Figure 1.
Identification of PDZ2 of SAP97 as a binding partner to the C-tail of the ß1-AR.
A, increasing concentrations of GST-tagged ß1-ARC-tail or GST were screened for interactions with SAP97 by GST pulldown using lysates from KEK-293 cell expressing SAP97-YFP. GST pull-downs (IP) were immunoblotted (IB) for SAP97 using a SAP97 antibody from Enzo Life Sciences. The optical density of each band was plotted against µM of ß1-AR protein to estimate the concentration of ß1-AR protein at 50% of the maximal optical density. These values from n = 3 experiments were used to calculate the apparent EC50 of their interaction. Input represents ∼4% of the lysate. B, schematic diagram of the protein-protein interacting domains of SAP97. C, 10 µl or 20 µl of a 1 µg/ml stock solution of GST-tagged PDZ1, PDZ2 or PDZ3 of SAP97 were slot-blotted onto nitrocellulose membranes. The membranes were hybridized with 16 µg/ml (0.34 µM) of MBP-ß1-AR(425–477) fusion protein. The binding of the ß1-AR to each PDZ was detected with the sc-567 anti-ß1-AR antibody (Santa Cruz). D, Western blot (IB) of purified GST or GST-PDZ1, -PDZ2 or -PDZ3 fusions. E, Equal amounts of protein lysates from HEK-293 cells expressing either the empty pcDNA 3.1 or FLAG WT ß1-AR were mixed with 10 µg of GST or 12.5 µg of the indicated GST-tagged PDZ in a total volume of 1 ml (0.38 µM of GST or GST-PDZ). Then ∼4% input lysates (lanes 1, 2) or GST pull-downs (IP) from cells that expressed the WT ß1-AR (lanes 3–6) or the empty vector (lanes 7–10) were subjected to Western blotting (IB) and probed with the anti-FLAG antibody.
Figure 2.
Effects of SAP97 knockdown and rescue on recycling and resensitization of the human ß1-AR.
A, in images a–n, HEK-293 cells stably expressing the FLAG-tagged WT ß1-AR and either the scrambled shRNA-EGFP (a–g), or hSAP97 shRNA-EGFP (h–n) were used. In images o-b’, HEK-293 cells stably expressing the FLAG-tagged WT ß1-AR and hSAP97 shRNA were transfected with the empty pIRES-EGFP vector (o–u) or rSAP97 in pIRES-EGFPII (v-b’). Rat SAP97 constructs (denoted by an asterisk) are resistant to hSAP97 shRNA because of sequence mismatches in the region targeted by the human SAP97 shRNA. Cells on glass slides were prelabeled for 1 h with Cy3-anti-FLAG antibody and fixed (images a, h, o, v). The rest of the slides were exposed to 10 µM isoproterenol for 30 min (b, i, p, w) then acid washed and fixed (c, j, q, x). The rest of the slides were subjected to recycling conditions for the indicated time period and then fixed. Slides incubated with alprenolol for 1 h were exposed to acid wash and then fixed (g, n, u, b’). The distribution of fluorescent pixels was obtained using confocal microscopy and the colors shown are pseudo colors. B, pixels inside a 300-nm boundary in isoproterenol/acid-washed cells (c, j, q and x) were set arbitrarily to 100% to indicate 100% internalization and the ratios in alprenolol-treated cells were calculated and expressed as % for each time period. The ratios from 20 independent images for each condition were calculated and expressed as mean ± S.E. for each time period and then compared among the three different groups by one-way ANOVA with Newman-Keuls post-tests. Statistical results are expressed as (*), (**), and (***) p<0.05, p<0.01, and p<0.001, respectively. Each scale bar represents 5 µm. C, comparison of adenylyl cyclase activities in response to desensitization by isoproterenol and resensitization in HEK-293 cells. For the first two sets of experiments, cells stably expressing the WT ß1-AR with either the scrambled or the hSAP97 shRNA were used. For the third set, cells expressing the ß1-AR, SAP97 shRNA and rSAP97 in pIRES-EGFPII were used. Following the desensitization/resensitization protocol, cell membranes were prepared and used to measure adenylyl cyclase activities for each condition. These experiments were repeated n = 3–5 times each in triplicate and were plotted as mean ± S.E. and then compared among different groups by one-way ANOVA with Newman-Keuls post-tests and expressed as described above. NS, indicates non significant difference among the compared groups.
Figure 3.
Mapping of the role of SAP97 PDZ domains in binding to and in recycling of the ß1-AR.
A, schematic diagram of SAP97 and deletion constructs used in panels B and C. B, cells stably expressing FLAG WT ß1-AR and hSAP97 shRNA were transiently transfected with rSAP97 or its deletion constructs in pIRES-EGFP II that are shown in panel 3A. The internalization and recycling assays were carried out as described in the legend of Fig. 2. Each scale bar represents 5 µm. C, HEK-293 cells expressing hSAP97 shRNA and FLAG-tagged ß1-AR were transfected with pIRES-EGFP harboring SAP97 or its various PDZ deletion mutants shown in panel A above. Equal amounts of cell lysates (input) were incubated with ∼8 µL of anti-FLAG M2 IgG resin overnight. Then the beads were washed and eluted into 50 µl of 2X-Laemmli sample buffer. The top panel shows Western blot analysis of 10% of each FLAG IP that were probed with anti-human ß1-AR IgG (IB) to index ß1-AR IP’s. The middle panel shows Western blot analysis of 80% of each FLAG IP that were probed with anti-SAP97 IgG (IB) to index SAP97 Co-IPs. To normalize for input protein levels, ∼4% of each cell lysate were subjected to Western blotting (IB) and probed with the anti-pan MAP kinase antibody.
Figure 4.
SAP97-AKAP79 interactions are involved in recycling of the ß1-AR.
A, schematic diagram of SAP97 L27/U1 and U5 deletion constructs used in panels B and C. B, recycling of the ß1-AR in HEK-293 cells stably expressing hSAP97 shRNA was rescued by ΔL27-I3I5SAP97 but not by ΔL27 I2I5 ß-isoform of SAP97. Each scale bar represents 5 µm. C–D, ΔL27-I3I5 SAP97 or ΔL27-I2I5 SAP97 in pIRES-EGFP were transfected into the FLAG-tagged ß1-AR and hSAP97 shRNA double-stable cell line. In C, FLAG immunoprecipitations were probed with an anti-pan PDZ antibody to detect SAP97. In D, FLAG immunoprecipitations were probed with anti-AKAP79 antibody (IB). Input represented ∼4% of total cell lysate from each condition.
Figure 5.
PDZ2 interference with the binding between SAP97 and the ß1-AR.
A, in vitro interference of PDZ2 in the binding between SAP97 and the ß1-AR. Equal amounts of extract prepared from cells co-expressing FLAG ß1-AR and SAP97-YFP was mixed with 2.5 µg purified GST or 3.2 µg of GST-fusions of individual SAP97 PDZs in a total volume of 1 ml (0.1 µM). FLAG IP’s were subjected to Western blotting (IB) and probed with anti-SAP97. B, determining the affinity of the interaction between PDZ2 and the ß1-AR. Membranes from HEK-293 cells expressing 1.2 pmoles of ß1-AR per mg protein were solubilized and mixed with the indicated concentrations of GST of GST-PDZ2. GST-pull downs were probed with the anti-ß1-AR antibody and optical densities were used to calculate the EC50 for their interaction. C, in vivo interference of PDZ2 with ß1-AR binding to SAP97. HEK-293 cells expressing the FLAG ß1-AR and YFP-SAP97 were transfected with the indicated SAP97 PDZ in pIRES-EGFP. FLAG IP’s or ∼4% of input lysate were subjected to Western blotting (IB) and probed with anti-SAP97 antibody (n = 3).
Figure 6.
Lysines 323 and 326 in PDZ2 influence SAP97-mediated recycling and binding to the ß1-AR.
A, ribbon diagram of the SAP97 PDZ2 structure with ESKV peptide bound. In PDZ2 domain helixes and strands are named according to Doyle et al [45] and shown in different colors. ESKV peptide is shown in magenta. B, ligand binding interactions between the C-terminal sequence (ESKV) of the human ß1-AR and SAP97 PDZ2 domain. Amino acids representing different strands and helixes in PDZ2 domain structure are shown in different colors: ßC strand, green; ßB strand, light blue; ßA strand, dark blue; alpha/B helix, light red. ESKV sequence is shown in magenta, oxygen atoms are depicted in red, nitrogen atoms are in navy blue and hydrogen bonds are indicated as dashed blue lines. Distance between the alpha-carbons of amino acids is shown in Å. C, cartoon of the K323A/K326A PDZ2 mutant that was used in Panel D. D, effect of mutating K323 and K326 to Ala on SAP97-mediated regulation of ß1-AR recycling. HEK293 cells stably expressing the FLAG-tagged WT ß1-AR and hSAP97 shRNA were transfected with K323A/K326A SAP97-YFP and the internalization and recycling of the ß1-AR were analyzed as described in Fig. 2. Each scale bar represents 5 µm. E, Cells stably expressing FLAG ß1-AR and hSAP97 shRNA were transfected with SAP97-YFP or K323A/K326A SAP97-YFP. FLAG IP’s from cell extracts were subjected to Western blotting (IB) and probed with anti-SAP97 antibody or with anti-pan MAP kinase antibody.
Figure 7.
Effect of the isolated PDZs of SAP97 on trafficking and resensitization of the ß1-AR.
A, B, each isolated PDZ of SAP97 in pIRES-EGFP was co-transfected into cells stably expressing the FLAG ß1-AR followed by determining their effect on internalization and recycling of the ß1-AR by confocal microscopy. Each scale bar represents 5 µm. C, Effect of overexpressing PDZ2 or K323A/K326A PDZ2 on recycling of the ß1-AR as determined by surface-biotinylation recycling assays. HEK-293 cells expressing the WT ß1-AR and K323A/K326A PDZ2 in pIRES-EGFP (lanes 1–5) or the WT ß1-AR with WT PDZ2 in pIRES-EGFP (lanes 6–10) were surface biotinylated and one set of plates was removed to determine total biotinylation for each cell line (lanes 1 and 6). The remaining sets of culture dishes were exposed to isoproterenol for 30 min at 37°C to induce ß1-AR internalization, followed by removal of isoproterenol and cleavage of the remaining surface biotin (1st cleavage) with glutathione (lanes 2–5 and 7–10). One set of plates was extracted to determine the amount of internalized biotinylated ß1-AR (lanes 2 and 7). The remaining cultures were switched to warm culture medium supplemented with 10 µM ß-adrenergic receptor antagonist alprenolol and then returned to 37°C to allow the internalized ß1-AR to recycle for 15, 30, and 60 min. After each time period, the cells were recleaved (2nd cleavage) to ensure cleavage of any newly appearing (recycled) surface biotin. After the second cleavage, the cells were solubilized in RIPA buffer, and equal amounts of clarified lysate protein were mixed with 50 µl of bovine serum albumin-blocked ultralink-neutra avidin beads at 4°C overnight. The beads were eluted and the eluates were probed by Western blotting using the anti-FLAG antibody. D, PDZ2-GST or PDZ2 (K323A/K326A)-GST were mixed with lysates prepared from cells expressing the wild-type ß1-AR, followed by analyzing the GST-pull downs for co-IP of the ß1-AR. E, effect of PDZ2 or K323A/K326A PDZ2 on desensitization and resensitization of ß1-AR-regualted adenylyl cyclase activity. Cells stably expressing the WT ß1-AR with either the PDZ2-EGFP or K323A/K326A PDZ2-GFP were subjected to desensitization and resensitization as described in the legend of Fig. 2. The activities of adenylyl cyclase in n = 5 experiments, each in triplicate were plotted as mean ± S.E. and then compared among different groups by one-way ANOVA with Newman-Keuls post-hoc tests. Statistical results are expressed as (*) and (**) to indicate p<0.05, p<0.01, respectively. NS, indicates non significant difference among the compared groups.
Figure 8.
Characterization of molecular interactions between the ß1-AR and SAP97 and AKAP79/150 in cardiac and neuronal tissue.
A, mouse neonatal cardiac myocytes were infected with an adenovirus harboring HA-tagged mouse ß1-AR. Then input lysates were immunoprecipitated with anti-HA resin (IP). The resin was eluted and equal amounts of eluate were subjected to Western blotting (IB) and probed with antibodies to AKAP150, SAP97 and ß1-AR. B, lysates of rat hippocampal membranes were incubated with 1∶200 dilution of anti-SAP97 IgG (Enzo Life sciences) or mouse IgG for 16 h at 4°C. Then 30 µl of protein G-agarose was added for 2 h, followed by 4 washes of the resin in complete RIPA buffer. The beads were suspended in 2X Laemmli sample buffer with 20 mM dithiothreitol for 45 min at 37°C. The cleared eluates were divided equally and subjected to SDS-PAGE on 8% (GluR1) or 10% (ß1-AR) gels followed by electroblotting to nitrocellulose membranes. These membranes were probed with anti-GluR1 (sc-28799) or ß1-AR (sc-568) antibodies from Santa Cruz Biotechnology. C, diagram representing the binding of SAP97 dimers to the ß1-AR in postsynaptic membranes and to the GluR1 subunit of AMPAR in a postsynaptic membrane specialization commonly categorized as the “perisynaptic membrane”. We hypothesize that activation of the ß1-AR signalosome ultimately activates PKA, which in turn phosphorylates the GluR1 subunit of AMPA receptors on Ser845 to initiate regulated trafficking of AMPA receptors from perisynaptic membranes to postsynaptic membranes [39]–[41].