Phosphodiesterase type 4 anchoring regulates cAMP signaling to Popeye domain-containing proteins

Fig. 1. POPDC1 co-localises with PDE4. A. Immunofluorescence showing colocalization of DAPI (nucleus, blue), PDE4A4-VSV (green; polyclonal against human PDE4A) and POPDC1-FLAG (red: polyclonal against POPDC1) in transiently transfected HEK293 (Pearson's coefficient = 0.6., n = 4) and NRVMs (polyclonal against rodent PDE4A, (Pearson's coefficient = 0.6, n = 4).). B. Proximity ligation analysis shows that POPDC1 and PDE4A colocalized in transfected HEK293 cells, NRVMs and ARVMs (upper panels). (same antibodies as in A, n = 4). No background was detected in controls where only the secondary antibodies were used (lower panels). Red dots indicate co-localisation of PODC1 and PDE4A5 and green is wheat germ agglutinin staining to show positions of relevant cells. C, D. POPDC1 and PDE4A5 are detected in the membrane fraction after subcellular fractionation of cell lysate isolated from (C) transfected HEK293 cells or (D) NRVMs (blots typical of experiments done n = 3). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 2. PDE4A and POPDC1 bind directly to each other. A. Myc-immunoprecipitation identifying POPDC1 in a complex with PDE4A5-VSV from transiently transfected HEK293 overexpressing PDE4A5-VSV and POPDC1-myc and B. POPDC1-immunopurification from NRVMs blotted for PDE4A5. C and D Recombinant purified POPDC1-GST and PDE4A4-MBP were mixed at equal concentrations prior to co-immunoprecipitations using amylose resin, that binds to the MBP tag (C.) or glutathione-agarose beads (D.). E. Far Western blotting where recombinant purified POPDC1-GST, UBC9-GST and P75-GST proteins were run on SDS-PAGE and blotted for GST, to ensure the proteins had been successfully transferred, and for MBP, to ensure there was no non-specific binding. F. Membranes were then overlaid with recombinant purified PDE4A4-MBP and re-probed with MBP to detect any interaction between the ‘bait’ proteins and the overlaid PDE4. G and H. A negative control experiment was carried out in the same manner but utilizing recombinant purified RHE PfPdx1-GST protein as it has not been shown to interact with PDE4A4. All blots in Fig. 2 are a representative of experiments done n = 3.

Fig. 3. Mapping the PDE4 docking region on POPDC1. A. Peptide arrays encompassing the full sequence of the POPDC1 protein were overlain with lysate from HEK293 cells over-expressing PDE4A4-VSV or PDE4D9-VSV. Each spot contains an immobilized 25mer sequence derived from the POPDC1 sequence. Dark spots indicate peptides that interacted with PDE4 (blotted for VSV tag). B. The initial binding site identified in A. (POPDC1 aa154 – aa178) was further interrogated by alanine scanning to identify single amino acids required for the interaction, conditions as in 3A. C. A motif scan array was constructed to evaluate the importance of the RLSILLK motif discovered in B. This was overlaid with PDE4A4-VSV, conditions as in 3A. D. To confirm whether the binding motif RLSILLK was important for the PDE4 interaction, an N-terminal truncation or E. a C-terminal truncation peptide array was analysed, conditions as in 3A. For all peptide array experiments negative control experiment was performed using mock transfected HEK293 cell lysate.

Fig. 4. POPDC1 binds to a site within the PDE4 UCR1 domain. A. A peptide array of 25mers shifted by 5 amino acids covering the full PDE4A4 sequence was overlaid with purified recombinant protein consisting of the Popeye domain of POPDC1 tagged with GST. A binding region was identified in the UCR1 domain of PDE4A and the modular structure of PDE4A is shown. A mock control was treated under same conditions but with BSA added instead of GST-Popeye domain. (n = 1) B. The Popeye binding domain in UCR1 contains 6 divergent regions (termed motifs 1–6 in red) when sequences from PDE4 A,B,C and are compared. Divergent amino acids are identified in bold and underlined. C. An alanine scanning array of the popeye binding domain was overlaid with purified recombinant protein consisting of the Popeye domain of POPDC1 tagged with GST. Amino acids that are alanised (or changed to Asp if alanine) are shown in red. Control array (Mock) was treated in same manner but the purified Popeye domain was replaced with BSA. D. A C-terminal truncation array was overlaid with purified recombinant protein consisting of the Popeye domain of POPDC1 tagged with GST. Control array (Mock) was treated in same manner but…

Fig. 5. POPDC1 interacts preferentially with the PDE4A sub-family. A. Co-immunoprecipitations were done using HEK293 cell lysate transiently transfected with POPDC1-myc and PDE4A4-VSV, PDE4B1-VSV or PDE4D7-VSV. Myc-agarose resin was used to precipitate POPDC1 and any interacting PDE4 was detected using VSV tag. Western blotting for both myc confirmed pull down of POPDC and VSV confirming successful transfection of PDE4 isoforms. Negative control immunoprecipitation experiments were performed using cell lysate from mock transfected HEK293. B. Specificity of the POPDC1 interaction for endogenous PDE4 isoforms was examined in NRVMs using PanPDE4A, PDE4B and PDE4D antibodies by PLA analysis. Red dots indicate co-localisation. C. Images were quantified using ImageJ. The intensity of the red (PLA) signal was taken for each cell by drawing around each individual cell using the wheat germ agglutinin as the cell boundary. Data represents an n of 3 with each n consisting of 15–20 cells. Significance was evaluated using a one-way ANOVA with Tukeys post-hoc analysis, ****p = 0.0001, compared to PDE4A control. (For interpretation of the references to colour in this figure legend, the reader is…

Fig. 6. A cell penetrating disruptor peptide dissociates the POPDC1-PDE4 complex. A. HEK293 cells transiently transfected with POPDC1-myc and PDE4A5-VSV were treated with 10 μM scrambled or disruptor peptide. Co-immunoprecipitations of POPDC1 and PDE4A5 performed using Myc-conjugated resin was used to evaluate ability of scrambled or disruptor peptide to compete with PDE4-POPDC1 interaction. The figure is a representative example chosen from three replicate experiments. B. HEK293 cells transiently transfected with POPDC1-myc and PDE4A5-VSV were treated with 10 μM scrambled peptide or disruptor peptide prior to fixation. After PLA staining of HEK293 cells, images were analysed using ImageJ. C. Significance was evaluated using a one-way ANOVA with Tukeys post-hoc analysis, ****p = 0.0001, compared to untreated and scrambled peptide controls. D. The experiment was repeated in NRVM staining endogenous proteins. E. After PLA staining of NRVMs, images were analysed in ImageJ for quantification of PLA signal per condition. The intensity of the red (PLA) signal was taken for each cell by drawing around each individual cell using the wheat germ agglutinin as the cell boundary. Data represents an…

Fig. 7. Disruption of the POPDC1-PDE4 complex reduces TREK1 binding. A. HEK293 cells stably expressing PDE4A4 were transiently transfected with POPDC-CFP and TREK1-YFP. After treatment with 25 μM Forskolin or 10 μM Rolipram, static measurements were taken every 5 s for 300 s per cell. Test data was normalised to mean data from an identical but untreated control group of cells. B. FRET ratio calulated as mean ± SEM, control n = 9 cells, forskolin n = 15 cells, rolipram n = 15 cells per experiment (n = 3). Significance was evaluated using a one-way ANOVA with Tukeys post-hoc analysis, ** p ≤0.005, ***p = 0.001, compared to control C. Calculation of FRET ratio after treatment with 10 μM scrambled peptide or 10 μM disruptor peptide. Static measurements were taken every 5 s for 300 s. Background fluorescence was subtracted from mean intensities and are plotted as mean subtracted intensities. Results represented as mean ± SEM, untreated n = 11 cells, scrambled peptide n = 11 cells, disruptor peptide n = 20 cells per experiment (n = 3). A one-way ANOVA with Tukey's post-hoc was used to analyze the FRET ratios.

Fig. 8. POPDC1-PDE4 disrupting peptide accelerates spontaneous Ca2+ transient frequency in isolated mouse SAN (A) Examples of spontaneous Ca2+ fluctuations recorded by 2-D confocal imaging in Fluo-4-loaded individual SAN cells embedded within the intact SAN. The SAN was incubated for 1 h with 5 μM disrupting peptide prior to Ca2+ measurements in normal Tyrode solution (Control) and in the presence of isoproterenol (Iso, 10 nM). (B) Representative traces of spontaneous Ca2+ transients in SAN cells incubated with 5 μM scrambled peptide (left) or 5 μM disrupting peptide (right) in control and Iso 10 nM conditions. Fluorescence traces are expressed as F/F0, where F is the fluorescence signal and F0 the diastolic fluorescence. (C) Average cycle length (time between two consecutive spontaneous Ca2+ peaks) in 6 SAN incubated with 5 μM scrambled peptide and 6 SAN incubated with 5 μM disrupting peptide in control and Iso 10 nM. Bars are mean values ± SEM. **: p < 0.01; ****: p < 0.0001; $$$: p < 0.001. Two-way ANOVA followed by a Sidak's post-hoc test.

Supplementary Fig. 1. Expression pattern of PDE4A5 and POPDC1 in adult rabbit ventricular myocytes. Immunofluorescence showing colocalization of DAPI (nucleus, blue), PDE4A5 (green: PDE4A5 specific polyclonal) and POPDC1 (red: POPDC1 polyclonal) in ARVMs. (Pearson's coefficient = 0.21).

Supplementary Fig. 2. Expression pattern of PDE4A5 and TREK-1 in adult rabbit ventricular myocytes. Immunofluorescence showing colocalization of DAPI (nucleus, blue), TREK-1 (green: TREK1 polyclonal) and POPDC1 (red: POPDC1 polyclonal) in ARVMs. (Pearson's coefficient = 0.54).

Supplementary Fig. 3. Relative levels of colocalisation between PDE4 and POPDC1 in ARVMS evaluated by PLA. A. ARVM cells were subjected to PLA following staining with antibodies against POPDC1 and either PDE4A5, pan-PDE4B or pan-PDE4D. B. PLA signals from 3 cells were counted for each treatment and a mean value plotted as PLA signals per cell (n = 1).

Supplementary Fig. 4. PDE4 activity or phosphorylation by PKA is not affected by POPDC1 association. A. Increasing amounts of Purified recombinant POPDC1-GST and PDE4A4-MBP were incubated together in a PDE assay. The activity of PDE4A4 was recorded. Graphs were constructed using GraphPad Prism 6™. Results are displayed in pmol cAMP/ min and are represented as a mean ± SEM, n = 3. B. Purified recombinant POPDC1-GST (purple) or GST (blue) were incubated with PDE4A4-MBP in a PDE assay containing increasing concentrations of Rolipram (blue). The activity of PDE4A4 was recorded and graphs were constructed using GraphPad Prism 6™. Results are represented as a mean ± SEM, n = 3. C. Equal molar concentrations of POPDC1-GST and PDE4A4-MBP were incubated with PKA catalytic subunit. Samples were separated by SDS-PAGE and probed via western-blotting for indicated proteins. D. The gel depicted in C. was blotted for phosphorylated-PKA substrates and a phospho-PDE4 UCR1 to detect changes in the level of PDE4 phosphorylation. E. Intensities from phospho-PKA substrate blots were evaluated using ImageJ and normalised to controls. n of 3, mean ± SEM. F. Intensities from phospho-UCR1 blots were…