these effects. Therefore, we hypothesized that cAMP enhances GC responses by down-modulating the AKT/mTOR pathway and that PDE4B, by blunting these inhibitory effects, is at the center of GC resistance (and rapamycin activity) in DLBCL. If this notion is accurate, PDE4B expression may impinge on the same genes and pathways that control GC response in malignant lymphocytes. To test this idea, we used gene set enrichment analysis (GSEA) and found a marked association between gene sets of DLBCLs defined by high or low PDE4B expression and ALLs classified by their resistance or sensitivity to GC-induced apoptosis(p and FDR =.014). To further our investigation, we created PDE4B gain or loss of function models in DLBCL cell lines. We reconstituted PDE4B expression (wild-type [WT] or phosphodiesterase inactive [PI] mutant) in the PDE4B-null DHL6 cell line and constitutively expressed a PDE4B-specific RNAi, which decreased this gene expression 14-fold, in the PDE4B-high cell line Ly3 (Ly3Ri). The functional consequences of these changes were confirmed by measuring the intra-cellular levels of cAMP and further validated in cell proliferation assays. DHL6-PDE4B-WT cells became ∼ 50% more resistant to the inhibitory effects of cAMP (forskolin - 10μM) than their PI counterpart (p<0.01) whereas the Ly3Ri cells were ∼ 30% more sensitive than the Ly3 control cells (p<0.01). Therefore, we were confident that these models could yield robust information regarding the role of PDE4B on the sensitivity of DLBCL to GC and rapamycin. In proliferation assays, Ly3Ri cells showed a ∼ 40% increase in sensitivity to dexamethasone (Dex, .5μM) and rapamycin (5nM) when compared to no-RNAi control cells (p<0.01). In agreement with these data, cells expressing inactive PDE4B were more sensitive to Dex and rapamycin than the PDE4B-WT cells, a difference specially pronounced when in combination with forskolin. To start to characterize the signaling molecules that may account for the PDE4B-dependent sensitization to GC and rapamycin in DLBCL, Ly3Ri cells and their control counterparts were treated for 3 hours with Dex (.5μM), rapamycin (5nM) or vehicle only and their lysates analyzed for the phosphorylated levels of the mTOR targets 4EBP1 (Thr37/46) and S6R (S235/6). A decrease in the phospho-S6R levels was evident with both drugs, particularly rapamycin, irrespective of the PDE4B levels. However, PDE4B RNAi cells showed a more striking decrease in the levels of phosho 4EBP1 upon Dex and rapamycin treatment than control cells. These findings are particularly important because phosphorylation of 4EBP1 controls the activity of eIF4E, a translational regulator that has been implicated in resistance to chemotherapy and rapamycin in murine models of lymphoma. Taken together, our findings implicate the AKT/mTOR pathway on the elusive interplay between cAMP signals and glucorticoid sensitivity. These data also indicate that pharmacological inhibition of PDE4B may improve the effectiveness of rapamycin and GC in lymphoid malignancies.
Author notes
Disclosure: No relevant conflicts of interest to declare.