Scholarly article on topic 'Homologous desensitization of signalling by the alpha (α) isoform of the human thromboxane A2 receptor: A specific role for nitric oxide signalling'

Homologous desensitization of signalling by the alpha (α) isoform of the human thromboxane A2 receptor: A specific role for nitric oxide signalling Academic research paper on "Biological sciences"

CC BY
0
0
Share paper
Keywords
{"Thromboxane receptor" / Alpha / Desensitization / Phosphorylation / "Nitric oxide" / "G protein coupled receptor"}

Abstract of research paper on Biological sciences, author of scientific article — Leanne P. Kelley-Hickie, Martina B. O'Keeffe, Helen M. Reid, B. Therese Kinsella

Abstract Thromboxane (TX) A2 plays a central role in hemostasis, regulating platelet activation status and vascular tone. We have recently established that the TPβ isoform of the human TXA2 receptor (TP) undergoes rapid, agonist-induced homologous desensitization of signalling largely through a G protein-coupled receptor kinase (GRK) 2/3-dependent mechanism with a lesser role for protein kinase (PK) C. Herein, we investigated the mechanism of desensitization of signalling by the TPα isoform. TPα undergoes profound agonist-induced desensitization of signalling (intracellular calcium mobilization and inositol 1,4,5 trisphosphate generation) in response to the TXA2 mimetic U46619 but, unlike that of TPβ, this is independent of GRKs. Similar to TPβ, TPα undergoes partial agonist-induced desensitization that occurs through a GF 109203X-sensitive, PKC mechanism where Ser145 within intracellular domain (IC)2 represents the key phospho-target. TPα also undergoes more profound sustained PKC- and PKG-dependent desensitization where Thr337 and Ser331, respectively, within its unique C-tail domain were identified as the phospho-targets. Desensitization was impaired by the nitric oxide synthase (NOS), soluble guanylyl cyclase (sGC) and PKG inhibitors l-NAME, LY 83583 and KT5823, respectively, indicating that homologous desensitization of TPα involves nitric oxide generation and signalling. Consistent with this, U46619 led to rapid phosphorylation/activation of endogenous eNOS. Collectively, data herein suggest a mechanism whereby agonist-induced PKC phosphorylation of Ser145 partially and transiently impairs TPα signalling while PKG- and PKC-phosphorylation at both Ser331 and Thr337, respectively, within its C-tail domain profoundly desensitizes TPα, effectively terminating its signalling. Hence, in addition to the agonist-mediated PKC feedback mechanism, U46619-activation of the NOS/sGC/PKG pathway plays a significant role in inducing homologous desensitization of TPα.

Academic research paper on topic "Homologous desensitization of signalling by the alpha (α) isoform of the human thromboxane A2 receptor: A specific role for nitric oxide signalling"

Biochimica et Biophytica Acta

www.elsevier.com/locate/bbamcr

Homologous desensitization of signalling by the alpha (a) isoform of the human thromboxane A2 receptor: A specific role for nitric oxide signalling

Leanne P. Kelley-Hickie, Martina B. O'Keeffe, Helen M. Reid, B. Therese Kinsella *

School of Biomolecular and Biomedical Science, Conway Institute of Biomolecular and Biomedical Research, University College Dublin, Belfield, Dublin 4, Ireland

Received 23 January 2007; received in revised form 14 March 2007; accepted 15 March 2007 Available online 28 March 2007

Abstract

Thromboxane (TX) A2 plays a central role in hemostasis, regulating platelet activation status and vascular tone. We have recently established that the TPp. isoform of the human TXA2 receptor (TP) undergoes rapid, agonist-induced homologous desensitization of signalling largely through a G protein-coupled receptor kinase (GRK) 2/3-dependent mechanism with a lesser role for protein kinase (PK) C. Herein, we investigated the mechanism of desensitization of signalling by the TPa isoform. TPa undergoes profound agonist-induced desensitization of signalling (intracellular calcium mobilization and inositol 1,4,5 trisphosphate generation) in response to the TXA2 mimetic U46619 but, unlike that of TPp., this is independent of GRKs. Similar to TPp., TPa undergoes partial agonist-induced desensitization that occurs through a GF 109203X-sensitive, PKC mechanism where Ser145 within intracellular domain (IC)2 represents the key phospho-target. TPa also undergoes more profound sustained

337 331

PKC- and PKG-dependent desensitization where Thr and Ser , respectively, within its unique C-tail domain were identified as the phospho-targets. Desensitization was impaired by the nitric oxide synthase (NOS), soluble guanylyl cyclase (sGC) and PKG inhibitors l-NAME, LY 83583 and KT5823, respectively, indicating that homologous desensitization of TPa involves nitric oxide generation and signalling. Consistent with this, U46619 led to rapid phosphorylation/activation of endogenous eNOS. Collectively, data herein suggest a mechanism whereby agonist-induced

145 331 337

PKC phosphorylation of Ser partially and transiently impairs TPa signalling while PKG- and PKC-phosphorylation at both Ser and Thr , respectively, within its C-tail domain profoundly desensitizes TPa, effectively terminating its signalling. Hence, in addition to the agonist-mediated PKC feedback mechanism, U46619-activation of the NOS/sGC/PKG pathway plays a significant role in inducing homologous desensitization of TPa. © 2007 Elsevier B.V. All rights reserved.

Keywords: Thromboxane receptor; Alpha; Desensitization; Phosphorylation; Nitric oxide; G protein coupled receptor

Available online at www.sctencedirect.com

W ScienceDirect

Biochimica et Biophysica Acta 1773 (2007) 970 - 989

1. Introduction

The prostanoid thromboxane (TX)A2, synthesized through the sequential actions of cyclooxygenase (COX) 1 and/or COX 2 and TXA2 synthase, plays a critical role in the dynamic control of hemostasis and vascular tone [1,2] and may induce

Abbreviations: C-tail, carboxyl-terminal tail; [Ca2+]i, intracellular calcium; COX, cyclooxygenase; FBS, foetal bovine serum; GPCR, G protein-coupled receptor; GRK, G protein-coupled receptor kinase; HA, hemagglutinin; HEK, human embryonic kidney; IP, prostacyclin receptor; IP3, inositol 1, 4, 5-trisphosphate; NO, nitric oxide; NOS, nitric oxide synthase; PAGE, polyacryla-mide gel electrophoresis; PG, prostaglandin; PK, protein kinase; PL, phospholi-pase; sGC, soluble guanylyl cyclase; TP, TXA2 receptor; TX, thromboxane * Corresponding author. Tel.: +353 1 7166727; fax: +353 1 2837211.

E-mail address: Therese.Kinsella@ucd.ie (B.T. Kinsella).

other diverse cellular responses including mitogenic and/or hypertrophic growth of vascular smooth muscle [3-5], inhibition of angiogenesis/neo-vascularization [6-8] and apoptosis of CD4/CD8+/- immature thymocytes [9]. Additionally, the TXA2 receptor (TP) mediates at least some of the adverse actions of the isoprostane 8-iso prostaglandin (PG)F2a, generated in situations of oxidative stress [10,11]. In keeping with these actions, imbalances in the levels of TXA2 or of its synthase or receptor (TP) have been implicated in a number of vascular diseases including thrombosis, unstable angina, bronchial asthma, hypertension and glomerulonephritis [3-5,12,13].

The TXA2 receptor, or TP, a member of the G protein coupled receptor superfamily, is primarily coupled to Gq/ phospholipase (PL)Cp> activation and to mobilization of Ca2+ from inositol 1,4,5 trisphosphate-operated intracellular stores

0167-4889/$ - see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.bbamcr.2007.03.012

[1]. In humans, TXA2 signals through two TP isoforms termed TPa and TPp. that are identical for their N-terminal 328 amino acids differing exclusively in their carboxyl terminal (C)-tail domains [14-16]. Whilst the significance of two receptors for TXA2 in humans, but not in non-primates, is unknown there is substantial evidence that they display critical differences in their signalling and profiles of expression and hence it is thought likely that they have distinct physiologic/pathophy-siologic roles [17]. TPa and TPp. exhibit distinct patterns of expression in a variety of cell/tissue types of vascular origin, including platelets [18-20] and such differences are due to the fact that TPa and TPp. are not only products of differential splicing but are also under the transcriptional control of distinct promoters within the single human TP gene [21-24]. While TPa and TPp. exhibit identical ligand binding and coupling to PLCp. [15,25,26], they oppositely regulate adenylyl cyclase [27] and TPa, but not TPp>, mediates activation of Gh, leading to PLC activation [28].

A central feature of the general GPCR signal transduction cascade is the regulation or desensitization of second messenger generation and signalling that occurs in response to the continued presence of the ligand, dampening or terminating the specific cellular response [29,30]. Such desensitization is typically initiated by GPCR phosphorylation leading to uncoupling of the receptor from its cognate G-protein [29,30] and may involve either cross-talk/heterologous desensitization between different signalling systems or agonist-induced/homologous desensitiza-tion terminating or modulating the response to the receptor's own ligand [29,30]. In studies investigating heterologous desensitiza-tion or cross-talk between TXA2 and other prostanoids, it has been recently established that while both TP isoforms have evolved to share a similar mechanism of protein kinase (PK)C-induced phosphorylation in response to signalling by the EP1 subtype of the prostaglandin (PG) E2 receptor and by the PGF2a receptor (FP), those phosphorylations/desensitizations occur at distinct PKC sites within the unique C-tail domains of TPa and TPp>, respectively [31,32]. On the other hand, signalling by TPa, but not TPp>, is subject to prostacyclin and PGD2-mediated desensitization in a mechanism involving direct protein kinase (PK) A phosphorylation of TPa very first residue

within its unique C-tail domain divergent from TPp. [33,34]. Consistent with the latter, TPa, butnotthe TP(S, is also a target for nitric oxide (NO)-induced heterologous desensitization of signalling that occurs through direct cGMP-dependent PKG/cGK phosphorylation at Ser331 also within its C-tail [35]. Collectively, these latter studies indicate critical differences in the modes of regulation of signalling by the individual TP isoforms. Specifically, they point to an essential role for TPa in prostacyclin- and NO-regulated vascular hemostasis and point to a redundant or an, as yet, unidentified role for TPp. in this essential physiologic process [33-35].

As stated, agonist-induced homologous desensitization of GPCRs is a central mechanism whereby the cellular responses to the receptor's own ligand is dynamically modulated and is most often initiated through phosphorylation of the ligand-engaged, conformationally active GPCR such as by members of the G protein-coupled receptor kinases (GRKs) and/or by the second-

messenger kinases, such as PKA or PKC [29,30,36]. While both TP isoforms undergo rapid agonist-induced phosphorylation in response to TXA2 mimetic U46619 [20,25], TPp>, but not TPa, is subject to GRK/p>-arrestin-dependent internalization following prolonged agonist exposure [37,38]. We recently examined the mechanism of agonist-induced homologous desensitization of signalling and second messenger generation by TPp. and established that it undergoes two main mechanisms of desensitization: (i) partial, transient PKC-dependent desensitization where Ser145 within IC2 was identified as the PKC phospho-target. In addition, (ii) TPp. also undergoes more profound and sustained agonist-induced desensitization involving GRK 2/3-phosphorylation followed by internalization and we proposed a model whereby phosphorylation of Ser357 within the C-tail domain led to the recruitment and orientation of GRK2/3 which, in turn, facilitated its phosphorylation of Ser239 within IC3 thereby disrupting Gq coupling by TPp>. Hence, GRK2/3 binding is dependent on sequences within the unique C-tail domain of TPp. whilst Gq uncoupling is due to GRK phosphorylation of Ser239 within IC3 and, to a lesser extent, PKC phosphoryla-tion of Ser145 within IC2.

In view of these findings, the aim of the current studies was to investigate the mechanism of homologous desensitization of signalling by TPa in response to the TXA2 mimetic U46619. Our data established that TPa is subject to homologous desensitization of signalling through two key mechanisms involving receptor phosphorylation: one involving a partial PKC-dependent mechanism and another involving a profound and sustained PKG-dependent mechanism engaging NO signalling and endothelial (e) nitric oxide synthase (NOS) activation. Unlike that of TPp>, we found no role for the GRK/p>-arrestins in homologous desensitization of TPa. Hence, taken together, these studies demonstrate critical differences in the mechanisms of agonist-induced desensitization of signalling between the TP isoforms and add to the increasing lines of evidence for distinct physiologic roles for TPa and TPp. in the mediation and the regulation of signalling by the potent autocoid TXA2 in humans.

2. Materials and methods

2.1. Materials

SQ29,548 and I-BOP were from the Cayman Chemical Company. Anti-Gaq (C15; SC-392) antibody was obtained from Santa Cruz Laboratories. [32P] orthophosphate (8000-9000 Ci/mmol) and [3H]SQ29,548 (50.4 Ci/ mmol) was from NEN Life Sciences Products. [3H] IP3 (20-40 Ci/mmol) was obtained from American Radiolabelled Chemicals Inc. Anti-HA 101R monoclonal antibody and anti-HA-3F10-peroxidase conjugated antibody were obtained from BABCO and from Roche Molecular Biochemicals, respectively. Monoclonal anti-eNOS antibody (Clone 3) was obtained from Transduction Laboratories. Anti-phospho-eNOSSer1177(Human)/1179(Bovine) antibody was purchased from Cell Signalling Technology. Thrombin was obtained from Sigma Chemical Co. FURA2/AM and d-myo-inositol 1,4,5-trisphosphate, 3-deoxyhexasodium salt, U46619, N-[2-((p-bromocinnamyl) amino)ethyl]-5-isoquino-linesulfonamide, 2HCL (H-89) and 2-[1-(3-dimethy-laminopropyl)-1H-indol-3-yl]-3-(1H-indol-3-yl)-maleimide (GF 109203X), NG-Nitro-l-arginine Methyl Ester, Hydrochloride (l-NAME), KT 5823, LY 83583 (6-Anilino-5, 8-quinolinequinone) were obtained from Calbiochem.

2.2. Subcloning and site-directed mutagenesis

The plasmids pCMV:Gaq, pHM:TPß, pHM:TPa, pHM:TPA 328, pHM: TPaS329A, pHM:TPaS331A, pHM:TPaT337A, pHM:TPaS329'331A, pHM:TPaA336 and pRK5:ßARK1495-689 have been previously described [10,32,33,35,39].

Fig. 1 illustrates the C-tail sequence of TPa and its mutated variants that were generated and used in this study. Site-directed mutagenesis was performed using QuickChangeTM (Stratagene), unless specified otherwise. Mutation of Ser145 to Ala145 of TPa and TPA328 to generate the plasmids pHM:TPaS145A and pHM:TPS145A' A328 was achieved using pHM:TPa and pHM:TPA328, respectively, as templates and sense/antisense primer pair (5'-GC CCG GCG GTC GCC gcg CAG CGC GCC-3'). Mutation of Ser239 to Ala239 of TPa to generate pHM:TPaS239 was achieved using pHM:TPa as template and sense/ antisense primer pair (5'-G CGT CCC CGG GAC gcc GAG GTG GAG A-3'). Conversion of both Ser331,Thr337 to Ala331,Ala337of TPa to generate pHM6: TPaS331,337A was achieved using pHM:TPaT337A as template and sense/ antisense primer pair (5'-G CCC AGG TCG CTG gcc CTC CAG CCC C-3'). Mutation of Ser340 to Ala340 of TPa, TPaS331A and TPaT337A to generate pHM: TPaS340A, pHM:TPaS331'340A and pHM:TPaT337'S340A was achieved using pHM:TPa, pHM:TPaS331A and pHM:TPa

lines stably over-expressing HA-epitope tagged forms of TPa

T337A, respectively, as templates and sense/antisense primer pair (5'-C ACG CAG CGC GCC GGG CTG CAG TAG G-3'). Mutation of Ser340 to Ala340 of TPaS331,T337A to generate

pHM:TPa'

S331,T337,S340A

was achieved using pHM:TPa'

S331,T337A

as template

and sense/antisense primer pair (5'-CAG CCC CAG CTC GCG CAG CGC GCC G-3'). Mutation of Ser145 to Ala145 of TPaS331'T337A to generate pHM: TPaS145'S331'T337A was achieved using pHM:TPaS331,T337A as template and sense/antisense primer pair (5'-GC CCG GCG GTC GCC gcg CAG CGC GCC-3'). For each primer pair above, sequence shown corresponds to the sense primer and in each case the identity of the mutator codon is in boldface italics.

2.3. Cell culture and transfections

Human embryonic kidney (HEK) 293 cells were cultured in minimal essential medium with Earle's salts (MEM) supplemented with 10% FBS (foetal bovine serum) and maintained at 37 °C in 5% CO2. The following HEK 293 cell lines stably over-expressing hemagglutinin (HA) epitope-tagged forms of TPp (HEK.TPp), TPa (HEK.TPa), TPA328 (HEK.TPA328), TPaS329A (HEK. TPaS329A), TPaS337A (HEK.TPaT337A), TPaS331A (HEK.TPaS331A), TPaS329'331A (HEK.TPaS329'331A) and TPaA336 have been previously described [32,33,35].

For transfections, routinely HEK 293 cells were plated in 10 cm dishes at a density of 2 x 10 cells/dish in 8 ml media 48 h prior to transfection. Cells were transiently transfected with 10 |g pADVA [40] and 25 |g of pcDNA-, pCMV- or pHM-based vectors using the calcium phosphate/DNA co-precipitation procedure as previously described [10]. For transient transfec-tions, cells were harvested 48 h post transfection. To create HEK 293 cell

TP« prlstrprslslqpqltqrsglq

T,p(-iS329A' 11U. -pp^s331Ä

TpaT340A TpaT337A

TPaA336

Fig. 1. Schematic of the carboxyl (C) tail domain of TPa. The amino acid sequence of the carboxyl terminal (C)-tail domain of TPa (residues 321343) is shown, where residues unique to TPa (residues 329-343) are underlined. The truncation (A) mutant TPa , generated by conversion of Leu336 codon to a stop codon, is indicated by the open arrow head while Ser/Thr to Ala substitutions to generate TPaS329A, TPaS331A, TPaT337A and

mutations are indicated by the solid arrows. The combination

substitutions TPaS329,331A, TPaS331,T337A, TPaS331,340A, TPaT337,S340A and TPaS331,T337,S340A were also generated. Mutations involving Ser145 or Ser239 within IC2 or IC3, respectively, either alone or in combination with the C-tail mutations are not shown.

TPaS145A), TP

TPaS340A (HEK.TPa!

,S145A,A328

(HEK.TP-

A), TPaS

S145A,A328

), TPaS

(HEK.TPaS

(HEK.TPaS239A),

A), TPa

S331,340A

(HEK.TPaS331,340A), TPaT337,S340A (HEK.TPaT337,S340A), TPaS331,T337,S340A

(HEK.TPaS331,T337,S340A),

S134,S331,T337 A (HEK TPaS145,S331,T337A)

cells were transfected with 10 |ig of Sca1-linearised pADVA plus 25 |ig of the appropriate Pvu1-linearised pHM6-based recombinant plasmids. Forty-eight hours post-transfection, G418 (0.8 mg/ml) was applied and after approximately 21 days, individual G418-resistant colonies were selected and individual pure clonal stable cell lines/isolates were examined for TP expression by analysis of radioligand binding.

2.4. Radioligand binding studies

Cells were harvested by centrifugation at 500xg at 4 °C for 5 min and washed three times with ice-cold Ca2+/Mg2+-free phosphate-buffered saline (PBS). TP radioligand binding assays were carried out at 30 °C for 30 min in 100 |l reactions in the presence of 0-40 nM [3H] SQ29,548 for Scatchard analysis or in the presence of 20 nM [3H] SQ29,548 for saturation radioligand binding experiments as previously described [10]. Protein determinations were carried out using the Bradford assay [41].

2.5. Measurement of intracellular calcium ([Ca2+]i) mobilization

Measurement of intracellular calcium mobilization ([Ca2+]i) in FURA2/ AM preloaded HEK 293 cell lines (2x106 cells/dish), each transiently co-transfected with pADVA (10 |g), and pCMV:Gaq (25 |g), was carried out as previously described [10]. To investigate the effect of GRK2/|3ARK1 on TP-mediated [Ca2+]j mobilization, HEK.TPa or HEK.TPp cells (2 x 106 cells/dish) were co-transfected with pADVA (10 |g), pCMV:Gaq (25 |g) along with pRK5:|3ARK1495-689 (25 |g), encoding a dominant negative form of GRK2/ pARK1[39]. Approximately 48 h post-transfection, cells were harvested by scraping, washed twice in ice-cold PBS and resuspended in HBSSHB (modified Ca2+/Mg2+-free Hank's buffered salt solution, containing 10 |M HEPES, pH 7.67, 0.1% bovine serum albumin (BSA)) buffer at 107 cells/ml and incubated in the dark with 5 |M FURA2/AM for 45 min at 37 °C. Cells were collected by centrifugation (900xg, 5 min), washed once in an equal volume of HBSSHB, and were finally resuspended in HBSSHB buffer at 107 cells/ml and kept at room temperature in the dark for 40 min. For each measurement of [Ca2+]i mobilization, aliquots of cells were diluted to 0.825 x 106 cells/ml in HBSSHB, 1 mM CaCl2 and FURA2 fluorescence was recorded (2 ml aliquots of cells) at 37 °C with gentle stirring with a Perkin Elmer-Cetus LS50-B spectrofluorometer at excitation wavelengths of 340 nm and 380 nm and emission wavelength of 510 nm, respectively.

Cells were stimulated with the TP agonist U46619 (1 | M) and changes in [Ca2+]i mobilization were monitored as a function of time. To assess agonist-mediated homologous desensitization, cells (0.825 x 10 cells/ml; 2 ml per assay) were stimulated with 1 |M U46619 (Primary stimulation) or, as a control, with the vehicle for 4 min; thereafter, cells were collected by centrifugation (900xg, 5 min); were washed twice in 2 ml HBSSHB, 1 mM CaCl2; and finally resuspended in 2 ml HBSSHB,1 mM CaCl2 and were left to recover over various time intervals (ranging from 0 to 60 min); thereafter, cells re-stimulated with 1 |M U46619 (Secondary stimulation) and changes in [Ca2+]i mobilization were monitored as a function of time. To assess the role of protein kinase (PK) A, PKC or PKG on U46619-mediated desensitization of TPa signalling, cells were pre-incubated with the PKA inhibitor H-89 (10 |M), the PKC inhibitor GF 109203X (50 nM) or the PKG inhibitor KT 5823 (50 nM), for 15 min and respective kinase inhibitors were maintained throughout the primary/secondary U46619-stimula-tions and during the washes. To assess the role of the nitric oxide (NO) pathway on U46619-mediated desensitization of TPa signalling, cells were pre-incubated with the NO synthase inhibitor l-NAME (1 | M) or the soluble guanylate cyclase (sGC) inhibitor LY 83583 (1 |M) for 15 min and respective inhibitors were maintained throughout the primary/secondary U46619-stimulations and during the washes.

In all cases, the drugs (agonist or inhibitors in 0.01% ethanol in HBS) were diluted in the vehicle HB SSHB such that addition of 20 |l of the diluted drug/inhibitor to 2 ml of cells resulted in the correct working concentration. The vehicle had no effect on

U46619-mediated [Ca2+]i mobilization. Calibration of the signal was performed in each sample by adding 0.2% Triton X-100 to obtain the maximal fluorescence ratio (Rmax) and then 1 mM EGTA to obtain the minimal fluorescence ratio (Rmju). The ratio of the fluorescence at 340 nm to that at 380 nm is a measure which assumes a Kd of 225 nM Ca2+ for FURA2/AM. The results presented in the figures are representative data from at least three or four independent experiments and are plotted as changes in Ca2+i mobilized (A[Ca2+]j (nM)) as a function of time (s) upon ligand stimulation. Changes in [Ca mobilization were determined by measuring peak rises in [Ca2+]j mobilized (A[Ca2+]j) and were calculated as mean changes in A[Ca2+];±S.E.M. (nM).

2.6. Measurement of IP3 levels

Intracellular IP3 levels were measured as previously described [33,43]. Briefly, cells were harvested by scraping, washed twice in ice-cold PBS and resuspended at approximately 2x106 cells/200 |il in HEPES-buffered saline (HBS; 140 nM NaCl, 4.7 mM KCl, 2.2 mM CaCl2, 1.2 mM KH2PO4, 11 mM glucose, 15 mM HEPES-NaOH, pH 7.4), 10 mM LiCl (Walsh et al., 2000). Cells (200 |l) were pre-incubated in HBS buffer at 37 °C for 10 min; where appropriate inhibitors (H-89, 10 |M; GF 109203X, 50 nM; KT 5823, 50 nM; l-NAME, 1 |M; LY 83583, 1 |M) or vehicle (HBS) were added to the cells and pre-incubated at 37 °C for 15 min. Thereafter, cells were stimulated with 1 |M U46619 at 37 °C for 2 min. To assess agonist-mediated homologous desensitization, cells (2 x 106 cells/200 |l) were stimulated with 1 |M U46619 at 37 °C for 2 min (Primary stimulation); thereafter, cells were collected by centrifugation (900xg, 5 min); were washed once in 200 |l HBS, 10 mM LiCl and finally resuspended in 200 | l HBS, 10 mM LiCl and were left to recover for 60 min; thereafter, cells were re-stimulated with 1 |MU46619 at 37 °C for 2 min (secondary stimulation). All ligands and kinases were pre-diluted in HBS such that 50 |l added to 200 |l of cell suspension in HBS gave the desired final concentration. To determine basal IP3 levels, an equivalent volume (50 |l) of the vehicle HBS was added instead of ligand. The level of IP3 produced was quantified by radio competition assay essentially as described [43]. Levels of IP3 produced by ligand stimulated cells over basal stimulation, in the presence of HBS, were expressed in pmol IP3/mg protein±S.E.M. and as fold stimulation over basal (fold increase±S.E.M.). In all cases, 4 independent experiments were performed, each in duplicate.

2.7. Measurement of agonist-mediated TP phosphorylation

Agonist-mediated TP phosphorylation in whole cells (1.8-2x106 cells/ 10 cm dish) was performed essentially as previously described [33]. Briefly, cells were transiently co-transfected with pADVA (10 |g), pCMV:Gaq (25 |g) plus pcDNA: p Arrestin2 (25 |g) approximately 48 h prior to labelling. Thereafter, cells were washed once in phosphate-free DMEM, 10% dialysed FBS and were metabolically labelled for 60 min in the same medium (2 ml/10 cm dish) containing 100 |Ci/ml [32P] orthophosphate (8000-9000 Ci/mmol) at 37 °C, 5% CO2. Where appropriate, the PKG kinase inhibitor KT 5823 (50 nM) or its vehicle was added during the labelling period. Thereafter, U46619 (1 |M) or an equivalent volume of vehicle (0.01% ethanolinHBS; 20 |l) were added and cells were incubated for 10 min at 37 °C, 5% CO2. Reactions were terminated by transferring the dishes to ice and aspirating the labelled medium and HA-epitope tagged TP receptors were processed and immunoprecipitated using the anti-HA 101R antibody as previously described [33]. Electroblots were then exposed to Xomat XAR (Kodak) film to detect 32P-labelled proteins. Thereafter, blots were subject to phosphorimage analysis, and the intensities of agonist-induced phosphorylation were expressed in arbitrary units relative to basal (vehicle-stimulated) levels. In parallel experiments, cells were incubated under identical conditions in the absence of [32P] orthophosphate; HA-tagged TP receptors were immunoprecipitated from the same cell lines using the anti-HA 101R antibody and immunoblots were screened using the anti-HA 3F10 horseradish peroxidase-conjugated antibody [33].

2.8. Analysis ofNOS expression and eNOS phosphorylation

RNA extracted from HEK 293 cells (1.4 |gin25 |l reactions) were converted to first strand (1°) cDNA using MMLV RT essentially as previously described [18].

Aliquots (3.5 |il) of each 1° cDNA were then used as templates in PCR reactions (25 |l) using forward (F) and reverse (R) primer pairs selective for human endothelial (e) nitric oxide synthase, (F-[5'-CCA GCTAGC CAA AGT CAC CAT-3'], R-[5'-GTC TCG GAG CCA TAC AGG ATT-3'], amplicon 354 bp; [44-46]); neuronal (n)NOS, (F-[5'-TTG GGG GCC TGG GAT TTC TGG-3'], R-[5'CGT TGG CAT GGG GGA GTG AGC-3'], amplicon 465 bp; [47]) and inducible (i) NOS, F-[5'-GAG GAA GTG GGC AGG AGA ATG-3'], R-[5'-GTA GTA GAA AGG GGA CAG GAC-3'], amplicon 294 bp; [48,49]). For each primer pair, PCRs carried out in the absence of template served as controls. For western analysis, aliquots of total HEK 293 cell protein were analysed by SDS-PAGE; blots were screened versus anti-eNOS antibody (Transduction Laboratories) following by chemiluminescence detection, essentially as previously described [33].

In order to assess eNOS phosphorylation, HEK.TPa, HEK.TPp or HEK 293 cells (2 x 106 cells/10-cm dish) were plated some 48 h prior to experimentation. Cells were incubated with vehicle, 1 |M U46619 or 10 U/ml Thrombin for 5 min at 37 °C and aliquots of whole cell protein (100 |g/lane) were analysed by SDS-PAGE/western blot. Blots were initially screened using the anti-phospho eNOS1177 as per supplier's instructions (Cell Signalling Technology) and, following stripping, were subsequently re-screened using the anti-eNOS (Transduction Laboratories).

2.9. Data analysis

Radioligand binding and Scatchard analysis data were analysed using GraphPad Prisim V4.0 programme (GraphPad Software Inc., San Diego CA, USA). Statistical analyses were carried out using the unpaired Student's t test using the Statworks Analysis Package. pvalues < 0.01 indicated statistically significant differences.

3. Results

3.1. Effect of primary and secondary agonist stimulation on TPa-signalling

We have recently investigated the mechanism of agonist-induced homologous desensitization of signalling by the TPe isoform of the human thromboxane (TX) A2 receptor [50]. Herein, the aim of the current study was to investigate the mechanism of homologous desensitization of signalling by the TPa isoform of that receptor. We initially investigated desensitization of TPa signalling ([Ca2+]i mobilization and IP3 generation) in response to the TXA2 mimetic U46619 by comparing it to that of the previously characterised TPA328 [50], a truncated variant devoid of the divergent residues within the C-tail domains of TPa and TPe [26,33].

Stimulation of HEK.TPa cells, stably over-expressing TPa (Table 1), with the TXA2 mimetic U46619 resulted in efficient mobilization of intracellular calcium ([Ca2 ]i, Fig. 2A), consistent with previous findings [26,33]. Pre-stimulation of HEK.TPa cells with U46619 almost completely desensitized TPa signalling following agonist-washout and secondary restimulation at 15 min following the initial U46619 stimulation (Fig. 2B, p=0.0001). Moreover, U46619-desensitization of TPa signalling was sustained and did not significantly recover even at 60 min following the primary stimulation (Fig. 2C, p=0.0001). On the other hand, pre-treatment of HEK.TPa cells with the vehicle, as opposed to U46619 itself, during the primary stimulation followed by washout confirmed that the cells retained the ability to fully signal in response to subsequent stimulation with U46619 (Fig. 2C, inset). Hence the loss of the

Table 1

Cell lines

(nM±S.E.M.) (pmol/mg protein±S.E.M.)

HEK.TPaWT 10.5±1.1 4.52±0.05

HEK.TPA32S 7.2±4.1 4.47±0.04

HEK.TPaS145A 6.1±1.3 3.03±0.04

HEK TPS145A.A328 6.1 ±3.7 3.83±0.03

HEK.TPaS239A 6.3 ±2.4 4.35±0.04

HEK.TPaS329A 7.1 ±3.2 4.43±0.04

HEK.TPaS331A 6.3±1.2 4.00±0.04

HEK.TPaA336 7.8±2.1 4.43±0.04

HEK.TPaT337A 6.5 ±1.7 4.23±0.04

HEK.TPaS340A 8.1 ±2.4 5.3±0.04

HEK.TPaS329'331A 8.2±1.65 2.9±0.19

HEK.TPaT337'S340A 7.5 ±5 4.9±0.04

HEK.TPaS331'T337A 6.9±3.1 3.23±0.03

HEK.TPaS331'T337'S340A 6.3±4 4.01±0.05

HEK.TPßWT 9.9±0.5 3.64±0.03

Scatchard analysis of HEK 293 cells stably over expressing HA-epitope tagged forms of TPa or TPp or their variant receptors were carried out in the presence of the TP antagonist [3H]SQ29,548 (50.4 Ci/mmol, 0-40 nM) using 75 ^g whole cell protein/assay. Radioligand binding data were analysed with the Graphpad prism V 3.0 computer program (GraphPad Software Inc.) to determine the Kd and Bmax values. Data are presented as the mean values of four independent experiments± standard error mean (S.E.M.). HEK 293 control cells expressed 154±4.1 fmol [3H]SQ 29,548/mg protein±S.E.M. (n = 3).

secondary U46619-response was not due to mechanical stress during agonist-washout per se but rather resulted from agonist-induced desensitization of TPa signalling.

While primary simulation of HEK.TPA328 cells also led to efficient U46619-induced [Ca2 ]i mobilization (Fig. 2D), consistent with previous data [50] pre-stimulation with U46619 only

ry . ^ a

partially desensitized [Ca ]i mobilization by TP such that ~ 80% of its primary U46619-response occurred following its restimulation at 15 min (Compare Fig. 2D versus Fig. 2E, p=0.2). Moreover, desensitization of U46619-mediated [Ca2 ]i mobilization by TPA328 was transient such that its signalling had fully recovered showing no desensitization at 60 min (Compare Fig. 2D versus Fig. 2F; p=0.22). Consistent with these data, primary stimulation of HEK.TPa and HEK.TPA328 cells with U46619 also yielded efficient increases in IP3 generation (Figs. 2G and H, respectively). While pre-incubation with U46619 almost completely inhibited IP3 generation by TPa following its secondary restimulation (Fig. 2G, p=0.0049), it had no significant effect on IP3 generation by TPA328 at 60 min following the primary stimulation (Fig. 2H, p = 1). Moreover, similar agonist-induced desensitization of [Ca2 ]i mobilization and IP3 generation occurred when I-BOP was used as the TP-stimulatory ligand (data not shown). Taken together these data confirm that like TPp>, TPa is subject to near complete and sustained agonist-induced homologous desensitization while, consistent with previous reports [50], TPA328 undergoes partial and transient desensitization suggesting that the major target site(s) involved in homologous desensitization of TPa may be located within its unique C-tail domain.

To investigate whether the second messenger-regulated protein kinases (PKs) may play a role in mediating homologous desensitization of TPa, the effects of the PKA and PKC inhibitors H-89 and GF 109203X, respectively, on U46619-induced [Ca2+]i mobilization were investigated. Pre-incubation

of HEK.TPa or HEK.TPA328 cells with H-89 had no significant effect on [Ca2+]j mobilization in response to their primary or secondary stimulation with U46619 (Fig. 3A and C, respectively). Whilst pre-incubation with GF 109203X had no

ry I a a A O^O

significant effect on [Ca ]i mobilization by TPa or TP in response to primary agonist stimulation (Fig. 3B and D, respectively), it partially and significantly impaired desensitiza-tion of TPa such that in the presence of GF 109203X, secondary U46619 stimulation of TPa yielded a response of 36% relative to that of its primary response (Compare Fig. 3A and B, primary and secondary stimulations). Consistent with previous reports [50], pre- incubation of HEK.TPA328 cells with GF 109203X fully impaired agonist-induced desensitization in response to its secondary U46619 stimulation (Compare Fig. 3C versus Fig. 3D, secondary responses).

Hence, it is evident that TPA328 undergoes partial and transient desensitization that occurs through a GF 109203X-sensitive, PKC mechanism, whilst TPa undergoes almost complete and more sustained desensitization that is also partially sensitive to GF 109203X/PKC inhibition. As TPA328 is devoid of C-tail residues of both TPa and TPp>, it is evident that its target PKC-sensitive site(s) is located at sites other than within the C-tail domain of TPa or TPp>. In fact through computational analysis of the amino acid sequence of TPA328 we identified a putative PKC site within the second intracellular loop (IC)2 where Ser145 represents the predicted phospho-target residue [51]. Moreover, through site-directed mutagenesis of both TPA328 and TPe, to generate TPS145A'A328 and TP(iS145A , respectively, we have previously established that Ser145 actually corresponds to the GF 109203X-sensitive, PKC site that accounts for the partial agonist-induced desensitization of both TPA328 and TPp, [50]. Hence, herein, to investigate if Ser145 within the IC2 of TPa had any role to play in its homologous desensitization, site-directed mutagenesis was used to generate TPa

Primary stimulation of HEK. cells (Kd: 6.1± 1.3 nM SQ29,548; Bmax: 3.03± 0.04 pmol/mg protein; Table 1) with U46619 resulted in efficient [Ca2+]j mobilization (Fig. 3E) and IP3 generation (data not shown) confirming that the mutation per se did not affect the basic ligand binding or signalling properties of TPaS145A. However, unlike the wild type TPa (Fig. 3E), pre-stimulation with U46619 did not fully desensitize signalling by TPaS145A such that secondary agonist restimulation yielded a response of 24% relative to the primary response (Fig. 3E). Similar data were obtained when agonist-induced IP3 generation was analysed (data not shown). Consistent with previous reports, pre-stimulation of TPS145A,A328 with U46619 did not desensitize [Ca2+]j mobilization in response to its secondary stimulation with U46619 at 15 min following the initial stimulation (Fig. 3F, compare primary and secondary stimulations;p= 0.55). Taken together, these data confirm that whilst Ser145 within IC2 represents the phospho-target residue involved in partial U46619-induced, GF 109203X-sensitive homologous desensitization of TPA328 [50], similar to that of TPp>, Ser145 also plays a minor though significant role in the homologous desensitiza-tion of TPa. However, the major determinant(s) of U46619-induced desensitization of TPa signalling appear to be located

Fig. 2. U46619-mediated desensitization of signalling by TPa. Panels A-F: HEK.TPa (Panels A-C) or HEK.TPA328 (Panels D-F) cells, transiently co-transfected with pCMV:Gaq, were stimulated with 1 |M U46619 for 4 min as primary stimulation (Panels A and D). Thereafter, cells were washed to remove the U46619 as indicated by the horizontal arrow and were then re-stimulated with 1 | M U46619 either 15 min (Panels B and E) or 60 min (Panels C and F) following the primary U46619 stimulation, where ligands were added at the times indicated by the vertical arrows. Panel C, inset: HEK.TPa cells were stimulated with vehicle for 4 min as the initial primary stimulation (data not shown) prior to washout and subsequent stimulation with 1 |M U46619 at 60 min. Data presented are plotted as changes in intracellular Ca2+ mobilization (A[Ca2+]i, nM) as a function of time (second, s). Actual mean changes in were as

follows: PanelA: A[Ca2+]i= 174±8.8 nM; PanelB: A [Ca2+]i = 0 nM; Panel C: A[Ca2+]i=0 nM; PanelCinset: A[Ca2+]i= 191.5±25.5 nM; Panel D: A[Ca2+]i = 180± 5.6 nM; Panel E: A[Ca2+]i= 158±3.8 nM; Panel F: A[Ca2+]i= 178±6.7 nM. Panels G and H: HEK.TPa (Panel G) or HEK.TPA328 (Panel H) cells, transiently co-transfected with pCMV:Gaq or, as controls, HEK 293 cells (Panels G and H; HEK 293) were stimulated with 1 |M U46619 for 2 min (U46619); alternatively, cells were stimulated with 1 |M U46619 for 2 min, washed to remove the agonist and re-stimulated 60 min following the primary stimulation with 1 |M U46619 for 2 min (U4,U4). Levels of IP3 produced in ligand-stimulated cells relative to vehicle (HBS)-treated cells (basal IP3) were expressed as fold stimulation of basal (fold increase in IP3±S.E.M.; n =4). The asterisks indicate that the level of U46619-mediated IP3 generation was significantly reduced following secondary stimulation compared to that of the primary stimulation, or that the level of U46619-mediated IP3 generation was significantly lower in HEK 293 cells than in each of the above cell lines where ** and *** indicatesp<0.01 andp<0.001, respectively. Basal levels of IP3 in HEK.TPa (Panel G), HEK.TPA328 (Panel H) and HEK 293 cells (Panels G and H) were found to be in the range of 0.27-0.39 nmol/mg protein. Panel I: HEK.TPa cells co-transfected with pCMV:Gaq (lane 1) or the control vector pCMV5 (lane 2) were analysed by SDS PAGE (75 |g whole cell protein analysed/lane) followed by western blot analysis using anti-Gaq antibody (Gaq/11 (C-19): S.C. 392). Data presented is a representative immunoblot from four independent experiments. The relative position of the 46-kDa molecular size marker is indicated to the right of Panel I.

within its unique C-tail domain and that PKC may also have a partial role in that desensitization.

3.2. The role of GRKs in homologous desensitization of TPa

In addition to the second messenger-regulated protein kinases, members of the G protein coupled receptor kinase (GRK) family are widely reported to play a central role in

homologous desensitization of various GPCRs [36]. Moreover, we have recently established that GRK2/3 plays a critical role in the homologous desensitization of signalling by TPß [50]. Hence, herein, we investigated the role of the ubiquitously expressed GRK2/3, also referred to as the ß adrenergic kinases (ß-ARK) 1/2, in the homologous desensitization of signalling by TPa. Initially we examined the effect of over-expressing GRK2/ßARK1495-689, a dominant-negative form of GRK2/

Fig. 3. Effect of H-89 and GF 109203X on U46619-mediated desensitization of TPa signalling. Panels A-F: HEK.TPa (Panels A and B) or, as controls, HEK.TPA328 (Panels C and D) cells were pre-incubated for 10 min with either 10 |M H-89 (Panels A and C) or 50 nM GF 19203X (Panels B and D) prior to stimulation for 4 min with 1 | M U46619 (primary stimulation; Panels A, B, C and D); cells were then washed to remove the U46619 and were then re-stimulated at 15 min following the primary stimulation with 1 |M U46619 in the presence of 10 |M H-89 (Secondary stimulation; Panels A and C) or 50 nM GF 109203X (Secondary stimulation; Panels B and D). Data presented are representative profiles from at least four independent experiments and are plotted as changes in intracellular Ca + mobilization (A[Ca +]i, nM) as a function of time (second, s), where ligands were added at the times indicated by the arrows. Actual mean changes in U46619-induced [Ca2+]i mobilization (nM±S.E.M.) in response to primary and secondary U46619-stimulations were: Panel A: A[Ca2+]i=220±9.2 nM, 0 nM; Panel B: A[Ca2+]i=250± 12 nM, 90±8 nM; Panel C: A[Ca2+]i = 180± 10 nM, 165±8 nM; Panel D: A[Ca2+]i = 200± 9.8 nM, 200 ±12 nM. Panels E and F: HEK.TPaS145A (Panel E) or HEK.TPS145A,A328 (Panel F) cells were stimulated for 4 min with 1 |M U46619 (Primary stimulation; Panels E and F); thereafter, cells were washed and re-stimulated at 15 min following the primary stimulation with 1 |M U46619 (Secondary stimulation; Panels E and F). Actual mean changes in U46619-induced [Ca2+]i mobilizations (nM±S.E.M.) in response to primary and secondary U46619-stimulations were: Panel E: A[Ca2+]i = 157±3.8 nM, 37±4.2 nM; Panel F: A[Ca2+]i = 174±7.7 nM, 170±9.9 nM.

P>ARK1[39], on U46619-induced desensitization of TPa comparing it to its effect on TPp. signalling, acting as a reference. Transient over-expression no

significant effect on primary U46619-induced [Ca2+]i mobilization by TPa or TPp. (Fig. 4A and C, respectively). However, while over-expression of p>ARK1495-689 almost fully impaired desensitization of TPp. signalling following its secondary stimulation with U46619 (Compare Fig. 4C versus Fig. 4D, p = 0.0001), it had no effect whatsoever on agonist-induced

desensitization of signalling by TPa (Compare Fig. 4A versus Fig. 4B). Thus, while GRK2/3 (p>ARK1/2) play a major role in the homologous desensitization of TPp. they do not appear to participate in desensitization of TPa.

As Ser239 has been identified as a critical site in the GRK2/ 3-mediated homologous desensitization of TPp. [50], a site that is also common to and found within the IC3 domain of TPa but is clearly not targeted through GRK2/3 (Fig. 4B), we extended our studies to investigate whether it may have any

Fig. 4. Role of GRK2/3 on U46619-mediated desensitization of signalling by TPa and TP p. HEK.TPa (Panels A and B), HEK.TP|3 (Panels C and D) or HEK. TPaS239A (Panels E and F) cells each transiently co-transfected with either pCMV:Gaq plus pRK5:|3ARK1495-689 (+GRKDN) or, as controls, with pCMV:Gaq plus pRK5 (- GRKdn) were stimulated with 1 |j.M U46619 for 4 min (Panels A, C and E). Thereafter, cells were washed as indicated by the horizontal arrow and re-stimulated at 15 min following the primary stimulation with 1 |M U46619 (Panels B, D and F), where ligands were added at the times indicated by the vertical arrows. Actual mean changes in U46619-induced [Ca2+]i mobilization (nM±S.E.M.) were: Panel A: +GRKDN, A[Ca2+]i = 150±7.5 nM; -GRKDN, A[Ca2+]i= 165±9 nM; Panel B: +GRKDN, A[Ca2+];=0 nM; -GRKDN, A [Ca2+]; = 0 nM; Panel C: +GRKDN, A[Ca2+]; = 195± 11 nM; -GRKDN, A[Ca2+]; = 180±6.7 nM; Panel D: +GRKDN, A[Ca2+]j = 210± 12 nM; -GRKDN, A[Ca2+]i = 0 nM. Panel E: +GRKDN, A[Ca2+]i= 180±9 nM; -GRKDN, A[Ca2+]i = 176± 10.3 nM; Panel F: +GRKDN, A [Ca2+]j = 0 nM; - GRKDN, A[Ca2+]; = 0 nM.

role to play in homologous desensitization of TPa, such as through phosphorylation by other members of the GRK family or other Ser/Thr kinases. Hence, site directed mutagenesis was used to generate TPaS239A, thereby disrupting the phospho-target. Primary stimulation of HEK.TPaS239A cells (Kd: 7.1 ±3.2 nM SQ29,548; Bmax: 4.35±0.04 pmol/mg protein; Table 1) with U46619 resulted in mobilization (Fig. 4E) and IP3 generation (data not shown) confirming that the mutation per se did not affect the basic ligand binding or signalling properties of TPaS239A. Similar to that of the wild type TPa (Fig. 2A and B), pre-stimulation with U46619 fully desensitized signalling by TPaS239A (Fig. 4E and F). Moreover, over-expression of the dominant negative ßARK1495-689 had no effect on the primary signalling or on the level of agonist-induced homologous desensitization of TPaS239A (Fig. 4E and F, +GRK DN). These data clearly indicate that while Ser239 within IC3 is a

critical GRK2/3-targeted site in the homologous desensitization of TPß signalling, that site appears to be redundant in TPa and does not participate in desensitization of signalling of TPa.

3.3. Investigation of the role of the unique C-tail domain in homologous desensitization of TPa.

Hence, it is evident that while Ser145 within the IC2 domain plays a role, at least in part, in the GF 109203X, PKC-dependent homologous desensitization of signalling by TPa, from studies with TPA328 it appears that the major determinants of TPa desensitization rely on sequences within its unique C-tail domain. Analysis of the C-tail sequence of TPa reveals several Ser/Thr residues that might represent target sites for agonist-induced phosphorylation or desensitization, including Ser329, Ser331, Thr337 and Ser340 (Fig. 1). Thus, Ala-scanning and/or

deletion mutagenesis was used to disrupt those putative phospho-targets either individually to generate TPaS329A, TPaS331A, TPaA336, TPaT337A and TPaS340A or, there after, in various specific combinations as relevant. It is noteworthy that Ser329 has previously been identified as a target residue for PKA phosphor-ylation [33] while Ser331 was identified as a target for PKG phosphorylation [35] and Thr337 as a target for PKC phosphoryla-

tion [32]. Initially, Scatchard analyses confirmed that values obtained for the affinity (Kd) and maximal radioligand binding (Bmax) for each of the HEK 293 cell lines stably over-expressing the respective variant receptors compared well to those previously reported for wild type TPa (Table 1) and clearly indicated that the mutations per se did not affect their ligand binding properties.

Stimulation of HEK.TPaS329A cells yielded efficient [Ca2+] mobilization and IP3 generation in response to U46619 (Fig. 5A and C). Moreover, pre- stimulation of HEK.TPaS329A cells with U46619 completely desensitized [Ca2+]i mobilization and IP3 generation in response to secondary agonist stimulation (Fig. 5A-C). These data concur with previous results using H-89 demonstrating that the PKA inhibitor H-89 had no effect on the homologous desensitization of TPa (Fig. 3A).

Consistent with previous reports [35], agonist stimulation of TPaS331A yielded efficient [Ca2+]i mobilization and IP3

generation

(Fig. 5D and F). Pre-stimulation of HEK.TPaS331A cells with U46619 impaired, but did not fully abolish [Ca2 ]i mobilization and IP3 generation (Fig. 5D-F) in response to secondary stimulation with U46619. Specifically, the level of [Ca2 ]i mobilization by TPaS331A following secondary stimulation corresponded to 39% of that mobilized following primary U46619 stimulation. These data suggest that Ser331 plays a significant role in the homologous desensitization of TPa.

U46619-stimulation of TPaA336, a truncated variant of TPa eliminating two potential target phosphorylation sites at Thr337 and Ser340, yielded efficient [Ca2+]i mobilization and IP3 generation in response to U46619 (Fig. 5G and I, respectively). Pre-stimulation of TPaA336 with U46619 impaired but did not abolish [Ca2+]j mobilization and IP3 generation in response to its secondary agonist stimulation (Fig. 5H and I, respectively). Specifically, the level of [Ca2+]; mobilization by TPaA336 in response to secondary U46619 stimulation corresponded to 20% of its response (compare Fig. 5G versus Fig. 5H, p< 0.0003). Moreover, while both TPaT337A (Fig. 5J and L, respectively) and TPaS340A (Fig. 5M andN, respectively) yielded efficient [Ca2+] mobilization and IP3 generation in response to primary stimulation, signalling by TPaS340A was fully desensitized following secondary U46619-stimulation (Fig. 5M-O) while that of TPaT337A was only partially desensitized (Fig. 5J-L). Specifi-

2+ t337a

cally, the level of [Ca ] mobilization by TPa in response to secondary stimulation corresponded to 25% of that of its primary response. These data suggest that Thr337, a site previously established to act as a PKC site [32], plays a significant role in the homologous desensitization of TPa and are consistent with findings herein indicating that desensitiza-tion of TPa is sensitive to the PKC inhibitor GF 109203X (Fig. 3B).

Hence, taken collectively, our data generated from the systematic Ala-scanning mutagenesis studies indicate a specific role for Ser145 located in IC2, Ser331 and Thr337 located within the C-tail domain in the agonist-induced desensitization of TPa, with no apparent role for Ser239 within IC3 or Ser329 or Ser340

in that

desensitization. Thereafter, to exclude the possibility that the latter sites might act in concert or in synergy with the former sites in contributing to the desensitization of signalling of TPa, site-directed mutagenesis was used to disrupt those sites in specific combinations (Table 1). Initially we investigated the role of Ser239 in combination with Ser331 or Thr337 to generate TPaS239,331A and

s239,t337a

TPa , respectively. While primary stimulation of

TPaS239,331A with U46619 yielded efficient increases in [Ca2+] mobilization and IP3 generation (Fig. 6A and data not shown), the level of its agonist-induced desensitization was not significantly different to that of TPaS331A (compare Fig. 6B versus E, respectively). Specifically, the level of [Ca2+]; mobilization by

s239,331a

TPa ' in response to secondary stimulation corresponded to 36% of the primary U46619 response. Moreover, mutation of Ser239 in combination with Thr337 to generate TPaS239,T337A did not affect the level of desensitization relative to that of TPaT337A alone (data not shown). These data clearly indicate that mutation of

239 331

Ser , located within IC3, in combination with either Ser or Thr337 does not affect the overall level of TPa desensitization and hence, consistent with our previous data (Fig. 5A-C), we conclude that Ser239 does not contribute to TPa desensitization.

Similarly, we also investigated the role of Ser340 in combination with Ser331 or Thr337 to generate TPaS331,340A

t337,340a

and TPa a, respectively. While primary stimulation of TPaS331'340A and TPaT337'340A with U46619 yielded efficient increases in [Ca2+]; mobilization (Fig. 6C and E, respectively) and IP3 generation (data not shown), in each case the level of agonist-induced desensitization was not significantly different to that of their respective single mutants TPaS331A (compare Fig. 6D versus Fig. 5E, respectively) or TPaT337A (compare Fig. 6F versus Fig. 5K, respectively). Collectively these data confirm that mutation of Ser340 in combination with either

Ser or Thr337 does not affect the overall level of TPa desensitization and hence, consistent with our previous data (Fig. 5M-O), we conclude that Ser340 does not contribute to TPa desensitization. Moreover, through a similar series of

experiments to generate TPaS329,331A and TPaS329,T337A (data

Fig. 5. Role of the C-tail domain in agonist-induced desensitization of TPa signalling. Panels A-N: HEK.TPaS329A (Panels A and B), HEK.TPaS331A (Panels D and E), HEK.TPaA336 (Panels G and H), HEK.TPaT337A (Panels J and K) or HEK.TPaS340A (Panels M and N) cells, each co-transfected with pCMV:Gaq, were stimulated with 1 |M U46619 for 4 min as primary stimulation (Panels A, D, G, J and M). Thereafter, cells were washed to remove the U46619 as indicated by the horizontal arrow and were then re-stimulated 15 min following the primary stimulation with 1 | M U46619 (Panels B, E, H, K and N), where ligands were added at the times indicated by the vertical arrows. Data presented are plotted as changes in intracellular Ca2+ mobilization (A [Ca2+]i, nM) as a function of time (second, s). Actual mean changes in U46619-induced [Ca2+]i mobilization (nM±S.E.M.) were as follows: Panel A: A[Ca2+]i=228±7.7 nM; Panel B: A[Ca2+]i=0 nM; Panel D: A[Ca2+]i= 207±9.1 nM; PanelE: A[Ca2+]i = 80±6.8 nM; Panel G: A[Ca2+]i = 240± 14 nM; PanelH: A[Ca2+]i = 50±3.4 nM, Panel J: A[Ca2+]i=200±9.2 nM; PanelK: A[Ca2+]i± 49±5 nM, Panel M: A[Ca2+]i= 180±6.1 nM; Panel N: A[Ca2+]i±0 nM. Alternatively: HEK.TPaS329A (Panel C), HEK.TPaS331A (Panel F), HEK.TPaA336 (Panel I), HEK.TPaT337A (Panel L), or HEK.TPaS340A (Panel O) cells, co-transfected with pCMV:Gaq, were stimulated with 1 |M U46619 for 2 min (Panels C, F, I, L and O; U46619); cells were then washed and re-stimulated at 60 min following the primary stimulation with 1 |M U46619 for 2 min (Panels C, F, I, L and O; U4, U4). As controls, HEK293 cells were stimulated for 2 min with 1 |MU46619 (Panels C, F, I, Land O; HEK293). Levels of IP3 produced in ligand-stimulated cells relative to the vehicle (HBS) treated cells (basal IP3) were expressed as fold stimulation of basal (fold increase in IP3±S.E.M.; n=4). The asterisks indicate that the level of U46619-mediated IP3 generation was significantly reduced following secondary stimulation compared to that of the primary stimulation, or that the level of U46619-mediated IP3 generation was significantly lower in HEK 293 cells than in each of the above cell lines where *, ** and *** indicates p <0.05, p <0.01 and p <0.001, respectively. Basal levels of IP3 in the latter cell lines were found to be in the range of 0.27±0.06-0.39±0.09 nmol/mg protein.

not shown), consistent with our previous data involving TPaS329 (Fig. 5A-C) and H-89 (Fig. 3A), we also excluded a role for Ser329 in the homologous desensitization of TPa.

Thereafter, we investigated the overall effect of mutating Ser145,

331 337

Ser and Thr in combination (Table 1). Stimulation of HEK. TPaS331'T337A cells yielded efficient [Ca2+]i mobilization (Fig. 6G)

and IP3 generation (data not shown) in response to U46619. However, pre-stimulation with agonist only partially desensitized signalling by TPaS331'T337A (Fig. 6G and H). Specifically, the level

| i a Q'5'51 T'X'XH A

of [Ca ]i mobilization by TPa ' in response to secondary stimulation corresponded to 33% of that [Ca2+]i mobilized in response to the primary U46619 stimulation. Consistent with these

s331,t337,

data, the level of agonist-induced desensitization of TPa ' ' s340a was some 34% of its primary response (compare A[Ca2 +]i=195±12 nM for the primary response versus A[Ca2+]i=67 ±8.2 nM for the secondary response; Fig. 6I and J), further confirming a role for Ser331 and Thr337, but not Ser340, in TPa desensitization.

Consistent with the other mutant cell lines, U46619-induced

stimulation of HEK.TPaS145,S331,T337A cells, in which Ser145

within IC2 was also mutated along with Ser331 and Thr337, yielded efficient [Ca2+]i mobilization confirming that the mutation per se did not affect its basic signalling. However, pre-stimulation of TPaS145,331,T337A with U46619 only partially desensitized that signalling by such that its level of [Ca2 ]i mobilization following secondary agonist stimulation corresponded to 55% of its primary response (data not shown). Thus, taken together it appears that TPa is subject to GF 109203X-sensitive desensitization, where Ser145 and Thr337 [32] within its IC2 and C-tail domains, respectively, have been identified as the putative PKC-phospho-targets. In addition, Ser331, a site previously identified as a cGMP-dependent PKG site [35] has also been found to be critical for agonist-induced desensitization of TPa.

3.4. Agonist-induced phosphorylation of TPa in whole cells (in vivo)

It has been previously reported that TPa is subject to rapid agonist-induced phosphorylation that is dependent, at least in part, on PKC activation [20,25]. Hence herein, to establish whether TPa undergoes agonist-induced PKG phosphorylation, we investigated the effect of KT 5823 on TPa and TPaS331A phosphorylation in whole cells/in vivo. Initially, the specificity of the anti-HA antibodies to immunoprecipitate TPs from the respective cell lines, but not from the parent HEK 293 cells, was confirmed (Fig. 7C). The presence of a discrete band of approximately 39 kDa and a broad diffuse band of 46-60 kDa representing their non-glycosylated and glycosylated forms, respectively, were evident inthe TPa andTPaS331A immunoprecipitates (Fig. 7C, lanes 1 and 2, respectively) whilst no immunoreactive bands were evident in immunoprecipitates prepared from non-transfected HEK 293 cells (Fig. 7C, lane 3). Stimulation of HEK.TPa cells with U46619 led to 5-7-fold increase in the level of TPa phosphorylation relative to vehicle treated cells (Fig. 7A, lanes 1 and 2). Moreover, pre-

incubation with KT 5823 reduced the level of agonist-induced TPa phosphorylation by 2.5 fold (Fig. 7A, lane 3). Consistent with this, while stimulation of HEK.TPaS331A cells with U46619 led to 2.5fold increase in the level of TPaS331A phosphorylation relative to basal levels in vehicle treated cells (Fig. 7B, lanes 1 and 2), agonist-induced phosphorylation of TPaS331A was not significantly affected by preincubation with KT 5823 (Fig. 7B, lanes 1 and 2). Moreover, TPaS331,T337A failed to undergo significant U46619-induced phosphorylation above basal levels (Fig. 7D).

3.5. Investigation of the role of Nitric oxide signalling in homologous desensitization of TPa

To further address whether PKG plays an important role in homologous desensitization of TPa, we examined the effect of various pharmacological inhibitors of the NO/cGMP signalling cascade on U46619-induced desensitization of TPa, and as a control, on TPp signalling. While pre-incubation of HEK.TPa cells with KT 5823 (50 nM) did not interfere with their ability to mobilize [Ca2+]j per se (Compare Fig. 2A versus Fig. 8A), it partially inhibited U46619-induced desensitization of TPa signalling in response to secondary stimulation with U46619 (Fig. 8B). Specifically, in the presence of KT 5823, the level of [Ca2+] mobilized by TPa in response to its secondary stimulation with U46619 corresponded to approximately 38% of its primary response (Fig. 8A and B). On the other hand, pre-incubation of HEK.TPp cells with KT 5823 had no significant effect on U46619-induced [Ca2+]j mobilization in response to its primary stimulation and did not impair agonist-induced desensitization of signalling by TP p. (Insets to Fig. 8A and B).

A potent stimulator of PKG is the activation of nitric oxide synthase (NOS) promoting the conversion of l-arginine to nitric oxide. NO, in turn, activates the soluble form of guanylate cyclase (sGC) leading to a rise in intracellular cGMP and activation of PKG [52-55]. Hence, we also investigated the actions ofsGC and NOS inhibitors LY 83583 and l-NAME, respectively, on TPa desensitization. While pre-incubation of TPa with LY 83583 did not interfere with its ability to mobilize [Ca2+]j per se (Compare Fig. 2A versus Fig. 8C), it partially inhibited desensitization of TPa signalling (Fig. 8D). In the presence ofLY 83583, the level of [Ca2+]j mobilized following secondary stimulation corresponded to approximately 30% of its primary response. LY 83583, on the other hand, had no significant effect on [Ca2+]j mobilization by TPp following its primary or secondary stimulation with U46619 (Insets to Fig. 8C and D). Similarly, the NOS inhibitor l-NAME had no effect on primary U46619-induced signalling by TPa (Compare Fig. 2A versus Fig. 8E) but significantly inhibited desensitization of TPa signalling (Fig. 8E and F). Specifically, in

Fig. 6. Role of the C-tail domain in agonist-induced desensitization of TPa signalling. Panels A-J: HEK.TPaS329,331A (Panels A and B), HEK.TPaS331,340A (Panels C and D), HEK.TPaT337'S340A (Panels E and F), HEK.TPaS331'T337A (Panels G and H), or HEK.TPS331'T337'S340A (Panels I and J) cells, each co-transfected with pCMV: Gaq, were stimulated with 1 |j.M U46619 for 4 min as primary stimulation (Panels A, C, E, G and I). Thereafter, cells were washed to remove the U46619 as indicated by the horizontal arrow and were then re-stimulated 15 min following the primary stimulation with 1 |M U46619 (Panels B, D, F, H and J), where ligands were added at the times indicated by the vertical arrows. Data presented are plotted as changes in intracellular Ca + mobilization (A [Ca +]j, nM) as a function of time (second, s). Actual mean changes in U46619-induced [Ca2+] mobilization (nM±S.E.M.) were as follows: Panel A: A[Ca2+]j= 176±9.4 nM; Panel B: A[Ca2+]j=63±4.4 nM; PanelC: A[Ca2+]j = 160±7.2 nM; PanelD: A[Ca2+]j=60± 10 nM; PanelE: A[Ca2+]j= 179± 11 nM; PanelF: A[Ca2+]j=40±3.4 nM; PanelG: A[Ca2+]j = 181±6.5 nM; Panel H: A[Ca2+]j=60±7.1 nM, Panel I: A[Ca2+]j = 195± 12 nM; Panel J: A[Ca2+]j = 67±8.2 nM.

Fig. 7. U46619-mediated phosphorylation of TPa. Panels A and B: HEK.TPa (Panel A) and HEK.TPaS331A cells (Panel B), metabolically labelled with [32P] orthophosphate, were pre-incubated for 15 min with 50 nM KT 5923 (Panels A and B, lane 3) prior to incubation for 10 min with the vehicle HBS (Panels A and B; lane 1) or 1 | M U46619 (Panels A and B; lanes 2 and 3). Immunoprecipitates were resolved by SDS-PAGE and electroblotted onto PVDF membranes. Blots were subject to PhosphorImage analysis and the intensities of U46619-mediated TPa phosphorylation relative to basal phosphorylation in the presence of HBS were determined and expressed in arbitrary units as follows: TPa: 1 |M U46619, 5-fold increase; 1 |M U46619 plus 50 nM KT 5923, 2-fold; TPaS331: 1 |M U46619, 2-fold increase; 1 |M U46619 plus 50 nM KT 5923, 1.5-fold. Panel C: HEK.TPa, HEK.TPaS331A and, as a control, HEK 293 cells (Panel C, lanes 1-3, respectively) were subject to immunoprecipitation with anti-HA antibody 101R and immunoblots screened using the anti-HA 3F10 horseradish peroxidase-conjugated antibody followed by chemiluminescence detection. The positions of the molecular weight markers (kDa) are indicated to the left and right of the panels A and C. The arrow to the left of Panel A indicates the approximate position of the phosphorylated TPa. Data presented are representative of three independent experiments. Panel D: HEK. TPa , HEK.TPaS331A and HEK.TPaS331'T337A cells, metabolically labelled with [32P]orthophosphate, were stimulated for 10 min with 1 |M U46619 (+) or with an equivalent volume of the vehicle HBS (—). Thereafter, HA epitope-tagged TPp receptors were immunoprecipitated as previously described. Blots were subject to PhosphorImage analysis and the intensities ofU46619-mediatedTPa phosphorylation relative to basal levels in the presence of HBS were determined and expressed in arbitrary units (Phosphorylation; fold increase, n=3).

the presence of l-NAME, the level of [Ca2+] mobilized by TPa following secondary U46619 stimulation corresponded to approximately 40% of the primary signal. l-NAME had no significant effect on [Ca2+] mobilization by TPp. following its primary or secondary signalling (Insets to Fig. 8E and F).

Thereafter we sought to establish whether any or all of the NOS isoforms are actually expressed in HEK 293 cells. RT-PCR permitted detection of endothelial (e)NOS and neuronal (n)NOS but not inducible (i)NOS transcripts (Fig. 9A) and western blot analysis also confirmed the abundant endogenous expression of eNOS (Fig. 9B) in that cell line. Moreover, consistent with the role of NO/PKG in agonist-induced desensitization of TPa,

stimulation of HEK.TPa cells (Fig. 9C) or indeed non-transfected HEK 293 or HEK.TPp. cells (data not shown) with U46619 (1 pM, 5 min) led to the rapid phosphorylation-dependent activation of eNOS, as detected using the anti-phospho eNOSSer1177 antisera (Fig. 9C). In fact the level of eNOS phosphorylation in response to U46619 was comparable, if not greater, than those levels detected in response to stimulation of cells with thrombin (Fig. 9C), an agent that has recently been established to lead to rapid phosphorylation of eNOS at Ser1177/1179 leading to its activation [56].

Hence taken collectively, data herein indicate that agonist-induced densensitization of TPa signalling involves two key

Fig. 8. Effect of KT 5823, LY 83583 and l-NAME on U46619-mediated desensitization of TPa signalling. Panels A-F: HEK.TPa (Panels A-F) or, as controls, HEK.TPp (Insets A-F) cells were pre-incubated for 15 min with either 50 nM KT 5823 (Panels and Insets A and B), 1 |M LY 83583 (Panels and Insets C and D) or 1 |M l-NAME (Panels and Insets E and F) prior to stimulation with 1 |M U46619 (primary stimulation; Panels and Insets A, C and E); cells were then washed to remove the U46619 as indicated by the horizontal arrow and were then re-stimulated at 15 min following the primary stimulation with 1 |M U46619 in the presence of 50 nM KT 5823 (Panels and Insets B), 1 |M LY 83583 (Panels and Insets D) or 1 |M l-NAME (Panels and Insets F). Data presented are representative profiles from at least four independent experiments and are plotted as changes in intracellular Ca2+ mobilization (A[Ca2+]j, nM) as a function of time (second, s), where the ligands were added at the times indicated by the arrows. Actual mean changes in U46619-induced [Ca2+] mobilization (nM±S.E.M.) were: Panel A: A[Ca2+]j=160±7.9 nM; Panel B: A[Ca2+]j=60±7 nM; Panel C: A[Ca2+]j=210± 12 nM; Panel D: A[Ca2+]j = 58±6 nM; Panel E: A[Ca2+]j=200±9 nM; Panel F: A[Ca2+]i = 80±5.9 nM. Inset A: A[Ca2+]j = 170±25 nM; Inset B: A[Ca2+]j=0 nM; Inset C: A[Ca2+]j = 185±26 nM; Inset D: A[Ca2+]j=0 nM; Inset E: A[Ca2+]j= 220±40 nM; Inset F: A[Ca2+]j=0 nM.

second messenger regulated systems, one involving PLCp-regulated PKC activation and phosphorylation of Ser145 and Thr337 and another involving eNOS phosphorylation and activation leading the NO/cGMP-regulated PKG phosphor-

ylation of Ser331. Hence, in addition to the agonist-mediated PKC-feedback mechanism, engagement of the NO-cGMP signalling pathway through U46619-activation of the NOS/ sGC/PKG pathway plays a significant role in inducing

Fig. 9. Analysis of nitric oxide synthase (NOS) expression and activation in HEK 293 cells. Panel A: RT-PCR analysis of 1° cDNA from HEK 293 cells as template and primer pairs to selectively amplify neuronal (n)NOS (lane 1), endothelial (e)NOS (lane 3) or inducible (i)NOS (lane 5); negative controls where individual primer pairs were added to the reaction without template cDNA are shown in lanes 2 (nNOS), 4 (eNOS) and 6 (i"NOS). Molecular weight markers are shown in lanes M, and the position of the 350- and 517-bp markers are indicated to the left and right of the panel. PCR products corresponding to (n)NOS (lane 1; 456 bp) and(e)NOS (lane 3; 354 bp) were specifically amplified. Panel B: western blot analysis of eNOS expression. Aliquots (100 |ig and 200 |ig, respectively) of HEK 293 cell protein was analysed by SDS-PAGE followed by western blot analysis using anti-eNOS antibody. The relative position of the 120-kDA marker is indicated to the left of the panel. Results shown in Panel C: Phosphorylation of eNOS in HEK.TPa cells. HEK.TPa cells were incubated at 37 °C for 5 min with either vehicle, 1 |M U46619 or 10 U/ml Thrombin. Whole cell protein was analysed by SDS-PAGE followed by western blot analysis using anti-phospho eNOS1177 and anti-eNOS, as indicated. The relative position of the 120-kDA marker is indicated to the left of the panel. Results shown in Panels A-C are representative of 3 independent experiments.

homologous desensitization of TPa, but not TPp>, in HEK 293 cells.

4. Discussion

Agonist-induced homologous desensitization represents a critical regulatory component of the integrated signalling modality of diverse members of the GPCR superfamily and provides a mechanism of terminating, at least in part, the primary signalling response [29,30]. Phosphorylation of the GPCR, typically on Ser/Thr residues within its intracellular loop (IC) or carboxyl-terminal (C) tail domains, initiates such desensitization uncoupling the GPCR from its cognate G-protein(s). The G protein-coupled receptor kinases (GRKs), of which GRK1-7 have been identified, can only phosphorylate the agonist-engaged conformationally active GPCR and hence are typically associated with homologous desensitization [29,30,57]. Recruitment of p>-arrestin1/2 adaptor(s) to the GRK-phosphorylated receptor sterically hinders G-protein coupling [30,36,58]. Moreover, depending on the GPCR, p>-arrestin recruitment may promote sequestration and internaliza-tion of the desensitized receptor into intracellular compartments, such as through clathrin- or caveolin-dependent mechanisms [30,36,58], ultimately leading to either depho-sphorylation and recyclization (resensitization), degradation or recruitment of the GPCR into scaffolds/microdomains for participation in other signalling systems [30,36,58-60]. The second messenger-regulated kinases, such as cAMP-dependent protein kinase (PK) A and diacylglycerol-regulated PKC, on the other hand, can phosphorylate GPCRs even in the absence of the specific agonist and, hence, are most frequently associated

with heterologous desensitization or cross-talk between different signalling systems, but may also contribute to homologous desensitization in certain cases [29,30].

The platelet-derived prostanoid TXA2 regulates a range of physiologic responses mainly within the vasculature and is implicated as a mediator of a host of vascular disorders [35,12,13]. TXA2 acts as a potent evanescent autocoid with an estimated half-life of 30 s, being rapidly hydrolysed to its inactive metabolite TXB2 [1]. However, in such cases as thrombosis, vascular occlusion and hypertension where TXA2 is produced in large quantities over prolonged periods [61,62], turning off the TXA2 signal may also occur at the level of its receptor TP [29,30,36]. TP also mediates many the actions ofthe isoprostane 8-iso-PGF2a that is generated in abundance in situations of oxidative injury [10,11]. Furthermore, in primates TXA2 signals through two TP receptor isoforms, termed TPa and TPp>, adding to the complexity whereby the responses to TXA2 and, indeed, 8-iso-PGF2a may be regulated in humans. Hence, bearing this in mind, we have recently investigated the mechanisms whereby the cellular responses of TXA2 mediated by the TPp. isoform of the human TXA2 receptor (TP) are regulated and established that GRK2/3 plays a major role in that homologous desensitization, with an additional minor role for PKC [50]. Specifically, it was found that agonist-induced GRK2/3 phosphorylation of Ser357 within the unique C-tail domain of TPp. provides a docking site to recruit and orientate GRK2/3, facilitating its phosphorylation of Ser239 within IC3, impairing Gq coupling, desensitizing TPp. signalling and promoting its agonist-induced internalization [50]. As TPa and TPp. diverge within their C-tail domains, sequences such as Ser357 determined to be critical for desensitization of TPp. are

not found in TPa clearly implying that the mechanism of agonist-induced desensitization of the TP isoforms may indeed differ. Moreover, consistent with that hypothesis, TPp but not TPa undergoes agonist-induced internalization largely through a GRK/p-arrestin-dependent mechanism [37,63]. Hence, the overall aim of the current study was to investigate the molecular mechanism(s) whereby the cellular responses of TXA2 mediated by the TPa isoform are specifically regulated thereby aiming to elucidate the mechanism of homologous desensitiza-tion of TPa.

It was initially confirmed that, similar to TPp, TPa was subject to complete agonist-induced homologous desensitiza-tion as assessed by measurement of [Ca2+]j mobilization and IP3 generation in response to the TXA2 mimetic U46619. Desensitization of TPa is sustained showing no significant recovery following agonist wash-out and re-stimulation some 60 min following the primary response. On the other hand and consistent with our previous data [50], TPA328, the truncated derivative devoid of the divergent residues of TPa and TPp, underwent partial and transient desensitization with its signalling fully recovered to that of its primary response at 60 min. Moreover, desensitization of TPA328 was blocked by pre-incubation with the PKC inhibitor GF 109203X or by mutation of Ser145 within IC2 to generate TPS145A,A328 [50]. While the PKA-inhibitor H-89 had no effect on TPa signalling, GF 109203X quite significantly impaired agonist-induced desensi-tization such that the magnitude of the secondary U46619-response by TPa was 36% of its primary response in the presence of GF 109203X. Mutation of Ser145 within IC2 to generate TPaS145A, a residue common to both TP isoforms and previously identified as the critical PKC phospho-target of both TPA328 and TPp [50], established that the level of desensitization of TPaS145A was partially impaired, signalling to 24% of its primary response following secondary agonist stimulation. Hence, collectively these data suggested that whilst Ser145 within IC2 represents the PKC phospho-target residue involved in transient and partial desensitization of TPA328 [50], similar to that of TPp [50] Ser145 also plays a minor though significant role in the homologous desensitization of TPa. However, these data also suggest that the major determinant(s) of desensitiza-tion of TPa signalling appear to be located within its unique C-tail domain.

As stated, GRK2/3 plays a major role in both agonist-induced desensitization of signalling and internalization of TPp

239 357

where Ser and Ser within the IC3 and C-tail domains, respectively, were identified as the critical functional targets [50]. Hence, herein, we initially investigated the role of the ubiquitously expressed GRK2/3 in the homologous desensitiza-tion of signalling by TPa by examining the effect of over-expressing GRK2/pARK1495-689, a dominant negative form that impairs GRK2/3 membrane translocation and activation [39]. While pARK1495-689 almost fully impaired desensitization by TPp, it had no significant effect on primary or secondary signalling by TPa. Moreover, mutation of Ser239 within the IC3 domain of TPa, a critical GRK2/3-target site within TPp [50], established that TPaS239A under-went complete agonist-induced desensitization of signalling similar to the wild type

TPa. Collectively these data clearly indicate that while Ser239 within IC3 is a critical GRK2/3-targeted site in TPp, it appears to be redundant in TPa and does not participate in its homologous desensitization and hence that residue is unlikely to be a target for phosphorylation/desensitization by other GRKs, such as by the ubiquitously expressed GRK5 or GRK6

[64]. It is also noteworthy that the sequences DS239E and DS357R

are identically found in both the human and chimp TPp sequences but neither phosphotarget sites are found in the single bovine, mouse or rat TP orthologs suggesting that both Ser239 and Ser357 have co-evolved to confer a common function to the TPp isoforms in primates and clearly account for the redundant role of Ser239, in functional terms, within the TPa primate orthologs. The finding herein that GRK/p-arrestins do not play a significant role in homologous desensitization of signalling by TPa are entirely consistent with reports that TPa does not undergo significant agonist-induced internalization [37].

Hence, while Ser145 within the IC2 domain plays a role in the GF 109203X-sensitive, PKC-dependent homologous desensitization of signalling by TPa, from studies with TPA328 it was evident that the major determinants of TPa desensitization rely on sequences within its unique C-tail domain. The C-tail sequence of TPa has several Ser/Thr residues that might represent target sites for agonist-induced phosphorylation or desensitization, including Ser329, Ser331, Thr337 and Ser340. Moreover, several of those sites have been identified as targets of heterologous desensitization [32-35]. Hence, Ala-scanning and/or deletion mutagenesis was used to systematically disrupt those putative phospho-targets either individually, to generate

s329a s331a a336 t337a s340a

TPa , TPa , TPa , TPa and TPa , or in specific combinations as relevant and the functional consequences of those mutations on ligand binding, primary signalling and agonist-induced desensitization were investigated. Through these mutational studies, it was established that neither Ser329 nor Ser340, mutated either alone or in combination with other putative phospho-target residues, had any role to play in the mediation of homologous desensitization of TPa. It is noteworthy that Ser329 of TPa had been previously identified as a target for PKA phosphorylation and heterologous desensitization [33]. Hence, our data herein eliminating a role for Ser329 in the homologous desensitization of TPa are in agreement with that study [33] and with the fact that the PKA inhibitor H-89 had no effect on the desensitization of TPa.

331 337

On the other hand, mutation of Ser or Thr alone or in combination with each other or with Ser145 led to partial or near complete impairment of agonist-induced desensitization of TPa. Specifically, the level of [Ca2+]j mobilization by

O'l'i 1 A T'X'XH A

TPa and TPa following their secondary stimulation corresponded to 39% and 25% of their primary responses, respectively. Moreover, secondary agonist stimulation of

TPaS331,T337A or TPaS145,S331,T337A yielded responses corresponding to 33% and 55% of their primary responses, respectively. The involvement of Thr337, a site previously identified as a target for PKC phosphorylation [32], is consistent with a significant role for PKC in the agonist-induced desensitization of TPa and clearly indicates that both Ser145 and Thr337 within the IC2 and C-tail domains are functional

PKC targets. In terms of Ser331, it was previously established that TPa, but not TPp>, is subject to desensitization in response to the potent vasodilator nitric oxide (NO) that occurs through direct type 1 PKG-mediated phosphorylation where Ser331 was identified as the PKG phospho-target [35]. Hence, these studies [35] and data herein suggested a direct role for the NO/PKG-mediated signalling in both the heterologous and homologous desensitization of TPa. Consistent with this, KT 5823, the selective PKG inhibitor, significantly reduced U46619-induced phosphorylation of TPa but had no affect on phosphorylation of TPaS331A or TPaS331,T337A. Moreover, pre-incubation with KT 5823 reduced agonist-induced desensitization of signalling by TPa such that in its presence the level of [Ca2 ]i mobilized following secondary agonist stimulation corresponded to ~ 38% of the primary response. Similarly, the sGC and NOS inhibitors LY 83583 (30%) and l-NAME (40%), respectively, yielded significant impairments in agonist-induced desensitization of TPa. On the other hand, consistent with previous findings that TPp. is not a target for NO-mediated heterologous desensitization or PKG phosphorylation [35], neither KT 5823, LY 83583 nor l-NAME had any measurable effect on TPp. signalling following its primary or secondary agonist stimulation. Phosphorylation of eNOS at Ser1177 (Human)/1179 by

phosphatidylinositol 3-kinase (PI3-K)-dependent AKT or, as more recently demonstrated, by Gq/PLCp>/PKC8 pathway, such as in response to vascular endothelial growth factor [65] or thrombin stimulation, respectively [56], plays a critical role in eNOS activation. Herein, RT-PCR and/or western blot analysis confirmed the endogenous expression of both eNOS and nNOS in HEK 293 cells. Moreover, agonist (U46619)-stimulation of HEK.TPa cells, or indeed of non-transfected HEK 293 or HEK. TPp. cells (data not shown), led to the rapid phosphorylation of endogenous eNOS at Ser1177 to levels comparable, if not greater, than those in response to thrombin stimulation.

Hence, based on data presented herein, we propose a model for the mechanism of homologous desensitization of TPa signalling, as outlined in Fig. 10. Agonist activation of TPa and TPp. primarily couple to Gq-dependent PLCp> activation leading to increases in IP3 generation and mobilization of [Ca2 ]i. This, in turn, triggers the rapid and dynamic activation calmodulin-sensitive NO synthases (NOS), such as through Ser1177 phosphorylation and activation of eNOS, promoting the synthesis of NO from l-Arg. NO in turn activates sGC leading to a rise in intracellular cGMP and, in turn, activation of cGMP-dependent PKG, or cGK [54, 55]. Type 1 PKGa/(S, in turn, phosphorylates and inhibits signalling by TPa where Ser331 represents the phospho-target. As TPp. is not a target of PKG phosphorylation [35], it is not subject to direct desensitization by this mechanism. Additionally, TPa/Gq/PLCp mediated-activation of PKC can phosphorylate Ser145 and Thr337 contributing, in part, to the desensitization of TPa signalling. Thus, agonist-induced homologous desensitization of TPa is mediated, at least in part, through both PKG and PKC-catalysed mechanisms where Ser and Ser /Thr located within IC2 and C-tail domains have been identified as the respective phospho-targets. It is noteworthy that collective mutation of those three sites to generate TPaS145,S331,T337A did not eliminate

Fig. 10. Proposed mechanism of U46619-mediated, homologous desensitization of TPa signalling. Ligand (U46619)-engagement of TPa stimulates Gq-mediated PLCp activation leading to increases in IP3 generation and mobilization of [Ca2+]i. The latter rise in [Ca2+]i triggers activation of endothelial nitric-oxide synthase (eNOS) promoting conversion of l-arginine to nitric oxide (NO). U46619-mediated TPa activation can also lead to increased phosphorylation of eNOS at Ser1177, through an as yet unknown mechanism, contributing to eNOS activation (not shown). NO in turn activates the soluble form of guanylyl cyclase (sGC) leading to a rise in intracellular cGMP and, in turn, activation of protein kinase (PK) G. PKG, in turn, phosphorylates and inhibits signalling by TPa where Ser331 has been identified as the target residue for PKG phosphorylation. U46619-mediated desensitization of TPa signalling may be inhibited by the NOS inhibitor l-NAME, by the sGC inhibitor LY 83583 or by the PKG inhibitor KT5823. Additionally, TPa/Gaq/PLCp mediated-increase in DAG leads to activation of PKC; PKC can phosphorylate TPa at Ser145 and Thr337 yielding partial U46619-mediated desensitization of TPa signalling. Thus, U46619-induced homologous desensitization of TPa is partially mediated through both PKG and PKC-catalysed mechanisms where Ser331 and Ser145/Thr337 have been identified as the respective phospho-targets.

desensitization of TPa, clearly indicating that other factors may also contribute to agonist-induced desensitization of TPa.

The most striking feature of the proposed mechanism of homologous desensitization of TPa presented herein is that it mainly involves feedback loops whereby the second messenger kinases PKC and PKG activated in response to its primary signalling bring about desensitization through phosphorylation of critical residues within its IC2 and C-tail domains. The involvement of PKC in the agonist-induced desensitization of TPa is in keeping with studies by Spurney et al., involving the single murine TP where it was determined that mutation of three putative PKC phospho-Ser residues within its C-tail domain significantly attenuated that desensitization [66]. The endogenous vasodilator NO is critical for the maintenance of normal blood pressure [67-69] and protects the blood vessel wall by inhibiting platelet aggregation, secretion, adhesion and fibrino-gen binding to its receptor glycoprotein Ilb/IIIa [70]. Many intracellular events involving inositol turnover and Ca2+ mobilization, that contributes to vasoconstriction and platelet activation status for example, are regulated by the NO/cGMP signalling cascade, mainly through the direct- or indirect-action of PKG. Amongst the key molecular targets of NO-regulated PKG thus far identified is the IP3 receptor [71], thereby inhibiting PLC activation and IP3-evoked Ca2+ release from intracellular stores [72], IP3 receptor-associated cGMP kinase substrate or IRAG [73,74], myosin binding substrate [75] and the vasodilator-stimulated phosphoprotein or VASP [76].

Moreover, in recent studies, it was also established that PKG1a indirectly attenuates signalling by the thrombin/protease-activated receptor-1 (PAR-1) through direct activation of regulator of G-protein signalling-2 (RGS-2), increasing the GTPase activity of Gq [77,78]. Furthermore, as stated, through further recent studies it has been demonstrated that thrombin

1177/1179

can directly lead to eNOS phosphorylation at Ser in

bovine aortic endothelial cells in a mechanism involving Gq/ PLCp>/PKC8 phosphorylation and Ca2+-dependent activation of eNOS [56]. Whilst our data have indeed confirmed that both thrombin and U46619 can readily lead to eNOS phosphorylation at Ser1177 in HEK 293 cells, we have not explored whether that eNOS activation occurs through PKC8 or through the classic PI-3K or other mechanism [56].

Whilst our finding herein that agonist-induced desensitiza-tion of TPa entails a closed feedback loop involving eNOS phosphorylation/activation, NO generation and subsequent PKG phosphorylation is a mechanism not typically associated with homologous desensitization of GPCRs in general, it is also in agreement with our previous studies demonstrating that TPa is indeed a direct target of NO/PKG phosphorylation and heterologous desensitization [35]. Moreover, increased NO synthesis and release is known to occur in response to TXA2 in other cell/tissue types, such as within the pulmonary micro-vasculature where the local release ofNO acts as a compensatory mechanism in response to the proinflammatory and vasoactive effects of TXA2 [79]. NO regulation and synthesis via activation of the Ca2+ dependent eNOS or nNOS isozymes can occur in response to a diverse array of agonists that signal through GPCRs, including the widely documented bradykinin and muscarinic acetylcholine receptors signalling pathways, in addition to TP [70,80]. Moreover, direct physical interaction between the bradykinin B2 receptor and eNOS yields a reversible inhibitory complex that is readily dissociated on ligand binding and/or elevations in [Ca2+]i, thereby providing another counter-regulatory mechanism between GPCR/Ca2+-and NOS/cGMP-regulated signalling systems [81]. While we have not been able to find any evidence for a direct interaction between eNOS and TPa, or indeed TPp>, such as through co-immunoprecipitation studies (Reid and Kinsella, unpublished), we cannot rule out the possibility that additional mechanisms other than direct phosphorylation of TPa may contribute to the PKG-, or indeed PKC-, induced desensitization of TPa. For example, it has been demonstrated that PKG1a can be phosphorylated both in vitro and in vivo by PKC at Thr58, generating a partially active kinase that is more sensitive to cGMP levels [82]. Whether TXA2-mediated PKC activation contributes to PKG1a activation and TPa phosphorylation/ desensitization through the latter mechanism remains to be investigated.

Hence, in summary, it is indeed evident that there are fundamental differences in the mechanisms of homologous desensitization of TPa and TPp. adding further to the increasing lines of evidence that the TP isoforms may indeed have distinct physiologic roles. A critical difference between those two mechanisms is that agonist-mediated desensitization of TPa does not involve its internalization while desensitization of TPp.

involves its recruitment to an intracellular compartment(s), making it susceptible to down-regulation or, alternatively, potentially available for participation in other signalling systems [30,36,58-60]. The regulation of TPa signalling provides for a dynamic shut down or desensitization of signalling of the potent autocoid TXA2 by the counter-regulatory NO, while TPp. remains unchecked by that mechanism. These findings coupled to the fact that TPa, but not TPp>, is also a direct target of prostacyclin-induced heterologous desensitization [33] provide further evidence for a critical role for TPa in vascular hemostasis whilst the role of TPp>, if any, in that essential physiologic process remains to be defined.

Acknowledgements

This research was supported by the Wellcome Trust, The Health Research Board (Ireland) and Science Foundation Ireland.

References

[1] S. Narumiya, Y. Sugimoto, F. Ushikubi, Prostanoid receptors: structures, properties, and functions, Physiol. Rev. 79 (1999) 1193-1226.

[2] Y. Sugimoto, Physiological functions of prostanoid receptors and their subtypes, Nippon Yakurigaku Zasshi 115 (2000) 131-141.

[3] K. Hanasaki, T. Nakano, H. Arita, Receptor-mediated mitogenic effect of thromboxane A2 in vascular smooth muscle cells, Biochem. Pharmacol. 40 (1990) 2535-2542.

[4] G.W. Dorn II, M.W. Becker, M.G. Davis, Dissociation of the contractile and hypertrophic effects of vasoconstrictor prostanoids in vascular smooth muscle, J. Biol. Chem. 267 (1992) 24897-24905.

[5] S. Ali, M.G. Davis, M.W. Becker, G.W. Dorn II, Thromboxane A2 stimulates vascular smooth muscle hypertrophy by up-regulating the synthesis and release of endogenous basic fibroblast growth factor, J. Biol. Chem. 268 (1993) 17397-17403.

[6] A.W. Ashton, R. Yokota, G. John, S. Zhao, S.O. Suadicani, D.C. Spray, J.A. Ware, Inhibition of endothelial cell migration, intercellular communication, and vascular tube formation by thromboxane A(2), J. Biol. Chem. 274 (1999) 35562-35570.

[7] A.W. Ashton, Y. Cheng, A. Helisch, J.A. Ware, Thromboxane A2 receptor agonists antagonize the proangiogenic effects of fibroblast growth factor-2: role of receptor internalization, thrombospondin-1, and alpha(v)beta3, Circ. Res. 94 (2004) 735-742.

[8] A.W. Ashton, J.A. Ware, Thromboxane A2 receptor signaling inhibits vascular endothelial growth factor-induced endothelial cell differentiation and migration, Circ. Res. 95 (2004) 372-379.

[9] F. Ushikubi, Y. Aiba, K. Nakamura, T. Namba, M. Hirata, O. Mazda, Y. Katsura, S. Narumiya, Thromboxane A2 receptor is highly expressed in mouse immature thymocytes and mediates DNA fragmentation and apoptosis, J. Exp. Med. 178 (1993) 1825-1830.

[10] B.T. Kinsella, D.J. O'Mahony, G.A. Fitzgerald, The human thrombox-ane A2 receptor alpha isoform (TP alpha) functionally couples to the G proteins Gq and G11 in vivo and is activated by the isoprostane 8-epi prostaglandin F2 alpha, J. Pharmacol. Exp. Ther. 281 (1997) 957-964.

[11] L.P. Audoly, B. Rocca, J.E. Fabre, B.H. Koller, D. Thomas, A.L. Loeb, T.M. Coffman, G.A. FitzGerald, Cardiovascular responses to the iso-prostanes iPF(2alpha)-III and iPE(2)-III are mediated via the thromboxane A(2) receptor in vivo, Circulation 101 (2000) 2833-2840.

[12] E.A. Lianos, B.A. Bresnahan, Effect of thromboxane A2 inhibition and antagonism on prostaglandin and leukotriene synthesis in glomerular immune injury, J. Lab. Clin. Med. 134 (1999) 478-482.

[13] N.N. Kahn, H.S. Mueller, A.K. Sinha, Impaired prostaglandin E1/I2 receptor activity of human blood platelets in acute ischemic heart disease, Circ. Res. 66 (1990) 932-940.

[14] M. Hirata, Y. Hayashi, F. Ushikubi, Y. Yokota, R. Kageyama, S. Nakanishi, S. Narumiya, Cloning and expression of cDNA for a human thromboxane A2 receptor, Nature 349 (1991) 617-620.

[15] M.K. Raychowdhury, M. Yukawa, L.J. Collins, S.H. McGrail, K.C. Kent, J.A. Ware, Alternative splicing produces a divergent cytoplasmic tail in the human endothelial thromboxane A2 receptor, J. Biol. Chem. 269 (1994) 19256-19261.

[16] M.K. Raychowdhury, M. Yukawa, L.J. Collins, S.H. McGrail, K.C. Kent, J.A. Ware, Alternative splicing produces a divergent cytoplasmic tail in the human endothelial thromboxane A2 receptor, J. Biol. Chem. 270 (1995) 7011.

[17] B.T. Kinsella, Thromboxane A2 signalling in humans: a 'Tail' of two receptors, Biochem. Soc. Trans. 29 (2001) 641-654.

[18] S.M. Miggin, B.T. Kinsella, Expression and tissue distribution of the mRNAs encoding the human thromboxane A2 receptor (TP) alpha and beta isoforms, Biochim. Biophys. Acta 1425 (1998) 543-559.

[19] S.M. Miggin, B.T. Kinsella, Thromboxane A(2) receptor mediated activation of the mitogen activated protein kinase cascades in human uterine smooth muscle cells, Biochim. Biophys. Acta 1539 (2001) 147-162.

[20] A. Habib, G.A. FitzGerald, J. Maclouf, Phosphorylation of the thromboxane receptor alpha, the predominant isoform expressed in human platelets, J. Biol. Chem. 274 (1999) 2645-2651.

[21] A.T. Coyle, S.M. Miggin, B.T. Kinsella, Characterization of the 5' untranslated region of alpha and beta isoforms of the human thromboxane A2 receptor (TP). Differential promoter utilization by the TP isoforms, Eur. J. Biochem. 269 (2002) 4058-4073.

[22] A.T. Coyle, B.T. Kinsella, Characterization of promoter 3 of the human thromboxane A receptor gene. A functional AP-1 and octamer motif are required for basal promoter activity, FEBS J. 272 (2005) 1036-1053.

[23] A.T. Coyle, B.M. O'Keeffe, B.T. Kinsella, 15-Deoxy Delta-prosta-glandin J suppresses transcription by promoter 3 of the human thromboxane A receptor gene through peroxisome proliferator-activated receptor gamma in human erythroleukemia cells, FEBS J. 272 (2005) 4754-4773.

[24] A.T. Coyle, B.T. Kinsella, Synthetic peroxisome proliferator-activated receptor gamma agonists rosiglitazone and troglitazone suppress transcription by promoter 3 of the human thromboxane A(2) receptor gene in human erythroleukemia cells, Biochem. Pharmacol. 71 (2006) 1308-1323.

[25] A. Habib, R. Vezza, C. Creminon, J. Maclouf, G.A. FitzGerald, Rapid, agonist-dependent phosphorylation in vivo of human thromboxane receptor isoforms. Minimal involvement of protein kinase C. J. Biol. Chem. 272 (1997) 7191 -7200.

[26] M. Walsh, J.F. Foley, B.T. Kinsella, Investigation of the role of the carboxyl-terminal tails of the alpha and beta isoforms of the human thromboxane A(2) receptor (TP) in mediating receptor:effector coupling, Biochim. Biophys. Acta 1496 (2000) 164-182.

[27] T. Hirata, F. Ushikubi, A. Kakizuka, M. Okuma, S. Narumiya, Two thromboxane A2 receptor isoforms in human platelets. Opposite coupling to adenylyl cyclase with different sensitivity to Arg60 to Leu mutation, J. Clin. Invest. 97 (1996) 949-956.

[28] R. Vezza, A. Habib, G.A. FitzGerald, Differential signaling by the thromboxane receptor isoforms via the novel GTP-binding protein, Gh, J. Biol. Chem. 274 (1999) 12774-12779.

[29] R.J. Lefkowitz, G protein-coupled receptors: III. New roles for receptor kinases and beta-arrestins in receptor signaling and desensitization, J. Biol. Chem. 273 (1998) 18677-18680.

[30] S.K. Shenoy, R.J. Lefkowitz, Multifaceted roles of beta-arrestins in the regulation of seven-membrane-spanning receptor trafficking and signalling, Biochem. J. 375 (2003) 503-515.

[31] M.T. Walsh, B.T. Kinsella, Regulation of the human prostanoid TPalpha and TPbeta receptor isoforms mediated through activation of the EP(1) and IP receptors, Br. J. Pharmacol. 131 (2000) 601 -609.

[32] L.P. Kelley-Hickie, B.T. Kinsella, EP1- and FP-mediated cross-desensi-tization of the alpha (alpha) and beta (beta) isoforms of the human thromboxane A2 receptor, Br. J. Pharmacol. 142 (2004) 203-221.

[33] M.T. Walsh, J.F. Foley, B.T. Kinsella, The alpha, but not the beta, isoform of the human thromboxane A2 receptor is a target for prostacyclin-mediated desensitization, J. Biol. Chem. 275 (2000) 20412-20423.

[34] J.F. Foley, L.P. Kelley, B.T. Kinsella, Prostaglandin D(2) receptor-mediated desensitization of the alpha isoform of the human thromboxane A(2) receptor, Biochem. Pharmacol. 62 (2001) 229-239.

[35] H.M. Reid, B.T. Kinsella, The alpha, but not the beta, isoform of the human thromboxane A2 receptor is a target for nitric oxide-mediated desensitization. Independent modulation of Tp alpha signaling by nitric oxide and prostacyclin, J. Biol. Chem. 278 (2003) 51190-51202.

[36] K.L. Pierce, R.T. Premont, R.J. Lefkowitz, Seven-transmembrane receptors, Nat. Rev., Mol. Cell Biol. 3 (2002) 639-650.

[37] J.L. Parent, P. Labrecque, M.J. Orsini, J.L. Benovic, Internalization of the TXA2 receptor alpha and beta isoforms. Role of the differentially spliced cooh terminus in agonist-promoted receptor internalization, J. Biol. Chem. 274 (1999) 8941-8948.

[38] J.L. Parent, P. Labrecque, M. Driss Rochdi, J.L. Benovic, Role of the differentially spliced carboxyl terminus in thromboxane A2 receptor trafficking: identification of a distinct motif for tonic internalization, J. Biol. Chem. 276 (2001) 7079-7085.

[39] W.J. Koch, B.E. Hawes, J. Inglese, L.M. Luttrell, R.J. Lefkowitz, Cellular expression of the carboxyl terminus of a G protein-coupled receptor kinase attenuates G beta gamma-mediated signaling, J. Biol. Chem. 269 (1994) 6193-6197.

[40] CM. Gorman, D. Gies, G. McCray, Transient production of proteins using an adenovirus transformed cell line, DNA Protein Eng. Tech. 2 (1990) 3-10.

[41] M.M. Bradford, A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72 (1976) 248-254.

[42] G. Grynkiewicz, M. Poenic, R. Tsien, A new generation of Ca2+ indicators with greatly improved fluorescence properties. J. Biol. Chem. 260 (1985) 3440-3450.

[43] P. Godfrey, Inositol lipids and phosphates, in: G. Milligan (Ed.), Signal Transduction: A Practical Approach, Irl Press, Oxford, 1992, pp. 105-122.

[44] P.A. Marsden, K.T. Schappert, H.S. Chen, M. Flowers, C.L. Sundell, J.N. Wilcox, S. Lamas, T. Michel, Molecular cloning and characterization of human endothelial nitric oxide synthase, FEBS Lett. 307 (1992) 287-293.

[45] S.P. Janssens, A. Simouchi, T. Quertermous, D.B. Bloch, K.D. Bloch, Cloning and expression of a cDNA encoding human endothelium-derived relating factor/nitric oxide synthase, J. Biol. Chem. 267 (1992) 22694.

[46] S.P. Janssens, A. Shimouchi, T. Quertermous, D.B. Bloch, K.D. Bloch, Cloning and expression of a cDNA encoding human endothelium-derived relaxing factor/nitric oxide synthase, J. Biol. Chem. 267 (1992) 14519-14522.

[47] M. Nakane, H.H. Schmidt, J.S. Pollock, U. Forstermann, F. Murad, Cloned human brain nitric oxide synthase is highly expressed in skeletal muscle, FEBS Lett. 316 (1993) 175-180.

[48] D.A. Geller, C.J. Lowenstein, R.A. Shapiro, A.K. Nussler, M. Di Silvio, S.C. Wang, D.K. Nakayama, R.L. Simmons, S.H. Snyder, T.R. Billiar, Molecular cloning and expression of inducible nitric oxide synthase from human hepatocytes, Proc. Natl. Acad. Sci. U. S. A. 90 (1993) 3491-3495.

[49] I.G. Charles, R.M. Palmer, M.S. Hickery, M.T. Bayliss, A.P. Chubb, V.S. Hall, D.W. Moss, S. Moncada, Cloning, characterization, and expression of a cDNA encoding an inducible nitric oxide synthase from the human chondrocyte, Proc. Natl. Acad. Sci. U. S. A. 90 (1993) 11419-11423.

[50] L.P. Kelley-Hickie, B.T. Kinsella, Homologous desensitization of signalling by the beta (beta) isoform of the human thromboxane A2 receptor, Biochim. Biophys. Acta 1761 (2006) 1114-1131.

[51] N. Blom, A. Kreegipuu, S. Brunak, PhosphoBase: a database of phosphorylation sites, Nucleic Acids Res. 26 (1998) 382-386.

[52] J.W. Denninger, M.A. Marletta, Guanylate cyclase and the NO/cGMP signaling pathway, Biochim. Biophys. Acta 1411 (1999) 334-350.

[53] O. Feron, Endothelial nitric oxide synthase expression and its functionality, Curr. Opin. Clin. Nutr. Metab. Care 2 (1999) 291 -296.

[54] S.M. Lohmann, A.B. Vaandrager, A. Smolenski, U. Walter, H.R. De Jonge, Distinct and specific functions of cGMP-dependent protein kinases, Trends Biochem. Sci. 22 (1997) 307-312.

[55] P. Ruth, Cyclic GMP-dependent protein kinases: understanding in vivo functions by gene targeting, Pharmacol. Ther. 82 (1999) 355-372.

[56] E.D. Motley, K. Eguchi, M.M. Patterson, P.D. Palmer, H. Suzuki, S. Eguchi, Mechanism of endothelial nitric oxide synthase phosphorylation and activation by thrombin, Hypertension 49 (2007) 577-583.

[57] R.B. Penn, A.N. Pronin, J.L. Benovic, Regulation of G protein-coupled receptor kinases, Trends Cardiovasc. Med. 10 (2000) 81-89.

[58] R.J. Lefkowitz, S.K. Shenoy, Transduction of receptor signals by beta-arrestins, Science 308 (2005) 512-517.

[59] S.K. Shenoy, R.J. Lefkowitz, Receptor regulation: beta-arrestin moves up a notch, Nat. Cell Biol. 7 (2005) 1059-1061.

[60] S.K. Shenoy, M.T. Drake, C.D. Nelson, D.A. Houtz, K. Xiao, S. Madabushi, E. Reiter, R.T. Premont, O. Lichtarge, R.J. Lefkowitz, beta-arrestin-dependent, G protein-independent ERK1/2 activation by the beta2 adrenergic receptor, J. Biol. Chem. 281 (2006) 1261-1273.

[61] J.A. Oates, G.A. FitzGerald, R.A. Branch, E.K. Jackson, H.R. Knapp, L.J. Roberts II, Clinical implications of prostaglandin and thromboxane A2 formation (1), N. Engl. J. Med. 319 (1988) 689-698.

[62] D.J. Fitzgerald, W. Rocki, R. Murray, G. Mayo, G.A. FitzGerald, Thromboxane A2 synthesis in pregnancy-induced hypertension, Lancet 335 (1990) 751-754.

[63] E. Hamelin, C. Theriault, G. Laroche, J.L. Parent, The intracellular trafficking of the G protein-coupled receptor TPbeta depends on a direct interaction with Rab11, J. Biol. Chem. 280 (2005) 36195-36205.

[64] H. Zhou, F. Yan, H.H. Tai, Phosphorylation and desensitization of the human thromboxane receptor-alpha by G protein-coupled receptor kinases, J. Pharmacol. Exp. Ther. 298 (2001) 1243-1251.

[65] D. Fulton, J.P. Gratton, W.C. Sessa, Post-translational control of endothelial nitric oxide synthase: why isn't calcium/calmodulin enough? J. Pharmacol. Exp. Ther. 299 (2001) 818-824.

[66] R.F. Spurney, Role of C-terminal serines in desensitization and phosphorylation of the mouse thromboxane receptor, J. Biol. Chem. 273 (1998) 28496-28503.

[67] P.L. Huang, Z. Huang, H. Mashimo, K.D. Bloch, M.A. Moskowitz, J.A. Bevan, M.C. Fishman, Hypertension in mice lacking the gene for endothelial nitric oxide synthase, Nature 377 (1995) 239-242.

[68] E.G. Shesely, N. Maeda, H.S. Kim, K.M. Desai, J.H. Krege, V.E. Laubach, P.A. Sherman, W.C. Sessa, O. Smithies, Elevated blood pressures in mice lacking endothelial nitric oxide synthase, Proc. Natl. Acad. Sci. U. S. A. 93 (1996) 13176-13181.

[69] W.G. Haynes, J.P. Noon, B.R. Walker, D.J. Webb, l-NMMA increases blood pressure in man, Lancet 342 (1993) 931-932.

[70] J.F. Arnal, A.T. Dinh-Xuan, M. Pueyo, B. Darblade, J. Rami,

Endothelium-derived nitric oxide and vascular physiology and pathology, Cell. Mol. Life Sci. 55 (1999) 1078-1087.

[71] L.S. Haug, V. Jensen, O. Hvalby, S.I. Walaas, A.C. Ostvold, Phosphoryla-tion of the inositol 1,4,5-trisphosphate receptor by cyclic nucleotide-dependent kinases in vitro and in rat cerebellar slices in situ, J. Biol. Chem. 274 (1999) 7467-7473.

[72] H.H. Schmidt, S.M. Lohmann, U. Walter, The nitric oxide and cGMP signal transduction system: regulation and mechanism of action, Biochim. Biophys. Acta 1178 (1993) 153-175.

[73] J. Schlossmann, A. Ammendola, K. Ashman, X. Zong, A. Huber, G. Neubauer, G.X. Wang, H.D. Allescher, M. Korth, M. Wilm, F. Hofmann, P. Ruth, Regulation of intracellular calcium by a signalling complex of IRAG, IP3 receptor and cGMP kinase Ibeta, Nature 404 (2000) 197-201.

[74] A. Ammendola, A. Geiselhoringer, F. Hofmann, J. Schlossmann, Molecular determinants of the interaction between the inositol 1,4,5-trisphosphate receptor-associated cGMP kinase substrate (IRAG) and cGMP kinase Ibeta, J. Biol. Chem. 276 (2001) 24153-24159.

[75] H.K. Surks, N. Mochizuki, Y. Kasai, S.P. Georgescu, K.M. Tang, M. Ito, T.M. Lincoln, M.E. Mendelsohn, Regulation of myosin phosphatase by a specific interaction with cGMP-dependent protein kinase Ialpha, Science 286 (1999) 1583-1587.

[76] S. Massberg, M. Sausbier, P. Klatt, M. Bauer, A. Pfeifer, W. Siess, R. Fassler, P. Ruth, F. Krombach, F. Hofmann, Increased adhesion and aggregation of platelets lacking cyclic guanosine 3',5'-monophosphate kinase I, J. Exp. Med. 189 (1999) 1255-1264.

[77] K.M. Tang, G.R. Wang, P. Lu, R.H. Karas, M. Aronovitz, S.P. Heximer, K.M. Kaltenbronn, K.J. Blumer, D.P. Siderovski, Y. Zhu, M.E. Mendelsohn, Regulator of G-protein signaling-2 mediates vascular smooth muscle relaxation and blood pressure, Nat. Med. 9 (2003) 1506-1512.

[78] E.N. Christensen, M.E. Mendelsohn, Cyclic GMP-dependent protein kinase Ialpha inhibits thrombin receptor-mediated calcium mobilization in vascular smooth muscle cells, J. Biol. Chem. 281 (2006) 8409-8416.

[79] J.K. Wright, L.T. Kim, T.E. Rogers, H. Nguyen, R.H. Turnage, Nitric oxide and thromboxane A2-mediated pulmonary microvascular dysfunction, Arch. Surg. 134 (1999) 293-298.

[80] A. Christopoulos, E.E. El-Fakahany, The generation of nitric oxide by G protein-coupled receptors, Life Sci. 64 (1999) 1-15.

[81] H. Ju, V.J. Venema, M.B. Marrero, R.C. Venema, Inhibitory interactions of the bradykinin B2 receptor with endothelial nitric-oxide synthase, J. Biol. Chem. 273 (1998) 24025-24029.

[82] Y. Hou, J. Lascola, N.O. Dulin, R.D. Ye, D.D. Browning, Activation of cGMP-dependent protein kinase by protein kinase C, J. Biol. Chem. 278 (2003) 16706-16712.