RAS and SYK Mediate cAMP Inhibition of PI3K/AKT in Mature B-Cells.

In normal and malignant lymphocytes the inhibitory cyclic-AMP (cAMP) signals are mainly controlled by the phosphodiesterase 4B (PDE4B). We previously showed that in diffuse large B-cell lymphoma (DLBCL) cAMP induces apoptosis via down regulation of PI3K/AKT activity independently of the effectors PKA and EPAC. These data implied the existence of a pathway linking cAMP to PI3K. However, the molecules that mediate this interaction are unknown. In B-cells, phosphorylation of PI3K’s regulatory subunit by SYK potently up-regulates its activity which is enhanced by stabilization of the catalytic domain to cell membrane via binding to RAS. Hence, we first asked if cAMP could decrease RAS activity and if that contributed to PI3K inhibition. To address this issue, we reconstituted PDE4B expression (wild-type gene [WT] or a phosphodiesterase inactive [PI] mutant) in PDE4B-null DLBCL cells. Subsequently, the intra-cellular levels of cAMP were increased with Forskolin (10μM) and the levels of activated RAS determined in pull-down assays with the Raf1- Ras Binding Domain-GST fusion protein. In cells lacking functional PDE4B (PI), high levels of cAMP resulted in a marked reduction of the RAS activity whereas it had no effects in the PDE4B-WT cells. After showing that cAMP decreases RAS activity in mature B-cells, we sought to link RAS inhibition to PI3K downregulation. Thus, we created cells expressing the dominant negative RASN17 mutant in a PDE4B-null background; the inhibitory role of these N17 mutants was confirmed in RAS activation assays. We reasoned that if RAS is necessary for cAMP-mediated inhibition of PI3K, its absence (N17 mutant) may block this event. To test this hypothesis we compared the effects of cAMP in PDE4B-WT, -PI or RASN17 cells on the phosphorylation of AKT (S473), a precise surrogate for PI3K activity in this model, and on downstream targets of mTOR, S6 ribosomal protein (S6R - S235/6) and 4EBP1 (Thr37/46). Whereas at baseline there was no significant difference in the phospho levels of these proteins, cAMP completely inhibited their phosphorylation in the PI cells but had a limited or no effect on RASN17 and WT cells, respectively. In agreement with these data, in proliferation assays RASN17 expression rendered PDE4B-null cells ∼ 50% more resistant (p=.001) to cAMP than their RAS intact counterpart. Taken together, these data indicate that RAS transduce most, but probably not all, of the cAMP effects towards PI3K/AKT. Therefore, we next investigated if modulation of the tyrosine kinase SYK could also be involved in the cAMP inhibition of PI3K. Using PDEB-null cells, we found that cAMP decreased the phosphorylation of residues essential for SYK function (Y525/26) to the same extent as the SYK inhibitor Piceatannol. As SYK is a candidate for targeted inhibition in a variety of disorders, this newly found interplay with cAMP/PDE4B may have clinical implications. For this reason, we created a PDE4B loss of function model (stable RNAi) to test the benefits of combining SYK and PDE4B inhibition in DLBCL. In these assays, Piceatannol (5–10μM) was more efficient in inhibiting the proliferation of PDE4B-RNAi cells than their control counterpart (60% vs. 30% inhibition, p=.004). These data suggest that in the absence of PDE4B cAMP signals lower the tonic activity of SYK and increase the effects of SYK inhibitors. Herein, we placed RAS and SYK in the pathway linking cAMP to PI3K and found evidence that combining PDE4 and SYK inhibitors may be beneficial in inflammatory, auto-immune and malignant conditions.