The protein kinase A anchoring protein mAKAP coordinates two integrated cAMP effector pathways

Abstract

Cyclic adenosine 3', 5'-monophosphate (cAMP) is a ubiquitous mediator of intracellular signalling events. It acts principally through stimulation of cAMP-dependent protein kinases (PKAs) but also activates certain ion channels and guanine nucleotide exchange factors (Epacs). Metabolism of cAMP is catalysed by phosphodiesterases (PDEs). Here we identify a cAMP-responsive signalling complex maintained by the muscle-specific A-kinase anchoring protein (mAKAP) that includes PKA, PDE4D3 and Epac1. These intermolecular interactions facilitate the dissemination of distinct cAMP signals through each effector protein. Anchored PKA stimulates PDE4D3 to reduce local cAMP concentrations, whereas an mAKAP-associated ERK5 kinase module suppresses PDE4D3. PDE4D3 also functions as an adaptor protein that recruits Epac1, an exchange factor for the small GTPase Rap1, to enable cAMP-dependent attenuation of ERK5. Pharmacological and molecular manipulations of the mAKAP complex show that anchored ERK5 can induce cardiomyocyte hypertrophy. Thus, two coupled cAMP-dependent feedback loops are coordinated within the context of the mAKAP complex, suggesting that local control of cAMP signalling by AKAP proteins is more intricate than previously appreciated.

Figure 1. Bi-directional control of the mAKAP associated PDE4D3 activity

A) Diagram depicting the modular composition and action of AKAR2. B) Diagram of AKAR2-PKA. Bottom) Pseudo-coloured images of FRET changes in HeLa cells stimulated with cAMP over a 15 min time course. C) Diagram of AKAR2-PKA-PDE. Bottom) Pseudo-coloured images of FRET changes in HeLa cells stimulated with cAMP over a 15 min time course. D) Amalgamated FRET measurements for AKAR-PKA (Red, n=10), AKAR-PKA-PDE (Green, n=13), AKAR-PKA + H89 (Blue, n=10) and AKAR-PKA-PDE + H89 (Black, n=7) for 15 minutes after cAMP stimulation with forskolin (arrow). E) Amalgamated FRET traces (n=10) using the AKAR-PKA-PDE reporter after application of the PDE3 inhibitor milrinone (0–12 min) and the PDE4 inhibitor rolipram (24–36 min). Stimulation with cAMP was at times 0 and 24 min (arrows). F) FRET measurements from HeLa cells expressing the AKAR-PKA-PDE reporter in the presence (black, n=11) or absence (green, n=7) of a dominant active MEK5 for 15 minutes after cAMP stimulation with forskolin (arrow). G) Phosphodiesterase activity (n=4, error bars show S.E.M.) in mAKAP immune complexes isolated from RNVs. Treatment conditions are indicated above each column. H) ERK activity in the mAKAP complex from heart extracts (n=3, error bars show S.E.M.). Treatments with kinase inhibitors are indicated. I) Co-precipitated ERK was detected by immunoblot using (top) pan-ERK, (middle) ERK5 specific and (bottom) MEK5 specific antibodies. P values <0.01 (**) are indicated relative to control (black) and sample (red).

Figure 2. Epac1 suppresses mAKAP associated ERK5 activity

A) Serum stimulated mAKAP associated ERK5 activity (n=3, error bars show S.E.M.) in parallel cultures pretreated with forskolin to elevate cAMP (columns 2&3) or in the presence of the PKA inhibitors H89, KT5720, or Rp-cAMPs (columns 4–6). The amount of ERK5 was detected in each sample. B) Immunoblot detection of Epac1 in mAKAP immune complexes from rat heart extracts. C–E) Fluorescent staining of hypertrophic RNV with antibody for Epac1 (C) and Alexa 568 phalliodin for the actin cytoskeleton (D). Composite image (E) shows the distribution of Epac1 (green) and actin (red, scale bar =20μm). F) The mAKAP complex was immunoprecipitated from cultured RNV following serum stimulation to activate ERK. Parallel cultures were pretreated with either the Epac-selective activator 007 or KT5720 prior to serum stimulation (n=3, error bars show S.E.M.). G & H) Serum stimulated mAKAP associated ERK activity in cells expressing of constitutively active RapGAP (J, n=4, error bars show S.E.M.) or control β-galactosidase (K, n=3, error bars show S.E.M.). Stimulation of intracellular cAMP or treatment with the kinase inhibitor KT5720 is indicated. P values <0.01 (**) are indicated relative to control (black) and sample (red).

Figure 3. The mAKAP complex facilitates cytokine induced cardiac hypertrophy

A–C) Leukemia inhibitory factor (LIF) induced changes in RNV size. The outline of RNV from control (A) and LIF treated (B) samples are presented, scale bar =20μm. C) Quantitation of cell size (μm²) in control (1) and LIF treated RNV (2–5). Pharmacological manipulation of ERK5 (PD98095), Epac1 (007) or PDE4 (rolipram) activity is indicated (error bars show S.E.M.). The number of experiments is indicated above each column. D) Expression of mAKAP was suppressed by RNA interference. Quantitation of LIF induced hypertrophy (cell size +LIF/−LIF) in control (1), mAKAP silenced (2) and RNV rescued with an mAKAP form resistant to the shRNA (3, error bars show S.E.M.). E) Displacement of mAKAP from the nuclear membrane was achieved by overexpression of the mAKAP targeting domain fragment using the TET OFF inducible promoter. Quantitation of cell size (μm²) from control (1), from LIF-stimulated RNV controls (2, TET promoter OFF), and from LIF-stimulated cells expressing the mAKAP 585–1286 fragment (3, TET promoter ON, error bars show S.E.M.). F–I) Schematics depicting the major findings of this study. F) ERK5 phosphorylation of PDE4D3 shuts down cAMP metabolism. G) PKA phosphorylation of PDE4D3 enhances cAMP metabolism. H) Activation of Epac mobilizes Rap1 to suppress ERK5 activation, and I) low cAMP represses the Epac mediated block of ERK5 allowing cardiac hypertrophy. P values <0.01 (**) and p values <0.05 (*) are indicated relative to control (black) and sample (red).