α-Dystrobrevin-1 recruits α-catulin to the α1D- adrenergic receptor/dystrophin-associated protein complex signalosome John S. Lyssanda, Jennifer L. Whitingb, Kyung-Soon Leea, Ryan Kastla, Jennifer L. Wackera, Michael R. Bruchasa, Mayumi Miyatakea, Lorene K. Langebergb, Charles Chavkina, John D. Scottb, Richard G. Gardnera, Marvin E. Adamsc, and Chris Haguea,1 Departments of aPharmacology and cPhysiology and Biophysics, University of Washington, Seattle, WA 98195; and bDepartment of Pharmacology, Howard Hughes Medical Institute, University of Washington, Seattle, WA 98195 Edited by Robert J. Lefkowitz, Duke University Medical Center/Howard Hughes Medical Institute, Durham, NC, and approved October 29, 2010 (received for review July 22, 2010) α1D-Adrenergic receptors (ARs) are key regulators of cardiovascu- pression increases in patients with benign prostatic hypertrophy lar system function that increase blood pressure and promote vas- (12). Through proteomic screening, we discovered that α1D-ARs cular remodeling. Unfortunately, little information exists about are scaffolded to the dystrophin-associated protein complex the signaling pathways used by this important G protein-coupled (DAPC) via the anchoring protein syntrophin (10). Coexpression α “ α receptor (GPCR). We recently discovered that 1D-ARs form a sig- with syntrophins increases 1D-AR plasma membrane expression, nalosome” with multiple members of the dystrophin-associated drug binding, and activation of Gαq/11 signaling after agonist protein complex (DAPC) to become functionally expressed at the activation. Moreover, syntrophin knockout mice lose α1D-AR– plasma membrane and bind ligands. However, the molecular stimulated increases in blood pressure, demonstrating the im- α α mechanism by which the DAPC imparts functionality to the 1D- portance of these essential GIPs for 1D-AR function in vivo (10). AR signalosome remains a mystery. To test the hypothesis that Proper organization of signaling molecules within cells by the previously unidentified molecules are recruited to the α1D-AR sig- DAPC is essential for the maintenance of cellular homeostasis at nalosome, we performed an extensive proteomic analysis on each synaptic junctions (13). Mutations in DAPC result in severe PHARMACOLOGY member of the DAPC. Bioinformatic analysis of our proteomic data muscle wasting diseases, such as Duchenne muscular dystrophy/ sets detected a common interacting protein of relatively unknown Becker muscular dystrophy, and as a result, the role of this function, α-catulin. Coimmunoprecipitation and blot overlay complex for proper skeletal muscle function has been thoroughly assays indicate that α-catulin is directly recruited to the α1D-AR studied (14). However, the DAPC performs many other func- signalosome by the C-terminal domain of α-dystrobrevin-1 and tions: it facilitates proper water transport across the blood–brain not the closely related splice variant α-dystrobrevin-2. Proteomic barrier by anchoring aquaporin (15, 16), clusters nicotinic ace- and biochemical analysis revealed that α-catulin supersensitizes tylcholine receptors to ensure signal transmission at para- α1D-AR functional responses by recruiting effector molecules to sympathetic synapses (13), and anchors neuronal NOS at the cell the signalosome. Taken together, our study implicates α-catulin membrane in cardiac myocytes to permit cardiodilation (17). We as a unique regulator of GPCR signaling and represents a unique previously demonstrated that α1D-ARs form a complex with the expansion of the intricate and continually evolving array of GPCR DAPC (10), but why this interaction is necessary for α1D-AR signaling networks. functional coupling is unknown. In this study, we postulated that molecules necessary for α protein-coupled receptors (GPCRs) are seven-transmembrane 1D-AR signaling are recruited by the DAPC. Using a sequential fi α spanning proteins that are responsible for communicating in- proteomic screening approach, we identi ed -catulin as a G α formation in the form of extracellular stimuli across lipid mem- unique member of the 1D-AR signalosome. The goal of these α branes into distinct intracellular signals with precise accuracy. After experiments was to understand how -catulin integrates into this ligand binding, GPCRs signal through the canonical heterotrimeric growing GPCR protein complex and to decipher the purpose of G protein signaling pathway to activate a diverse array of down- this relatively unstudied protein in GPCR signaling networks. stream effectors (1). Recently, it has become evident that most Results GPCRs collaborate with one or more additional proteins at specific α α points in their lifecycle. These GPCR interacting proteins (or GIPs) -Catulin: Unique Member of the 1D-AR/DAPC Signalosome. We α are largely receptor subtype and cell context specific, include both previously demonstrated that syntrophins are required for 1D- α membrane and cytosolic proteins, and typically play a highly specific AR function in vitro and in vivo by anchoring 1D-ARs to the fi DAPC (10). Our working hypothesis is that the DAPC facilitates supporting role for GPCR function (i.e., traf cking, ligand binding, α enhancing signaling, signal termination, and/or degradation) (2–4). 1D-AR function by acting as a multiprotein scaffold to arrange Recently, we used yeast two-hybrid and proteomic screens to signaling molecules in close proximity to the receptor. To test this hypothesis, we fused tandem-affinity purification (TAP) epitopes identify GIPs for a clinically important GPCR, the α1D-adrenergic receptor (AR) (5). A member of the adrenergic family (α1, α2, β), α1D-ARs are ubiquitously expressed on blood vessels and are responsible for increasing blood pressure during exercise, injury, Author contributions: J.S.L., J.L. Whiting, J.L. Wacker, M.R.B., M.M., L.K.L., C.C., J.D.S., R.G.G., α M.E.A., and C.H. designed research; J.S.L., J.L. Whiting, K.-S.L., R.K., M.R.B., M.M., L.K.L., stress, or cardiovascular disease (6). 1D-AR knockout mice are M.E.A., and C.H. performed research; J.S.L., J.L. Whiting, J.L. Wacker, M.R.B., C.C., J.D.S., hypotensive and resistant to high salt diet-induced hypertension R.G.G., M.E.A., and C.H. analyzed data; and J.S.L., J.D.S., R.G.G., M.E.A., and C.H. wrote the (7, 8), yet this GPCR has been largely ignored over the past 20 y paper. fl because after transfection into cell culture α1D-ARs are seques- The authors declare no con ict of interest. tered in the endoplasmic reticulum (9, 10). Clinical interest in This article is a PNAS Direct Submission. the α1D-AR as a drug target has recently increased with the dis- 1To whom correspondence should be addressed. E-mail: [email protected]. α coveries that 1D-ARs are the predominant subtype expressed in This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. epicardial coronary arteries (11) and that α1D-AR prostate ex- 1073/pnas.1010819107/-/DCSupplemental. www.pnas.org/cgi/doi/10.1073/pnas.1010819107 PNAS Early Edition | 1of6 α α α containing streptavidin/calmodulin-binding proteins to the α1D- -Catulin Is Recruited to the 1D-AR Signalosome by -DB1. We α α α β β AR and all members of the DAPC, including, α-syntrophin, β - detected -catulin as a positive interactor in the 1D-AR, / 1/ 2- 1 α fi syntrophin, β -syntrophin, α-dystrobrevin-1 (α-DB1), and α-DB2. syntrophin, and -DB1 proteomic screens, making it dif cult to 2 α HEK293 cell lines stably expressing each individual clone were predict which protein directly binds and recruits -catulin to the α created. TAP-tagged and associated proteins were purified from 1D-AR/DAPC signalosome. An important clue was that α-catulin was not detectable in the α-DB2 screen (Fig. 1A and lysates of each stable cell line, digested with trypsin, and the Table S1). α-DB1 and α-DB2 are C-terminal splice variants of tryptic peptides subjected to liquid chromatography (LC) tandem the α-DB gene (20). Following their common 504-aa core, mass spectrometry (MS/MS). α-DB1 has a unique 184-aa C terminus, whereas α-DB2 has a 9- fi Analysis of the peptides identi ed by LC-MS/MS revealed aa unique C terminus. Because α-catulin interacts with α-DB1 both predicted and unique components of the scaffold (Fig. 1A but not with α-DB2, we suspected that α-catulin is most likely and Table S1). The core components of the scaffold (dystrophin, recruited by association with the α-DB1 C-terminal domain. utrophin, α/β-DB, and α/β1/β2-syntrophin) were cross-identified To test this possibility, we used HEK293 cells stably expressing in each of the individual pulldowns. Our experiments identified TAP-α-DB1 or TAP-α-DB2. Coimmunoprecipitation experi- α known interactors [i.e., CASK and LIN7C with β2-syntrophin, ments revealed that TAP- -DB1 interacts with both endogenous Gα subunits with α-syntrophin (18, 19)] for each bait protein, and transiently transfected α-catulin-flag, whereas TAP-α-DB2 confirming the validity of our experimental data sets. One pro- failed to coimmunoprecipitate either protein (Fig. 2A). These tein, α-catulin (CTNNAL1) was consistently identified in the results were corroborated by experiments demonstrating that α-catulin-flag coimmunoprecipitates specifically with endoge- majority of our proteomic screens (Table S1). α α Next, we confirmed these findings using immunoprecipitation nous -DB1 but not -DB2 (Fig. 2B). Next, we used nitrocellulose blot overlay assays to test whether experiments of full-length proteins in living cells. Shown in Fig. α α α the interaction between -DB1 and -catulin is direct. TAP- 1B, lysates from HEK293 cells stably expressing TAP- 1D-ARs α-DB1, TAP-α-DB2, or untransfected HEK293 cell lysates were α α were precipitated and probed for -catulin, -DB1, and syntro- separated by SDS/PAGE, transferred to nitrocellulose, and in- phin.