G protein-coupled receptor kinase 5 phosphorylation of hip regulates internalization of the chemokine receptor CXCR4

Abstract

Regulation of the magnitude, duration, and localization of G protein-coupled receptor (GPCR) signaling responses is controlled by desensitization, internalization, and downregulation of the activated receptor. Desensitization is initiated by the phosphorylation of the activated receptor by GPCR kinases (GRKs) and the binding of the adaptor protein arrestin. In addition to phosphorylating activated GPCRs, GRKs have been shown to phosphorylate a variety of additional substrates. An in vitro screen for novel GRK substrates revealed Hsp70 interacting protein (Hip) as a substrate. GRK5, but not GRK2, bound to and stoichiometrically phosphorylated Hip in vitro. The primary binding domain of GRK5 was mapped to residues 303-319 on Hip, while the major site of phosphorylation was identified to be Ser-346. GRK5 also bound to and phosphorylated Hip on Ser-346 in cells. While Hip was previously implicated in chemokine receptor trafficking, we found that the phosphorylation of Ser-346 was required for proper agonist-induced internalization of the chemokine receptor CXCR4. Taken together, Hip has been identified as a novel substrate of GRK5 in vitro and in cells, and phosphorylation of Hip by GRK5 plays a role in modulating CXCR4 internalization.

Figure 1. In vitro phosphorylation of purified Hip by GRK5

(A) Hexa-histidine tagged Hip was expressed in E. coli and purified using Ni⁺ agarose and size exclusion chromatography. One microgram of purified Hip was separated by SDS-PAGE and Coomassie stained. (B) 100 nM of purified GRK2 or GRK5 was incubated with 100 nM of purified Hip for various times at 30°C in a buffer containing radiolabeled ATP. Samples were separated by SDS-PAGE and subjected to autoradiography. (C) The radioactively labeled bands from B were excised and quantitated by scintillation counting to determine the stoichiometry of phosphorylation. Data represent the mean ± S.D. from six separate experiments. (D) Mixed micelles were generated by sonication of resuspended PC. The in vitro kinase assays were performed as in B & C in the presence or absence of 0.1 mg/ml PC vesicles. Data represent the mean ± S.D. from four separate experiments.

Figure 2. GRK5 binds to Hip

(A) GST and GST-Hip were expressed in E. coli and purified using glutathione beads (Coomassie stain). 21 pmol of purified GST or GST-Hip was then incubated with 3 pmol of purified GRK2 or GRK5 for 1.5 hr at room temperature with rocking. The beads were washed and the bound proteins were eluted with the addition of SDS sample buffer. The proteins were then separated by SDS-PAGE, transferred to nitrocellose and immunoblotted for GRK2 or GRK5. The experiment was performed nine times. (B) HEK293 cells were transiently transfected with pcDNA, HA-tagged Hip Wt and GRK5 as indicated. Cells were lysed 48 hr post-transfection and immunoprecipitated using a polyclonal anti-HA antibody. Immunoprecipitates were then immunoblotted for GRK5 and Hip. Cell lysates were immunoblotted for Hip, GRK4–6 and tubulin expression. The data shown are representative of three independent experiments.

Figure 3. GRK5 binding to Hip

(A) Hip is divided into five domains: the N-terminal domain (residues 1–99, white), three tetracopeptide repeat domains (residues 116–226, gray labeled TPR), a highly charged domain (residues 228–283, white labeled + − + −), a domain enriched in glycine, methionine and proline (residues 280–303, black labeled G), and a C-terminal domain (residues 303–369, dark grey). (B) GST tagged Hip domain deletions were expressed in E. coli and purified using glutathione beads (Coomassie stain). 21 pmol of GST or GST-Hip deletion mutants were incubated with 3 pmol of purified GRK5 for 1.5 hr at room temperature. The beads were then extensively washed and SDS sample buffer was added to elute the bound proteins. The proteins were separated by SDS-PAGE, transferred to nitrocellulose and immunoblotted with a monoclonal GRK4–6 antibody. (C) The western blot in B was quantitated using Odysessy software. The data (mean ± S.D., n=3) are from three separate experiments. (D) GST pulldowns were performed as described in B using serial truncation mutants of Hip. (E) The western blot in D was quantitated using Odyssey software. The data (mean ± S.D., n=3) are from three separate experiments.

Figure 4. GRK5 phosphorylates Hip between residues 303 and 369

(A) Hexa-histidine tagged Hip deletion mutants were expressed in E. coli and purified using Ni⁺ agarose and gel filtration chromatography. One microgram of purified protein was separated by SDS-PAGE and Coomassie stained. (B) 100 nM of purified GRK5 was incubated with 100 nM of purified Hip deletion mutants for various times at 30°C in a buffer containing radiolabeled ATP. Samples were separated by SDS-PAGE and subjected to autoradiography. (C) The radioactively labeled bands from B were excised and quantitated by scintillation counting to determine the stoichiometry of phosphorylation. Data represent the mean ± S.D. from five separate experiments.

Figure 5. Hip S346A mutant is not phosphorylated by GRK5

(A) The amino acid sequence of Hip between residues 273 and 369 is shown. The underlined portion highlights the residues in the Hip Δ303–369 construct that are deleted. The arrows depict the sites of cleavage by trypsin that result in a 7 kDa fragment. Ser-318 and Ser-346 are in bold. (B) Hexa-histidine tagged Hip wild type, S318A and S346A mutants were expressed in E. coli and purified using Ni⁺ agarose and gel filtration chromatography. One microgram of purified protein was separated by SDS-PAGE and Coomassie stained. (C) 100 nM of purified GRK5 was incubated with 100 nM of purified Hip point mutants for various times at 30°C in a buffer containing radiolabeled ATP. Samples were separated by SDS-PAGE and subjected to autoradiography.

Figure 6. GRK5 phosphorylates Hip in cells

(A) HEK293 cells were transfected with pcDNA3 or HA Hip Wt. 48 hr post-transfection, cells were lysed and immunoprecipitated with a monoclonal anti-HA antibody. The immunoprecipitates were then subjected to dephosphorylation by λ-phosphatase for 30 min at 30°C. The immunoprecipitates were separated by SDS-PAGE, transferred to low-fluorescence PVDF membrane and stained with Pro-Q Diamond Stain. The stain was visualized using a Typhoon scanner. The PVDF membrane was then immunoblotted for Hip. (B) HEK293 cells were transfected with control (ctrl) or GRK specific siRNA and HA-Hip Wt in pcDNA3. Seventy two hours post-transfection the samples were prepared as described in A without λ-phosphatase treatment. Equivalent amounts of cell lysates were probed for GRK expression. (C) The amount of Pro-Q Diamond stain was quantitated using ImageQuant 5.2 software and was normalized to the amount of staining in the control siRNA treated samples. The data shown are the mean ± S.D. (* p≤0.05, ** p≤0.01, *** p≤0.001) from four separate experiments.

Figure 7. GRK5 phosphorylates Hip at Ser-346

(A) HEK293 cells were transfected with control (ctrl) or GRK specific siRNA and HA-Hip Wt or S346A in pcDNA3. Seventy two hr post-transfection, cells were lysed and immunoprecipitated with a monoclonal anti-HA antibody. The immunoprecipitates were separated by SDS-PAGE, transferred to low-fluorescence PVDF membrane and stained with Pro-Q Diamond Stain. The stain was visualized using a Typhoon scanner. The PVDF membrane was then immunoblotted for Hip. Equivalent amounts of cell lysates were probed for GRK5 expression. (B) The amount of Pro-Q Diamond stain was quantitated using ImageQuant 5.2 software and was normalized to the amount of staining in the control siRNA treated samples. The data shown are the mean ± S.D. (** p≤0.01, *** p≤0.001) from four separate experiments.

Figure 8. Hip Ser-346 mediates internalization of CXCR4

(A) Control (Ctrl) or Hip shRNA expressing HeLa cells were transiently transfected with FLAG-CXCR4 Wt and pcDNA3-Hip Wt, S346A or S346D. Equivalent amounts of cell lysate were separated by SDS-PAGE, transferred to nitrocellulose and probed for Hip expression. The membrane was stripped and probed for tubulin as a loading control. (B) Cells transfected as in A were serum starved and stimulated with 50 nM CXCL12 for 30 min. The cells were then fixed with 4% PFA and the remaining amount of surface CXCR4 was determined by cell surface ELISA. The data (mean ± S.D., n=3, ** p≤0.01, *** p≤0.001) are from five separate experiments and are expressed as the fraction of cell surface expression relative to the 0 min untreated control sample.

Figure 9. CXCR4 activation promotes dephosphorylation of Hip

(A) HEK293 cells were transfected with Flag-tagged CXCR4 and HA-tagged Wt or S346A mutant Hip. Forty eight hr post-transfection, cells were stimulated with 50 nM CXCL12 for 15 min, lysed and immunoprecipitated with a monoclonal anti-HA antibody. The immunoprecipitates were separated by SDS-PAGE, transferred to low-fluorescence PVDF membrane and stained with Pro-Q Diamond Stain. The stain was visualized using a Typhoon scanner. The PVDF membrane was then immunoblotted for Hip. (B) The amount of Pro-Q Diamond stain was quantitated using ImageQuant 5.2 software and was normalized to the amount of staining in the control siRNA treated samples. The data shown are the mean ± S.D. (* p≤0.05) from three separate experiments.