Protein kinase A activates protein phosphatase 2A by phosphorylation of the B56delta subunit

Abstract

Our previous studies of DARPP-32 in striatal slices have shown that activation of D1 receptors leads to cAMP-dependent dephosphorylation of Thr-75, the Cdk5 site in DARPP-32. In the current study, we have elucidated a mechanism whereby protein phosphatase 2A (PP2A) is activated by a cAMP/PKA-dependent pathway, leading to dephosphorylation of Thr-75. PP2A consists of a catalytic C subunit that associates with the scaffolding A subunit and a variety of B subunits. We have found that the A/C subunits of PP2A, in association with the B56delta (or PPP2R5D) regulatory subunit, is an active DARPP-32 phosphatase. The B56delta subunit expressed in HEK293 cells forms a heterotrimeric assembly that catalyzes PKA-mediated dephosphorylation at Thr-75 in DARPP-32 (also cotransfected into HEK293 cells). The B56delta subunit is phosphorylated by PKA, and this increases the overall activity of PP2A in vitro and in vivo. Among four PKA-phosphorylation sites identified in B56delta in vitro, Ser-566 was found to be critical for the regulation of PP2A activity. Moreover, Ser-566 was phosphorylated by PKA in response to activation of D1 receptors in striatal slices. Based on these studies, we propose that the B56delta/A/C PP2A complex regulates the dephosphorylation of DARPP-32 at Thr-75, thereby helping coordinate the efficacy of dopaminergic neurotransmission in striatal neurons. Moreover, stimulation of protein phosphatase activity by this mechanism may represent an important signaling pathway regulated by cAMP in neurons and other types of cell.

Fig. 1.

B56δ is phosphorylated in vitro and in intact cells by PKA. The schematic shows the position of sites in rat B56δ phosphorylated in vitro by PKA. The peptide sequences illustrated were used to generate phospho-specific antibodies specific for each site (see SI Fig. 8). (a) The wild-type (wt) and S55/68/566A and S56/68/81/566A mutant forms of the B56δ subunit were expressed and purified from Sf9 insect cells by using baculovirus. Proteins (10 ng per lane) were incubated with PKA and [³²P]ATP for 3 h at 30°C, and reactions were analyzed by SDS/PAGE (4–20% acrylamide), staining with Coomassie blue and autoradiography. (b) Flag-B56δ subunit was transfected into HEK293 cells, cells were treated without or with dibutyryl cAMP (dbcAMP, 100 μM) for 30 min, and B56δ subunit was immunoprecipitated with agarose-conjugated anti-FLAG antibody. The phosphorylation of each site (as indicated) was analyzed with its respective phospho-specific antibody. Total B56δ was analyzed by immunoblotting using anti-Flag antibody.

Fig. 2.

Exogenously expressed B56δ subunit regulates the dephosphorylation of DARPP-32 in HEK293 cells. (a) Cells were transfected with Myc-DARPP-32 without or with B56δ or Bα subunit. Cells were treated with forskolin (10 μM) for 30 min, and the phosphorylation at Thr-34 or Thr-75 of DARPP-32 was analyzed by immunoblotting using phospho-specific antibodies corresponding to each site. Total DARPP-32 expression was analyzed by using anti-Myc antibody. The expression levels of B56δ or Bα were very similar, and similar amounts of heterotrimeric complex were formed by both subunits (data not shown). (b) The phosphorylation levels of Thr-34 (Left) and Thr-75 (Right) were normalized to values obtained for untreated cells expressing only DARPP-32. Data represent means ± SEM (n = 3) ∗, P < 0.01 compared with untreated, DARPP-32-only cells, Student's t test.

Fig. 3.

Phosphorylation of B56δ subunit regulates PP2A-dependent dephosphorylation of DARPP-32. HEK293 cells were cotransfected with Myc-tagged DARPP-32 and either vector, wild-type (wt) B56δ, or mutant B56δ in which Ser-53, 68, 81, and 566 were mutated to Ala (B56S53/68/81/566A). Cells were treated with forskolin (10 μM) for the indicated times, and the phosphorylation and total levels of DARPP-32 were analyzed by immunoblotting (data not shown). The phosphorylation levels of Thr-34 (a) and Thr-75 (b) were normalized to values obtained at zero time. Data represent means ± SEM (n = 3) ∗, P < 0.01 compared with untreated, DARPP-32-only cells, Student's t test.

Fig. 4.

Phosphorylation of B56δ by PKA activates PP2A. (a) HEK293 cells were transfected with either wild-type or mutant Flag-B56δ (as indicated) and cells incubated without (vehicle, DMSO, filled bars) or with forskolin (10 μM, open bars) for 30 min. Cells were lysed and B56δ immunoprecipitated by using agarose-conjugated anti-FLAG antibody. Recombinant DARPP-32 phosphorylated at either Thr-34 (Left) or Thr-75 (Right) were used as substrates for the immunoprecipitated PP2A. The results are expressed as the percentage of total ³²P released from either substrate. Data represent means ± SEM (n = 3) ∗, P < 0.01 compared with untreated control, Student's t test. (b) HEK293 cells were transfected as in a in the absence of any cell treatment. The B56δ subunit was immunoprecipitated with anti-Flag antibody, and samples were incubated without (filled bars) or with (open bars) PKA in vitro. PP2A activity was measured as in a. Data represent means ± SEM (n = 3) ∗, P < 0.01 compared with unphosphorylated PP2A, Student's t test.

Fig. 5.

Phosphorylation of Ser-566 by PKA regulates PP2A activity. HEK293 cells were transfected with wild-type B56δ or mutant B56δ in which individual phosphorylation sites were mutated (as indicated). The B56δ subunit was then immunoprecipitated with anti-Flag antibody, and samples were incubated without or with PKA in vitro. PP2A activity was measured by using DARPP-32 phosphorylated at Thr-34 (a) or Thr-75 (b). Data are expressed as the relative activation of PP2A activity in the absence or presence of phosphorylation by PKA. Data represent means ± SEM (n = 3).

Fig. 6.

B56δ is phosphorylated at Ser-566 by PKA in striatal slices in response to activation of D1 dopamine receptors. Mouse striatal slices were incubated with SKF38393 (10 μM) for various times (as indicated), and phosphorylation of Ser-566 in B56δ was measured by immunoblotting (see Inset for representative experiment). Ser-566 phosphorylation was normalized to total B56δ levels, and then values were normalized to that obtained at zero time in each experiment. Results shown represent the average from three experiments.

Fig. 7.

Expression of a nonphosphorylatable form of B56δ in mouse striatum attenuates the cAMP-dependent dephosphorylation of Thr-75 of DARPP-32. (a) By using AAV-mediated infection, wild-type or mutant Flag-B56δ (B56δS53/68/81/566A) was expressed in mouse striatum after stereotaxic injection. The control used empty AAV. After 14 days, striatal slices were prepared and incubated without or with SKF81297 (10 μM for 5 min), and the phosphorylation and total levels of DARPP-32 were analyzed by immunoblotting as described above. (b) The phosphorylation levels of Thr-34 (Left) and Thr-75 (Right) were normalized to values obtained for untreated slices. Data represent means ± SEM (n = 3); ∗, P < 0.01 compared with untreated samples from control striatal slices, Student's t test. ∗∗, P < 0.01 comparing B56δ and B56δS53/68/81/566A in the presence of SKF81297.