The PDZ and band 4.1 containing protein Frmpd1 regulates the subcellular location of activator of G-protein signaling 3 and its interaction with G-proteins

Abstract

Activator of G-protein signaling 3 (AGS3) is one of nine mammalian proteins containing one or more G-protein regulatory (GPR) motifs that stabilize the GDP-bound conformation of Galphai. Such proteins have revealed unexpected functional diversity for the "G-switch" in the control of events within the cell independent of the role of heterotrimeric G-proteins as transducers for G-protein-coupled receptors at the cell surface. A key question regarding this class of proteins is what controls their subcellular positioning and interaction with G-proteins. We conducted a series of yeast two-hybrid screens to identify proteins interacting with the tetratricopeptide repeat (TPR) of AGS3, which plays an important role in subcellular positioning of the protein. We report the identification of Frmpd1 (FERM and PDZ domain containing 1) as a regulatory binding partner of AGS3. Frmpd1 binds to the TPR domain of AGS3 and coimmunoprecipitates with AGS3 from cell lysates. Cell fractionation indicated that Frmpd1 stabilizes AGS3 in a membrane fraction. Upon cotransfection of COS7 cells with Frmpd1-GFP and AGS3-mRFP, AGS3-mRFP is observed in regions of the cell cortex and also in membrane extensions or processes where it appears to be colocalized with Frmpd1-GFP based upon the merged fluorescent signals. Frmpd1 knockdown (siRNA) in Cath.a-differentiated neuronal cells decreased the level of endogenous AGS3 in membrane fractions by approximately 50% and enhanced the alpha2-adrenergic receptor-mediated inhibition of forskolin-induced increases in cAMP. The coimmunoprecipitation of Frmpd1 with AGS3 is lost as the amount of Galphai3 in the cell is increased and AGS3 apparently switches its binding partner from Frmpd1 to Galphai3 indicating that the interaction of AGS3 with Frmpd1 and Galphai3 is mutually exclusive. Mechanistically, Frmpd1 may position AGS3 in a membrane environment where it then interacts with Galphai in a regulated manner.

FIGURE 1.

Interaction of Frmpd1 and AGS3 in vitro. A, schematic diagram of Frmpd1 indicating the location and length of the Frmpd1 fragment (Arg⁸⁷¹–Leu¹⁵⁴⁹) isolated in the yeast two-hybrid screen. B, rat brain lysate (1 mg of protein) was incubated with 1 μm GST, GST-LKB1-CT (Asp³³⁰–Gln⁴³⁶), or GST-Frmpd1 (Arg⁸⁷¹–Leu¹⁵⁴⁹) for 1 h at 24 °C and processed by GST pull-down assays, SDS-PAGE, and immunoblotting as described under “Experimental Procedures.” Membrane transfers were first probed with affinity purified AGS3 antibody (PEP32 0.22 μg/ml, left panel). The Input lane contains 1/10 of the lysate volume used for each interaction assay. Data represent at least four independent experiments. C, the continually truncated GST-Frmpd1 proteins (1 μm) were incubated with rat brain lysate and processed with GST pull-down assays as described for panel B. The two upper blots in C were stripped and reprobed with GST antibody (middle panels) to verify protein loading. A Coomassie Blue stain of the GST fusion proteins was included to indicate the quality of the preparation (lower right panel). The location and length of truncated Frmpd1 deletion constructs are indicated in the lower panel. Data shown are representative of at least three independent experiments with similar results using different brain lysates and/or fusion protein preparations.

FIGURE 2.

Coimmunoprecipitation of Frmpd1 and AGS3 from COS-7 cell lysates. COS-7 cells were transfected, harvested, and processed for immunoprecipitation as described under “Experimental Procedures.” Precleared lysates (800 μg of protein in 500 μl total volume) were immunoprecipitated with affinity purified AGS3 antibody (PEP32) and processed for SDS-PAGE and immunoblotting. The Input lanes represent 1/15 of the lysate volume used for immunoprecipitation. The blot was probed with Frmpd1 antibody (0.4 μg/ml) and AGS3 antibody (PEP32, 0.22 μg/ml). The data presented are representative of at least four independent experiments. IP, immunoprecipitation.

FIGURE 3.

Influence of Frmpd1 on fractionation of AGS3. A, COS7 cells were transfected with pcDNA3::Frmpd1, pcDNA3::AGS3, or co-transfected with both constructs, lysed, and processed for fractionation as described under “Experimental Procedures.” Forty μg of cytosol protein (100,000 × g supernatant) and 60 μg of washed membrane protein (100,000 × g pellet, washed with membrane buffer: 50 mm Tris-HCl, pH 7.4, 0.6 mm EDTA, 5 mm MgCl₂) were then processed for SDS-PAGE and immunoblotting. Data represent at least eight independent experiments. B, COS7 cells were transfected as described in A and fractionated as described under “Experimental Procedures” to generate a washed and unwashed membrane pellet. Sixty μg of cytosol proteins, 54 μg of washed membrane proteins, and 60 μg of unwashed membrane proteins were processed for SDS-PAGE and immunoblotting. C, data are presented in the bar graph as the mean ± S.E. (whole cell lysate: AGS3, n = 3; Frmpd1, n = 2; 100,000 × g supernatant and pellet: AGS3, n = 8; Frmpd1, n = 6). The pixel intensity of AGS3 or Frmpd1 in each fraction observed when cells were cotransfected was divided by the pixel intensity of the corresponding immunoreactive species observed in that specific fraction when cells were transfected with AGS3 or Frmpd1 alone and normalized according to actin expression level. Statistical significance was evaluated with the Student's t test (^*, p < 0.05). The following antibodies were used: Frmpd1 (0.4 μg/ml), AGS3 (PEP32, 0.22 μg/ml), Gα_i3 antisera (1:2,000), and actin (0.5 μg/ml).

FIGURE 4.

Influence of Frmpd1 on subcellular localization of AGS3. A, COS7 cells (80–90% confluence) were transfected with plasmids as indicated and processed for fluorescence microscopy as described under “Experimental Procedures.” DNA was stained by 4,6-diamidino-2-phenylindole (DAPI) (1 μg/ml) (blue). The white arrows point to the cell periphery. Images were taken from approximately the middle plane of the cell and are presented at ×63 magnification. The enlarged images are defined by the dashed inset demonstrating co-localization of AGS3-mRFP and Frmpd1-GFP at the cell periphery. Data are representative of three independent experiments. B, COS7 cells transfected with pEGFPN1::AGS3, pcDNA3::Gα_i3 or both constructs were processed for immunocytochemistry as described under “Experimental Procedures.” The patch with enhanced GFP background fluorescence reflects a section of the cell that folded back on itself during preparation. The images shown are representative of at least four independent experiments. Scale bar, 10 μm.

FIGURE 5.

Influence of siRNA-mediated knockdown of Frmpd1 on fractionation of AGS3 and receptor-mediated regulation of cAMP in CAD cells. A, CAD cells (60–70% confluence in 100-mm dish) were transfected with 80 nm Frmpd1 siRNA number 1 plus 80 nm Frmpd1 siRNA number 2 and processed for fractionation as described under “Experimental Procedures.” Forty μg of whole lysate protein and cytosol protein or 100 μg of membrane protein were electrophoresed and membrane transfers were probed with Frmpd1 antibody (0.4 μg/ml), AGS3 antibody (PEP32, 0.22 μg/ml), and Gα_i3 antisera (1:2,000). B, data are presented in the bar graph as the mean ± S.E. (n = 4). The pixel intensity of AGS3 or Frmpd1 in each fraction observed following siRNA-mediated knockdown of Frmpd1 was divided by the pixel intensity of the corresponding immunoreactive species observed in that specific fraction when cells were transfected with control siRNA duplexes and normalized according to actin expression level. The transfection control refers to cells treated with transfection reagent but without siRNA. C, CAD cells were transfected with control siRNA or Frmpd1 siRNA as described above for A. 24 h post-transfection, the cells were replated on 24-well plates and treated with 1 μm forskolin plus or minus the α₂-adrenergic receptor selective agonist UK-14304 (10 nm to 10 μm). Total cellular cAMP was measured as described under “Experimental Procedures.” Basal cAMP: control siRNA, 5.75 ± 0.86 pmol/well, n = 9; Frmpd1 siRNA, 5.71 ± 0.77 pmol/well, n = 9. Forskolin-induced cAMP: control siRNA, 24.12 ± 3.59 pmol/well, n = 9; Frmpd1 siRNA, 21.16 ± 2.46 pmol/well, n = 9. The right panel indicates knockdown of Frmpd1 following Frmpd1 siRNA transfection from a representative experiment. Data are presented as the mean ± S.E. Statistical significance was evaluated with the Student's t test (^*, p < 0.05); n, number of independent experiments.

FIGURE 6.

Interaction of Frmpd1 and Gα_i3 with AGS3 in COS7 cells. A, COS-7 cells were transfected with pcDNA3::Frmpd1 (10 μg) and pcDNA3::AGS3 (3 μg) or cotransfected with pcDNA3::Giα3 (2 or 4 μg). Cell lysates were prepared and processed for immunoprecipitation as described under “Experimental Procedures.” Precleared lysates (800 μg of protein in 500 μl total volume) were immunoprecipitated (IP) with AGS3 antibody (PEP32) and processed for SDS-PAGE and immunoblotting (IB). Membrane transfers were probed with Frmpd1 antibody (0.4 μg/ml), AGS3 antibody (PEP32, 0.22 μg/ml), and Gα_i3 antisera (1:2,000). Data are representative of three experiments using different transfected cell lysates.

FIGURE 7.

Schematic representation of AGS3 positioning within the cell. AGS3 may find its binding partners already positioned at the cell cortex (A) or it may find them in another subcellular compartment (B) and the complex would then move to the cell cortex. The bracketed entity and dashed arrows illustrate a postulated transient ternary complex. Up to four Gα_i subunits may be docked to AGS3 (6, 53) and only two Gα_i subunits are indicated in the schematic for the AGS3-Gα_i complex simply for ease of presentation. Refer to “Results and Discussion” for further discussion.