THE CXCR1 AND CXCR2 RECEPTORS FORM CONSTITUTIVE HOMO AND HETERO-DIMERS SELECTIVELY AND WITH EQUAL APPARENT

AFFINITIES

Shirley Wilson, *Graeme Wilkinson and Graeme Milligan

Molecular Pharmacology Group Division of Biochemistry and Molecular Biology

Institute of Biomedical and Life Sciences University of Glasgow

Glasgow G12 8QQ Scotland, U.K.

*Biological Chemistry

AstraZeneca Mereside, Alderely Park

Cheshire, SK10 4TG England, U.K.

Running title: CXCR1/CXCR2 interactions

Correspondence to: Graeme Milligan, Davidson Building, University of Glasgow,

Glasgow G12 8QQ, Scotland, U.K. Tel. (44) 141 330 5557, FAX (44) 141 330 4620, e-mail, g.milligan@bio.gla.ac.uk

1

JBC Papers in Press. Published on June 9, 2005 as Manuscript M413475200

Copyright 2005 by The American Society for Biochemistry and Molecular Biology, Inc.

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

Both homo- and hetero-dimeric interactions between the CXCR1 and CXCR2 chemokine receptors were observed following co-expression of forms of these receptors in HEK293 cells using assays including co-immunoprecipitation, single cell imaging of fluorescence resonance energy transfer, cell surface time-resolved fluorescence resonance energy transfer and bioluminescence resonance energy transfer. These interactions were constitutive and unaffected by the presence of the agonist interleukin 8 and selective as no significant interactions were noted between either the CXCR1 or CXCR2 receptor and the α1A-adrenoceptor. Saturation bioluminescence resonance energy transfer indicated that hetero-meric interactions between CXCR1 and CXCR2 were of similar affinity as the corresponding homo-meric interactions. A novel endoplasmic reticulum trapping strategy demonstrated that these interactions were initiated during protein synthesis and maturation and prior to cell surface delivery. These studies indicate that CXCR1/CXCR2 hetero-dimers are as likely to form in cells co-expressing these two chemokine receptors as the corresponding homo-dimers and stand in contrast to previous studies indicating an inability of the CXCR1 receptor to homo-dimerize or to interact with the CXCR2 receptor (Trettel et al., (2003) J. Biol. Chem. 278, 40980-40988).

2

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

1Abbreviations: APC, allophycocyanin; -BRET, bioluminescence resonance energy transfer; CFP, cyan fluorescent protein; ER, endoplasmic reticulum; GFP, green fluorescent protein; FRET, fluorescence resonance energy transfer; IL8, interleukin 8/CXCL8; TR-FRET, time-resolved fluorescence resonance energy transfer; YFP, yellow fluorescent protein.

3

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

In recent times the concept that G protein-coupled receptors (1GPCRs) can exist as dimers and/or higher order oligomers has become increasingly accepted (1-5). Techniques ranging from the co-immunoprecipitation of co-expressed but differentially epitope-tagged forms of a single receptor species (6-7) to the use of resonance energy transfer-based methods (8-12) have provided support for the presence of such dimers/oligomers for many GPCRs in transfected cell systems and the application of atomic force microscopy has shown the presence of dimers and arrays of dimers of rhodopsin in murine rod outer segment discs (13-14). In many cases, GPCR quaternary structure seems to be defined early in the processes of receptor synthesis and maturation with the GPCR being transported to the cell surface as a preformed dimer/oligomer (15-17), the structure of which is unaffected by the presence of agonist ligands. The potential quaternary structure of a range of chemokine receptors has also been explored using similar approaches (18-23). Although a substantial number of chemokine receptors have been shown to possess such quaternary structure, a number of features of certain chemokine receptors are either controversial or seem not to follow the general model outlined above. For example, dimerization/oligomerization of a number of chemokine receptors appears to be promoted by the binding of chemokine ligands (18-19). Equally, it appears that mutation of certain chemokine receptors to prevent dimerization does not restrict membrane delivery (24). Amongst the chemokine receptors (25) the closely related CXCR1 and CXCR2 receptors share a common agonist ligand in interleukin 8 (IL8, also called CXCL8). They are widely co-expressed on immune cells, including neutrophils, CD8(+) T cells and mast cells and non-competitive allosteric inhibitors of these receptors have been suggested to offer a general means to inhibit polymorphonuclear cell recruitment in vivo (26). Recently, Trettel et al., (23) have reported that the CXCR2 receptor forms a constitutive dimer when expressed in HEK293 cells and also in cerebellar neurons in which it is expressed endogenously. By contrast, these workers

reported that the CXCR1 receptor was unable to dimerize (23). As well as homo-dimeric/oligomeric interactions, many related GPCRs have been shown to have the capacity to form hetero-dimers. This can result in alterations in receptor pharmacology and signal transduction characteristics (27-29). A number of chemokine receptors have been reported to have the capacity to form hetero-dimers (30) but again, Trettel and co-workers (23) reported lack of interactions between the CXCR1 and CXCR2 receptors following their co-expression. Given the general parsimony of structure and function of closely related proteins we decided to re-examine the capacity of the CXCR1 and CXCR2 receptors to form homo- and hetero-dimers. Using a wide range of approaches we show that both CXCR1 and CXCR2 form homo-dimer/oligomers at an early stage in synthesis and maturation and that when co-expressed CXCR1 and CXCR2 form hetero-dimers as effectively as homo-dimers. The extent of neither the CXCR1-CXCR2 hetero-dimer nor the corresponding homo-dimers is affected by the presence of IL8. Methods Materials [125I]IL8 (2000 Ci/mmol) was from Amersham Biosciences. All reagents for BRET2 studies were from Packard Biosciences. Oligonucleotides were purchased from Interactiva, Ulm, Germany. The anti human CXCR1 antibody was from R&D Systems, Abingdon, U.K.) Molecular Constructs The human CXCR1 receptor was used as a PCR template for all CXCR1 constructs For the N-terminally modified forms of the receptor, primers encoded the appropriate epitope tag sequence and introduced a stop codon after the last amino acid of the receptor sequence. FLAG-CXCR1 Sense:5’AAAAGAATTCGCCACCATGGACTACAAGGACGACGATGATGA TAAGTCAAATATTACAGATCCAC 3’ Anti-sense: 5’AAAAGAATTCTCAGAGGTTGGAAGAGACATTGAC 3’. EcoRI sites are underlined and the

4

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

amplified fragment digested and ligated into pcDNA3. c-myc-CXCR1 Sense:5’AAAAGAATTCGCCACCATGGAACAAAAACTTATTTCTGAAGAA GATCTGTCAAATATTACAGATCCAC 3’ Anti-sense:5’AAAAGAATTCTCAGAGGTTGGAAGAGACATTGAC 3’. EcoRI sites are underlined and the amplified fragment digested and ligated into pcDNA3. HA-CXCR1 Sense:5’AAAAGGTACCGCCACCATGTATCCCTACGACGTCCCCGATTAT GCGTCAAATATTACAGATCCAC 3’ Anti-sense: 5’ AAAAGAATTCTCAGAGGTTGGAAGAGACATTGAC 3’. A KpnI site present in the sense primer and an EcoRI site present in the anti-sense primer are underlined and the amplified fragment digested and ligated into pcDNA3. To generate the endoplasmic reticulum (ER) trapped form of HA-CXCR1 (HA-CXCR1-ER) primers encoding the C-terminal 14 amino acid segment of the α2c-adrenoceptor were annealed to HA-CXCR1. Sense:5’AATTCAAGCATATCCTCTTTCGAAGGAGGAGAAGGGGCTTCAGGCAATGAT 3’, Anti-sense:5’CTAGATGATTGCCTGAAGCCCCTTCTCCTCCTTCGAAAGAGGA TATGCTTG 3’. An EcoRI site present in the sense primer and the XbaI site present in the anti-sense primer are underlined and the fragment digested and ligated downstream of the HA-CXCR1 fragment in frame in pcDNA3. Carboxyl-terminally tagged constructs In each case the primers were designed to amplify the CXCR1 receptor and remove the stop codon; CXCR1-GFP2 Sense:5’AAAAGAATTCGCCACCATGTCAAATATTACAGATCCAC-3’, Anti-sense:5’ AAAAGGTACCGAGGTTGGAAGAGACATTGAC 3’. The EcoRI and KpnI sites present in the sense and anti-sense primers respectively

are shown underlined. The amplified fragment was digested and ligated into pGFP2-N2 (Packard Biosciences) in frame with GFP2. CXCR1-Renilla luciferase Sense:5’AAAA AAGCTTGCCACCATGTCAAATATTACAGATCCAC 3’, Anti-sense: 5’AAAACTCGAGGTTGGAAGAGACATTGAC 3’. The HindIII and XhoI sites present in the sense and anti-sense primers respectively are underlined. The amplified fragment was digested and ligated into pcDNA3 upstream and in frame with Renilla luciferase ligated between XhoI and XbaI. CXCR1-YFP Sense: 5’ AAAA AAGCTTGCCACCATGTCAAATATTACAGATCCAC 3’, Anti-sense: 5’ AAAAGCGGCCGCGAGGTTGGAAGAGACATTGAC 3’. The HindIII and NotI sites encoded in the sense and anti-sense primers are underlined. The amplified fragment was digested and ligated into pcDNA3.1 (+) upstream and in frame with YFP ligated between NotI and XhoI. CXCR1-CFP Sense: 5’ AAAA AAGCTTGCCACCATGTCAAATATTACAGATCCAC3’, Anti-sense: 5’AAAAGGTACCGAGGTTGGAAGAGACATTGAC3’. The HindIII and KpnI sites encoded in the sense and anti-sense primers are underlined. The amplified fragment was digested and ligated into pcDNA3 upstream and in frame with CFP ligated between HindIII and XhoI. CXCR2 Human CXCR2 was used as a PCR template for all CXCR2 constructs. These were generated in a similar fashion to the CXCR1 constructs. For the N-terminally modified forms of the receptor, primers encoded the appropropriate epitope tag sequence and introduced a stop codon after the last amino acid of the receptor sequence.

5

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

FLAG-CXCR2 Sense:5’AAAAGAATTCGCCACCATGGACTACAAGGACGACGATGATAA GGAAGATTTTAACATGGAG 3’, Anti-sense: 5’AAAAGAATTCGAGAGTGGAAGTGTGCCC 3’. EcoRI sites present in both sense and anti-sense primers are underlined and the amplified fragment digested and ligated into pcDNA3. c-myc-CXCR2 Sense:5’AAAAGAATTCGCCACCATGGAACAAAAACTTATTTCTGAAGAAGA TCTGGAAGATTTTAACATGGAG 3’ Anti-sense:5’AAAAGAATTCGAGAGTGGAAGTGTGCCC 3’. EcoRI sites are underlined and the amplified fragment digested and ligated into pcDNA3. VSVG-CXCR2 Sense:5’AAAAGGTACCGCCACCATGTACACCGACATCGAAATGAACCGC CTTGGTAAG - 3’, Anti-sense: 5’AAAAGAATTCGAGAGTGGAAGTGTGCCC3’. The KpnI and EcoRI sites present in the sense and anti-sense primers respectively are underlined and the amplified fragment digested and ligated into pcDNA3. For the C-terminally modified forms of the receptor, primers were designed to amplify the sequence and remove the stop codon. CXCR2-GFP2 Sense:5’ AAAAGAATTCGCCACCATGGAAGATTTTAACATGGAC 3’, Anti-sense:5’ AAAAGGTACCGAGAGTAGTGGAAGTGTGCCC 3’. The EcoRI and KpnI sites present in the sense and anti-sense primers respectively are underlined. The amplified fragment was digested and ligated into pGFP2-N2 in frame with GFP2. CXCR2-Renilla luciferase Sense:5’AAAAAAGCTTGCCACCATGGAAGATTTTAACATGGAG3’, Anti-sense:5’AAAACTCGAGGAGCGTCGTGGAAGTGTG 3’. The HindIII and XhoI sites present in the

sense and anti-sense primers respectively are underlined. The amplified fragment was digested and ligated into pcDNA3 upstream and in frame with Renilla luciferase ligated between XhoI and XbaI. CXCR2-YFP Sense: 5’AAAAAAGCTTGCCACCATGGAAGATTTTAACATGGAG 3’, Anti-sense: 5’AAAAGCGGCGCGAGAGTAGTGGAAGTGTGCCC 3’. The HindIII and NotI sites encoded in the sense and anti-sense primers are underlined. The amplified fragment was digested and ligated into pcDNA3.1 (+) upstream and in frame with YFP ligated between NotI and XhoI. CXCR2-CFP Sense:5’ AAAAAAGCTTGCCACCATGGAAGATTTTAACATGGAG 3’, Anti-sense 5’ AAAAGGTACCGAGAGTAGTGGAAGTGTGCCC 3’. The HindIII and KpnI sites encoded in the sense and anti-sense primers are underlined. The amplified fragment was digested and ligated into pcDNA3 upstream and in frame with CFP ligated between HindIII and XhoI. Cell membrane preparation Pellets of cells were resuspended in 10mM Tris, 0.1mM EDTA, pH 7.4 (TE buffer) and the cells homogenized using 40 strokes of a glass on Teflon homogenizer. Samples were centrifuged at 1000 x g for 10 min at 4oC to remove unbroken cells and nuclei. The supernatant fraction was removed and passed through a 25 gauge needle 10 times before being transferred to ultra-centrifuge tubes and subjected to centrifugation at 50000 x g for 30 min. The supernatant was discarded and the pellet resuspended in TE buffer. Protein concentration was assessed, membranes diluted to 1mg/ml and stored at –80oC until required. Radioligand binding Reaction mixtures were established in a volume of 100µl containing 5µg of membrane protein, 100pM [125I]IL8, and a range of concentrations of non-radiolabelled IL8. Samples were incubated

6

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

for 90 min at room temperature prior to filtration through Whatman GF/C filters. Data was analysed using Graphpad Prism, and IC50 values determined via non-linear regression using one site competition analysis. The equilibrium dissociation constant for the binding of IL8 was calculated using the Cheng-Prusoff equation (31). Co-immunoprecipitation studies Cells were harvested 24 h following transfection and resuspended in RIPA buffer (50 mM HEPES, 150 mM NaCl, 0.5% sodium deoxycholate, 0.1% SDS, 1% Triton X-100, 10 mM NaF, 5 mM EDTA, 0.1 mM NaPO4, 5% ethylene glycol). The cell pellet was disrupted as above and placed on a rotating wheel for 1 h at 4°C. Samples were then centrifuged for 10 min at 14000 x g at 4°C and the supernatant transferred to a fresh tube containing 200µl of 1X RIPA and 50µl of Protein G beads (Sigma) to pre-clear the samples. Following incubation on a rotating wheel for 1 h at 4°C the samples were re-centrifuged at 14000 x g at 4°C for 1 min and protein concentration of the supernatant determined. Samples containing equal protein amounts were incubated overnight with 40µl Protein G beads, 5µg M2 anti-FLAG antibody (Sigma) at 4°C on a rotating wheel and fractions reserved to monitor protein expression in the cell lysates. Samples were centrifuged at 14000 x g for 1 min at 4°C and the Protein G beads washed with 3 times with RIPA buffer. Following addition of Laemmli buffer and heating to 85°C for 4 min, both immunoprecipitated samples and cell lysate controls were revolved by SDS-PAGE using pre-cast 4-12% acrylamide Novex Bis-tris gels (Invitrogen BV). Proteins were transferred onto nitrocellulose. These membranes were incubated in 5% (w/v) low fat milk, 0.1% Tween 20/PBS (v/v) solution at room temperature on a rotating shaker for 2 h and then with primary antibody overnight in 5% (w/v) low fat milk, 0.1% Tween 20/PBS (v/v) solution at 4°C. The membrane was washed 3 x in 0.1% Tween 20/PBS before addition of secondary antibody. Following further washes the membrane was subsequently developed

using ECL solution (Pierce). Confocal laser scanning microscopy Cells were imaged using a laser scanning confocal microscope (Zeiss LSM 5 Pascal) equipped with a 63x oil-immersion Plan Fluor Apochromat objective lens with a numerical aperture of 1.4. A pinhole of 20 and an electronic zoom of 1 or 2.5 was used (Carl Zeiss Inc., Thornwood, NY). The excitation laser line for GFP and YFP was the 488nm argon laser with detection via a 505-530 band pass filter. Alexa594 label was detected using a 543nm helium/neon laser and detected via a 560 long-pass filter. The images were manipulated using MetaMorph imaging software (version 6.1.3; Universal Imaging Corporation, Downing, PA). In some experiments, fixed cells were used. Cells grown on coverslips were transiently transfected and washed 3 times with ice-cold PBS. Cells were fixed for 10 min at room temperature using 4% paraformaldehyde in PBS/5% sucrose. The cells were washed a further 3 times in ice-cold PBS prior to being fixed onto microscope slides with 40% glycerol in PBS. Fluorescent microscopy and FRET imaging in living cells HEK293T cells were grown on poly-D-lysine treated coverslips and transiently transfected with appropriate CFP/YFP fusion proteins. Coverslips were placed into a microscope chamber containing physiological saline solution (130nM NaCl, 5mM KCl, 1mM CaCl2, 1mM MgCl2, 20mM HEPES, 10mM D-glucose pH 7.4). Cells were visualized using a Nikon Eclipse TE2000-E fluorescence inverted microscope and images obtained individually for eYFP, CFP and FRET filter channels using an Optoscan monochromator (Cairn Research, Faversham, Kent, UK) and a dichroic mirror 86002v2bs (Chroma Inc., Rockingham, VT). The filter sets used were; YFP (excitation – 500/5nm; emission – 535/30nm), CFP (excitation 430/12nm; emission – 470/30nm) and FRET (excitation – 430/12nm; emission – 535/30nm). The illumination time was 250ms and binning modes 2x2.

7

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

MetaMorph imaging software was used to quantify the FRET images using the sensitized FRET method. Corrected FRET was calculated using a pixel-by-pixel methodology using the equation FRETc = FRET – (coefficient B x CFP) – (coefficient A x YFP), where CFP, YFP and FRET values correspond to background corrected images obtained through the CFP, YFP and FRET channels. B and A correspond to the values obtained for the CFP (donor) and YFP (acceptor) bleedthrough co-efficients respectively, calculated using cells singly transfected with either the CFP or YFP protein alone. To correct the FRET levels for the varying amounts of donor (CFP) and acceptor (YFP), normalized FRET was calculated using the equation FRETn = FRETc / CFP x YFP, where FRETc, CFP and YFP are equal to the fluorescence values obtained from single cells. Time resolved (Tr) FRET 10cm2 dishes of HEK293T cells were transfected to express N-terminally c-myc or FLAG-tagged forms of CXCR1 or CXCR2 individually or in combination. 48 h after transfection the cells were harvested. Cell pellets were resuspended in 200µl of ice-cold PBS. Anti-c-myc Eu3+ and anti-FLAG allophycocyanin (APC) antibodies were diluted in 50% newborn calf serum: 50% PBS to final concentrations of 5nM and 15nM respectively. Samples were mixed and incubated on a rotating wheel at room temperature for 2 h. Samples were covered in aluminium foil to minimise exposure of the fluorophores to light. Samples were centrifuged at 1000 x g for 1 minute and the antibody mix removed from the cell pellet. The pellet was then washed 2 X in ice-cold PBS and resuspended in 250µl of PBS. To investigate agonist effect on energy transfer, 90µl of cells were transferred to a fresh tube and incubated with the chosen concentration of agonist at 37ºC. To measure the energy transfer, 40µl of each sample was dispensed in triplicate into a black 384-well plate. Blank wells containing PBS were also included. Tr-FRET was determined using a Victor2 plate reader (Packard Bioscience). Excitation was at 340nm and emission filters generated data representing donor (615nm)

and acceptor (665nm) fluorescence. Normalized FRET was calculated using the equation; Normalized FRET = ((A665-BLK)/D615)-C Where A665 is the fluorescent emission from the acceptor, D615 is the fluorescent emission from the donor and BLK represents the background reading at 665nm from wells containing PBS. C represents the cross-talk between the donor and acceptor windows for the samples incubated with only anti-c-myc Eu3+ and is equal to A665-BLK/D615. BRET2 Single Point BRET2 Cells were washed twice in PBS supplemented with 1g/l glucose and resuspended in a final volume of 1ml. 160µl of cells were dispensed into a white-walled 96 well plate (Optiplate, PerkinElmer) and either 20 µl of agonist or PBS/glucose added. If agonist was tested then the plate was incubated for 30 min at 37oC. DeepBlueC (PerkinElmer) substrate was diluted 1:20 in PBS/glucose and the mix kept protected from light until required. 20µl of substrate was added to each well resulting in a final concentration of 10µM and BRET2 measured using a Mithras LB940 (Berthold Technology, Bad Wildbad, Germany). Readings were taken using 410nm (band pass 80nm) corresponding to light emission resulting from Renilla luciferase catalysing the substrate to coelenteramide. Transferred energy emitted by GFP2 was detected using a 515nm (band pass 30nm) filter and a ratiometric reading obtained corresponding to the ratio of light intensity (515nm) to light intensity (410nm). Saturation BRET2 In saturation BRET2 experiments cells were transfected with a constant amount of the energy donor (Renilla luciferase) construct and a varying amount of energy acceptor (GFP2) construct. Cells were harvested, membranes prepared and diluted to 0.5mg/ml. BRET2 was assessed as above for intact cells. Luminescence and fluorescence measurements were also obtained. 50µl of cell membranes were dispensed into white-walled 96-well plates (PerkinElmer) for luminescence

8

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

measurements and black-walled 386-plates (Costar, Cambridge, MA) for fluorescence measurement. For luminescence measurement h-coelenterazine (5µM) was added and the plate incubated at room temperature for 30 min prior to measurement at 410nm using a Mithras LB 940. GFP2 fluorescence was assessed using Victor2 1420 Multilabel counter (PerkinElmer). BRET2 readings were corrected for energy transfer resulting from bleedthrough of signal the Renilla luciferase construct expressed alone but detected in the GFP2 channel. Fluorescence readings were corrected for endogenous fluorescence of HEK293T cell membranes alone. Graphpad Prism 4 was used to analyse data using a one site binding hyperbola equation yielding BRETMAX and BRET50 values. Immunostaining protocol Cells were grown onto coverslips and transiently transfected. 24 h later medium was removed and the cells incubated with 20mM HEPES/DMEM containing the appropriate dilution of primary antibody for 40 min at 37oC in 5% CO2. IL8 (50nM) was added and the coverslip incubated for 30 min. Following 3 washes with PBS cells were fixed by incubating with 4% paraformaldehyde in PBS/5% sucrose for 10 min at room temperature. Following 3 further washes cells were permeabilized with 0.15% Triton-X-100/3% nonfat milk/PBS for 10 min. The coverslips were incubated with a secondary antibody (5mg/ml) conjugated to an Alexa 594

fluorophore. Following incubation for 1 h cells were washed twice in 0.15% Triton-X-100/3% nonfat milk/PBS and three times in PBS. Coverslips were then mounted onto microscope slides with 40% glycerol in PBS. Endoplasmic reticulum trapping and quantitation studies HEK293T cells were transfected to express either HA-CXCR1 or an ER-retained version of this construct that has the C-terminal 14 amino acids of the α2c-adrenoceptor attached to the C-terminal tail (HA-CXCR1-ER). Such cells were co-transfected with N-terminally FLAG-tagged forms of various receptors. 48h following

transfection cells were harvested and counted using a haemocytometer. 5 x 105 cells were dispensed into individual eppendorf tubes and the cells incubated with 15nM APC labelled anti-FLAG antibody and 1µM Hoescht nuclear stain on a rotating wheel for 1h. The cells were centrifuged at 1000 x g for 1 min and the cell pellet washed three times with PBS. The cells were re-suspended in 200µl PBS and 40µl replicates dispensed into a black 384-well plates. Fluorescence corresponding to APC was quantified using a Victor2 1420 Multilabel counter (PerkinElmer). Controls measured fluorescence in similarly treated but non-transfected HEK293T cells and this value subtracted from the other readings. To ensure equal cell number between wells, fluorescence representing Hoechst staining was measured in parallel. Such data ensured well to well cell number variation was less than 20%. cAMP measurements The inhibition of forskolin-stimulated cAMP generation was determined using the HitHunter cAMP XS assay kit (32) (DiscoverX, Birmingham, U.K). Transfected cells were harvested and resuspended in 1X PBS containing 0.5mM isobutylmethylxanthine. Cells were dispensed into white-walled 96 well plates (Optiplate, PerkinElmer) at approximately 30,000 cells/well and incubated with 10µM forskolin and varying concentrations of GRO-α for 30 minutes at 37°C. The assay kit was then used in accordance with the manufacturers’ instructions and luminescence detected using a Mithras LB940 (Berthold Technology, Bad Wildbad, Germany). Results The CXCR1 and CXCR2 receptors share interleukin 8 (IL8) as a high affinity agonist ligand. Because a significant number of both N-and C-terminally modified variants of the human forms of these two receptors were to be utilized in these studies we initially demonstrated that such modifications did not prevent IL8-mediated binding and internalization of the receptors.

9

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

Following introduction of the c-myc epitope tag sequence into the extreme N-terminus of both the CXCR1 and CXCR2 receptors these constructs were expressed transiently in HEK293T cells. Both constructs were shown to be located predominantly at the cell surface (Figures 1a, b). Following treatment of these cells with 50nM IL8 for 30 min a substantial fraction of both modified receptor constructs were re-located into punctate intracellular vesicles (Figures 1a, b). Equivalent experiments using N-terminally FLAG-tagged forms of the receptors produced similar results (data not shown). Both the CXCR1 and CXCR2 receptors were also modified by the addition of a variety of forms of the Aequoria victoria green fluorescent protein (GFP) to the C-terminal tail. As with the N-terminally modified variants, transient expression in HEK293T cells of the CXCR1 and CXCR2 receptors tagged with, for example, the bioluminescence resonance energy transfer (BRET) acceptor competent fluorescent protein GFP2 resulted in a predominantly plasma membrane localization (Figures 1c, d) and treatment with 50nM IL8 for 30 min also caused marked internalization into punctate vesicles (Figures 1c, d). Forms of the CXCR1 and CXCR2 receptors with the BRET energy donor Renilla luciferase linked in-frame to the C-terminus were also effectively delivered to the cell surface following transient expression in HEK293T cells (Figure 1e and data not shown). Modification of the receptors did not alter the binding affinity for IL8. Direct comparisons of the specific binding of [125I]IL8 to the unmodified CXCR1 receptor and its self-competition by non-radiolabelled IL8, with forms of this receptor N-terminally modified to include the c-myc, FLAG or HA epitope tags or with forms of the receptor C-terminally labelled with GFP2, cyan fluorescent protein (CFP), yellow fluorescent protein (YFP) or Renilla luciferase indicated each to bind the ligand with pKi close to -9.5 M (Table 1). Similarly, N-terminal modification of the CXCR2 receptor with each of the FLAG, c-myc or VSV epitope tags, or addition of CFP, YFP, GFP2 or Renilla luciferase to the C-terminal tail did not alter the affinity of this receptor to bind

IL8 (Table 1). Based on the specific binding of 100pM [125I]IL8 and the calculated Ki values for IL8, expression levels of the various forms of CXCR1 and CXCR2 were estimated. These ranged from 50–500 fmol/mg membrane protein and as previously noted by (33), for equal amounts of transfected cDNAs, levels of the CXCR1 constructs were higher than for the equivalent CXCR2 variant. Protein-protein interactions indicative of quaternary structure of GPCRs have been monitored in a range of ways. The most widely used has been the co-immunoprecipitation of differentially tagged but co-expressed polypeptides. Co-expression of both N-terminally c-myc and FLAG-tagged forms of the CXCR1 receptor allowed their co-immunoprecipitation (Figure 2a). Following immunoprecipitation with anti-FLAG antibody, samples were resolved by SDS-PAGE and immunoblotted with anti-c-myc antibody. This resulted in detection of principally a 38kDa polypeptide with small amounts of a band of some 80kDa that suggests a fraction of the immunoprecipitated dimer/oligomer was not effectively separated by the SDS-PAGE conditions employed (Figure 2a). No anti-c-myc reactive bands were present in the immunoprecipitates when either the FLAG or c-myc-tagged forms of CXCR1 were expressed individually, although direct immunoblots of lysates produced from such cells confirmed the expression of the individual forms of the CXCR1 receptor in each the samples anticipated (Figure 2a). Equally, mixing of cell lysates expressing either FLAG-CXCR1 or c-myc-CXCR1 prior to immunoprecipitation with the anti-FLAG antibody did not result in co-immunoprecipitation, indicating that co-expression was required to allow interaction. Equivalent experiments using FLAG and c-myc-tagged forms of CXCR2 produced equivalent data (Figure 2b), with the monomer migrating as a polypeptide of some 35 kDa. Again, a small fraction of the co-immunoprecipitated anti-c-myc immunoreactivity remained as a dimer following migration through SDS-PAGE. Co-expression of FLAG-CXCR1 with c-myc-CXCR2 also allowed their co-immunoprecipitation (Figure 2c), consistent

10

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

with the capacity of these two closely related GPCRs to form a hetero-dimer/oligomer complex in cells in which they are co-expressed. Co-immunoprecipitation experiments require cellular fragmentation and detergent-mediated dissolution of membranes. Therefore, in recent times a series of approaches based on resonance energy transfer techniques have been employed to examine potential protein-protein interactions in living cells. Initially we employed imaging of cells expressing forms of the CXCR1 and CXCR2 receptors tagged at the C-terminus with either CFP or YFP as these are well established fluorescence resonance energy transfer (FRET) partners. Expression of either CXCR1-CFP (Figure 3) or CXCR1-YFP (Figure 3) in HEK293 cells allowed the selective imaging of each construct in individual cells. Co-expression of CXCR1-CFP and CXCR1-YFP resulted in FRET (Figure 3). Co-transfection of the isolated forms of CFP and YFP did not result in significant levels of FRET although imaging studies confirmed their co-expression in individual cells (data not shown, but see (34)). These results define that the positive FRET signals obtained with co-expression of CXCR1-CFP and CXCR1-YFP did not reflect significant mutual affinity between the two fluorescent proteins and rather, therefore, support dimeric/oligomeric protein-protein interactions involving the CXCR1 receptor. This was a selective interaction because in cells co-transfected with CXCR1-CFP and α1A-adrenoceptor-YFP, only low levels of FRET were recorded (Figure 3) although imaging of individual cells demonstrated their co-expression and we ensured levels of expression of the α1A-adrenoceptor-YFP were similar to the levels of expression of CXCR1-YFP in the earlier experiments by direct measures of the levels of YFP fluorescence (Figure 3). Equivalent studies with CFP and YFP-tagged forms of the CXCR2 receptor also resulted in positive FRET signals when the two forms of this receptor were co-expressed (Figure 4). As with the CXCR1 receptor, these effects were selective as only weak interactions could be recorded between the CXCR2

receptor and the α1A-adrenoceptor-YFP (Figure 4). Finally, co-expression of CXCR1-CFP and CXCR2-YFP, or co-expression of the alternate CXCR2-CFP and CXCR1-YFP pairing, also generated strong positive FRET signals indicative of the capacity of these two related receptors to form a hetero-dimer/oligomer complex (Figure 5). As all of these experiments were conducted in the absence of IL8, these studies are also consistent with both homo- and hetero-dimer/oligomers of these receptors forming constitutively without need for agonist. Both the co-immunoprecipitation and FRET imaging studies offer initial insights into the capability of the CXCR1 and CXCR2 to form homo-dimeric/oligomeric and, when co-expressed, hetero-dimeric/oligomeric complexes. However, such studies can offer little insight into the relative propensity of pairs of GPCRs to interact. To assess this we employed saturation bioluminescence resonance energy transfer2 (saturation BRET2) experiments. Proteins tagged with Renilla luciferase and with GFP2 can allow BRET upon addition of an appropriate substrate for the luciferase if interactions between the partner proteins bring the luciferase and the GFP2 into proximity (35). By varying the ratio of energy acceptor (the GFP2-tagged protein) and energy donor (the luciferase-tagged protein) BRET2 saturation curves can be generated in which half-maximal signal provides a measure of the relative affinity of protein-protein interactions (12, 36). Following expression in HEK293 cells of varying ratios of CXCR1-Renilla luciferase and CXCR1-GFP2 addition of the luciferase substrate DeepBlue C to intact cells resulted in BRET2 signals that approached an asymptote with increasing [acceptor]/[donor] ratios and with an estimated BRET2 50% of 3.9 +/- 0.6 (Figure 6A). By contrast, when CXCR1-Renilla luciferase was co-expressed with the isolated GFP2 in varying ratios no measurable BRET2 signal was obtained (Figure 6A), confirming that the BRET2 signals obtained with co-expression of CXCR1-Renilla luciferase and CXCR1-GFP2 reflect protein-protein interactions involving the CXCR1 receptor rather than direct interactions between Renilla

11

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

luciferase and GFP2. Saturation BRET2

experiments in which CXCR2-Renilla luciferase and CXCR2-GFP2 were co-expressed also confirmed the capacity of this receptor to form homo-dimer/oligomers and the measured BRET2 50% of 2.2 +/- 0.1 (Figure 6A) indicated that the affinity of interactions between these forms of the CXCR2 receptor is at least as high as for the CXCR1 receptor. Co-expression of various ratios of CXCR1-Renilla luciferase and CXCR2-GFP2 confirmed the capacity of these two receptors to hetero-dimerize/oligomerize and, because the BRET2 50% ratio in these experiments, 3.6 +/- 0.1 was highly similar to those obtained for the homo-dimer pairings, suggests the propensity of CXCR1 and CXCR2 to hetero-dimerize is similar to that of the individual receptors to form homo-dimers. Again, it was important to obtain clear negative controls. To do so we co-expressed CXCR2-Renilla luciferase along with a GFP2-tagged form of the α1A-adrenoceptor that we have used in previous studies (12). BRET2 signals produced from this pairing were low and, importantly, were fit adequately by a straight line in which signal increase was a direct reflection of the energy acceptor to energy donor ratios (Figure 6A) suggesting that this pair of GPCRs has no substantial mutual affinity. As a ratiometric measure, BRET2signals that reflect specific protein-protein interactions should be independent of absolute expression levels. Increasing signals with receptor expression levels may reflect physical crowding that is not related to true interactions. Such effects have been termed bystander effects (36). Although the expression levels measured by the specific binding of [125I]IL8 (see earlier) were far lower than those previously reported to be required to observe bystander effects (36) we tested this directly. HEK293 cells were transfected with different amounts of CXCR1-Renilla luciferase (energy donor) and CXCR1-GFP2 (energy acceptor) cDNAs but in each case in a 1:1 ratio. Differing expression levels of CXCR1-Renilla luciferase were monitored by direct measurement of luciferase activity using h-coelenterazine as substrate to generate the luminescent signal because the emission spectrum from

oxidation of this substrate is not suited for energy transfer to GFP2 (12). Luminescence increased across the full range of cDNA amounts employed (Figure 6B). Similarly, relative levels of CXCR1-GFP2 were monitored by direct fluorescence and increased linearly with cDNA amount transfected (Figure 6B). Despite this, the BRET2 ratios recorded in such cells upon addition of DeepBlueC as luciferase substrate were not different for the varying levels of the constructs expressed (Figure 6B). As monitored using BRET2 neither homo-dimeric CXCR1 and CXCR2 nor hetero-dimeric CXCR1-CXCR2 interactions were modified by the addition of IL8 (Figure 7), indicating, as in the FRET imaging experiments, that each of these interactions is generated constitutively and not in response to agonist binding. Because the Renilla luciferase and GFP2

tags that act as BRET2 partners are both inside intact cells it is impossible to determine the cellular location of the dimeric/oligomeric GPCRs from such studies. Although the distribution of GFP2 -

tagged forms of both CXCR1 and CXCR2 monitored by confocal microscopy indicated significant plasma membrane delivery (Figure 1) this can only be fully established by 3D-cellular reconstruction from z-plane confocal slices through a single cell. We therefore wished to confirm quaternary structure of CXCR1 and CXCR2 homo- and hetero-dimers at the surface of transfected cells and to assess whether interactions between at least this population of these receptors might be modulated by IL8. To do so we employed time-resolved fluorescence resonance energy transfer (Tr-FRET) following expression of N-terminally FLAG and c-myc epitope-tagged forms of the CXCR1 and CXCR2 receptors. Addition of a combination of Eu3+-labelled anti-c-myc antibody, as long-lived energy donor, and allophycocyanin (APC)-labelled anti-FLAG antibody as energy acceptor to intact HEK293 cells co-expressing N-terminally c-myc- and FLAG-tagged forms of the CXCR1 receptor resulted in a substantial signal corresponding to Tr-FRET (Figure 8). This was not observed with addition of only the Eu3+ -labelled anti-c-myc antibody

12

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

or the APC-labelled anti-FLAG antibody (data not shown). Signals were also negligible when both labelled antibodies were added to mixed cell populations each expressing either c-myc-CXCR1 or FLAG-CXCR1 alone (Figure 9). Equivalent results were obtained with expression of c-myc and FLAG-tagged forms of CXCR2 (Figure 9). Equally strong Tr-FRET signals were also obtained from cells co-expressing pairs of N-terminally tagged CXCR1 and CXCR2 receptors (Figure 9), with results being equivalent whether CXCR1 or CXCR2 acted to bind the energy donating Eu3+-labelled anti-c-myc antibody (Figure 8). As in the BRET2 experiments addition of IL8 (100nM, 15 min) did not modify the Tr-FRET signal (Figure 8).

Recent studies have provided evidence that protein-protein interactions involving GPCRs can occur at an early stage in biogenesis and in advance of delivery to the cell surface (16-17). To examine this for CXCR1/CXCR2 receptors we employed an endoplasmic reticulum trapping strategy. As shown earlier (Figure 1), following transient expression in HEK293 cells a substantial fraction of both N and C-terminally modified forms of the CXCR1 receptor is delivered to the cell surface. Following transient expression of N-terminally HA-tagged CXCR1 in HEK293 cells, antibody staining in non-permeabilized and permeabilized cells confirmed substantial membrane delivery of this construct (Figure 9a). By contrast, although a fraction of a N-terminally HA-tagged form of α2C-adrenoceptor was able to reach the cell surface and hence be detected without cellular permeabilization (Figure 9a) the vast majority of this construct was retained intracellularly, with a distribution pattern consistent with labelling of the endoplasmic reticulum (Figure 10a). Indeed, previous work has shown this to reflect the presence of an endoplasmic reticulum retention motif within the C-terminal 14 amino acids of α2C-adrenoceptor (37-38). We generated a form of the CXCR1 receptor with these 14 amino acids from the α2C-adrenoceptor added to the C-terminal tail. In contrast to the wild type CXCR1 receptor, little of the modified CXCR1 receptor was trafficked to

the cell surface (Figure 9a). This was not a reflection of poor expression of this construct. Following permeabilization of the cells, the modified CXCR1 receptor (HA-CXCR1-ER) was shown to be largely intracellular (Figure 9a). The cell surface expression of FLAG-CXCR1 was monitored when this construct was co-expressed with either HA-tagged wild type CXCR1 or HA-CXCR1-ER. Only some 50% as much of the FLAG-tagged wild type CXCR1 reached the cell surface when its expression partner was HA-CXCR1-ER (Figure 9b). The same was true when the ability of FLAG-CXCR2 to reach the cell surface was monitored. Co-expression with HA-CXCR1-ER resulted in a substantial reduction of cell surface FLAG-CXCR2 compared to when this receptor was co-expressed with HA-CXCR1 (Figure 9b). As in the BRET2 experiments we wished to establish the selectivity of the endoplasmic reticulum trapping strategy. When FLAG--α1A-adrenoceptor was employed, cell surface delivery of this construct was the same whether co-expressed with HA-CXCR1 or HA-CXCR1-ER (Figure 9b), confirming a lack of significant interactions between the CXCR1 receptor and the α1A-adrenoceptor.

To assess if interactions between the CXCR1 and CXCR2 receptors altered potency or functionality of ligands we examined the ability of the CXCR2 selective agonist GRO-α to inhibit cAMP generation stimulated by forskolin (10µM) in cells individually expressing CXCR1 or CXCR2, cells individually expressing the two receptors but mixed prior to the assay or cells co-expressing CXCR1 and CXCR2. As anticipated, GRO-α -mediated inhibition of cAMP production was 30 fold more potent in cells expressing CXCR2 than CXCR1 (Figure 10). In cells co-expressing CXCR1 and CXCR2 the potency of GRO-α was intermediate between the values for the individually expressed receptors and was not different from that noted for cells individually expressing CXCR1 or CXCR2 that were mixed prior to assay (Figure 10).

13

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

Discussion It has become widely accepted that many members of the rhodopsin-like, family A class of GPCRs have the capacity to homo-dimerize and as both direct experimental data (39) and models based on the dimensions of GPCRs and G proteins (14) are consistent with a GPCR dimer binding a single G protein hetero-trimer, dimerization may be integral to GPCR function (6). There is increasing evidence that dimers of many GPCRs form during synthesis and maturation and that this can be important for cell surface delivery (16-17). Equally, there are many examples in which the extent of GPCR dimerization is unaffected by the binding of agonist ligands and indeed, where the GPCR is internalized into the cell as a dimer/oligomer in response to agonist challenge (40-41). The chemokine receptors are members of the class of family A GPCRs (25). However, although the literature is complex (see (5) for review), a number of reports on specific chemokine receptors are inconsistent with the simple pattern outlined above. For example, early studies indicated the chemokine SDF-1α to produce dimerization of the CXCR4 receptor that was almost undetectable in the absence of the ligand (19). In complete contrast, combinations of BRET and sedimentation studies indicated the CXCR4 receptor to be a constitutive dimer that was unaffected by the presence of SDF-1α (22). Ligand-induced dimerization has also been reported for the CCR2 receptor on addition of monocyte chemoattractant protein-1 (18) and the capacity of HIV-1 to utilize the CCR5 receptor as a ‘co-receptor’ for cell entry is blocked by dimerization of the receptor produced by chemokine agonists (42). However, the CCR5 receptor has also been reported to be both a ligand independent, constitutive dimer/oligomer (21) and a monomer (22). At this time a clear pattern is therefore difficult to discern. In many other aspects of structure, function and regulation family A GPCRs show considerable parsimony, as might be expected for a family of homologous proteins (43) and thus we wished to use as wide a range of approaches as possible to re-examine aspects of the interactions between CXCR1 and CXCR2 receptors. In

contrast to previously published work (23), herein we demonstrate the ability of the CXCR1 receptor, as well as the CXCR2 receptor, to form a homo-dimer/oligomer and that when co-expressed the CXCR1 and CXCR2 receptors are able to hetero-dimerize/oligomerize. Such conclusions are based on data from five distinct techniques, including, co-immunoprecipitation, resonance energy transfer approaches and intracellular trapping by an ER-retained version of theCXCR1 receptor. Three of these approaches were particularly enlightening. Firstly, the ER-trapping strategy showed that an ER-retained form of the CXCR1 receptor limited cell surface delivery of N-terminally FLAG-tagged forms of both CXCR1 and CXCR2. Specificity of this assay was established by unaltered cell surface delivery of a N-terminally FLAG-tagged form of the α1A-adrenoceptor, a receptor shown by other means to interact with the CXCR1 and CXCR2 receptors with minimal affinity. This assay established that CXCR1 homo-dimerization and CXCR1-CXCR2 hetero-dimerization occurs, as previously established for the β2-adrenoceptor (17), during receptor synthesis and maturation. Secondly, the application of cell surface Tr-FRET showed that constitutively established forms of each of CXCR1 and CXCR2 homo-dimers/oligomers and CXCR1-CXCR2 hetero-dimers were present at the surface of cells in the absence of IL8 and that such interactions were not modified substantially by the presence of the agonist. Thirdly, use of saturation BRET2 techniques (36) showed that the propensity of CXCR1 and CXCR2 to generate hetero-interactions was not different from their ability to homo-dimerize. Thus, unless specific cellular mechanisms exist in particular cell types to ensure that mRNAs encoding these GPCRs are trafficking to different sections of the ER machinery then it must be expected that CXCR1-CXCR2 hetero-dimers as well as the corresponding homo-dimers will exist and in ratios determined by expression levels of the individual receptors. Although experimental evidence indicates the hypothesis that the propensity of pairs of distinct GPCRs to form hetero-dimers will simply reflect the sequence homology

14

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

between individual GPCRs (44) is too simplistic (see 5 for review) the high similarity between the CXCR1 and CXCR2 would certainly be consistent with the equivalent apparent interaction affinities for homo- and hetero-dimerization that we observed in the BRET saturation studies. Similarly, for the three opioid receptor subtypes, very recent BRET saturation studies have indicated similar interaction affinities between all possible hetero-dimeric pairs (45) although earlier, technically more limited studies, had indicated that certain pairs were unable to hetero-dimerize (46). In each of these regards then the CXCR1 and CXCR2 receptors behave as ‘typical’ family A GPCRs. As noted above, the issue of ligand dependence of GPCR dimerization has been particularly contentious for the chemokine receptors. A substantial number of RET-based experiments have shown small effects of agonist ligands above substantial signals corresponding to constitutive dimers/oligomers (see (5) for review). Despite this, an emerging consensus favours family A GPCRs with small endogenous ligands that bind within the seven transmembrane domains predominantly as constitutive dimers. Observed effects of ligands are then most likely due to conformational alterations within the complex resulting in small changes in the orientation or distance between the RET reporters (47). However, for family A GPCRs with larger peptide and small protein ligands that have contact points for binding outwith the topology of the seven transmembrane helix bundle the situation is more complex. Each of neuropeptide Y Y4 receptor (48), TSH receptor (49) and type A cholecystokinin receptor (50) dimers have been reported to be constitutively formed but dissociated by agonists whereas, by contrast, the GnRH (51), lutropin (52) and a number of other receptors, have been reported to increase in

aggregation state in response to agonists. As such, hetero-dimerization between GPCR pairs might alter ligand pharmacology and function, particularly for receptors such as the chemokine receptors where ligand binding is defined at least in part by elements in the extracellular N-terminal region of the receptor. Indeed, recent studies (53) have provided evidence for negative binding co-operativity within CCR5-CCR2b hetero-dimers. As such, in the current studies, we examined if the effectiveness or potency of the CXCR2 selective agonist ligand GRO-α to inhibit cAMP production was affected by co-expression of CXCR1 and CXCR2. It was not, thus at least for this ligand, which has relatively modest selectivity between the two GPCRs there was no indication of the hetero-dimer displaying a distinct pharmacology or function. Further work, perhaps with less closely related GPCRs, will be required to explore such possibilities more effectively. For example, the ability of the CCR5 chemokine receptors to hetero-dimerize with other co-expressed GPCRs, including opioid receptors has recently been reported (54). At the moment there is no information on the relative affinity of such interactions and therefore their likely importance for physiology and function. However, as other examples of GPCR hetero-dimerization, including those involving interactions between the angiotensin AT1 receptor and each of the AT2 receptor (55), the bradykinin B2 receptor (56) and the mas proto-oncogene (57) have been reported to have physiological and patho-physiological consequences, application of the type of techniques used in these studies would provide a useful starting point to understand the importance of interactions between chemokine receptors with other members of the GPCR family.

Acknowledgements S.W. thanks the Biotechnology and Biosciences Research Council (BBSRC) for a CASE studentship. Part of these studies were supported by the Wellcome Trust.

15

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

References 1. Angers, S., Salahpour, A., and Bouvier, M. (2002) Annu. Rev. Pharmacol. Toxicol. 42, 409-435 2. Milligan, G., Ramsay, D., Pascal, G., and Carrillo, J.J. (2003) GPCR dimerisation. Life Sci. 74, 181-188 3. Breitwieser, G.E. (2004) Circ. Res. 94, 17-27 4. Javitch, J.A. (2004) Mol. Pharmacol. 66, 1077-1082 5. Milligan, G. (2004) Mol. Pharmacol. 66, 1-7 6. Hebert, T.E., Moffett, S., Morello, J.P., Loisel, T.P., Bichet, D.G., Barret, C., and Bouvier, M. (1996) J. Biol. Chem.271, 16384-16392 7. Cvejic, S. and Devi, L.A. (1997) J. Biol. Chem. 272, 26959-26964 8. Angers, S., Salahpour, A., Joly, E., Hilairet, S., Chelsky, D., Dennis, M., and Bouvier, M. (2000) Proc. Natl. Acad. Sci. U S A 97, 3684-3689 9. Overton, M.C., and Blumer, K.J. (2000) Curr. Biol. 10, 341-344 10. Rocheville, M., Lange, D.C., Kumar, U., Patel, S.C., Patel, R.C. and Patel, Y.C. (2000) Science 288, 154-157 11. McVey, M., Ramsay, D., Kellett, E., Rees, S., Wilson, S., Pope, A.J., and Milligan, G. (2001) J. Biol. Chem. 276, 14092-14099 12. Ramsay, D., Carr, I.C., Pediani, J., Lopez-Gimenez, J.F., Thurlow, R., Fidock, M., and Milligan, G. (2004) Mol. Pharmacol. 66, 228-239 13. Liang, Y., Fotiadis, D., Filipek, S., Saperstein, D.A., Palczewski, K., and Engel, A. (2003) J. Biol. Chem. 278, 21655-21662 14. Fotiadis, D., Liang, Y., Filipek, S., Saperstein, D.A., Engel, A., and Palczewski, K. (2004) FEBS Lett. 564, 281-288 15. White, J.H., Wise, A., Main, M.J., Green, A., Fraser, N.J., Disney, G.H., Barnes, A.A., Emson, P., Foord, S.M., and Marshall, F.H. (1998) Nature 396, 679-682 16. Terrillon, S., Durroux, T., Mouillac, B., Breit, A., Ayoub, M.A., Taulan, M., Jockers, R., Barberis, C., and Bouvier, M. (2003) Mol. Endocrinol. 17, 677-691 17. Salahpour, A., Angers, S., Mercier, J.F., Lagace, M., Marullo, S., and Bouvier, M. (2004) J. Biol. Chem. 279, 33390-33397 18. Rodriguez-Frade, J.M., Vila-Coro, A.J., de Ana, A.M., Albar, J.P., and Martinez-A, C., and Mellado, M. (1999) Proc. Natl. Acad. Sci. U S A 96, 3628-3633 19. Vila-Coro, A.J., Rodriguez-Frade, J.M., Martin De Ana, A., Moreno-Ortiz, M.C., Martinez-A, C., and Mellado, M. (1999) FASEB J. 13, 1699-1710 20. Mellado, M., Rodriguez-Frade, J.M., Vila-Coro, A.J., Fernandez, S., Martin de Ana, A., Jones, D.R., Toran, J.L. and Martinez-A., C. (2001) EMBO J. 20, 2497-2507 21. Issafras, H., Angers, S., Bulenger, S., Blanpain, C., Parmentier, M., Labbe-Jullie, C., Bouvier, M., and Marullo, S. (2002) J. Biol. Chem. 277, 34666-34673 22. Babcock, G.J., Farzan, M., and Sodroski, J. (2003) J. Biol. Chem. 278, 3378-3385 23. Trettel, F., Di Bartolomeo, S., Lauro, C., Catalano, M., Ciotti, M.T., and Limatola, C. (2003) J. Biol. Chem. 278, 40980-40988 24. Hernanz-Falcon, P., Rodriguez-Frade, J.M., Serrano, A., Juan, D., del Sol, A., Soriano, S.F., Roncal, F., Gomez, L., Valencia, A., Martinez-A, C., and Mellado, M. (2004) Nature Immunol. 5, 216-223 25. Horuk, R. (2001) Cytokine Growth Factor Rev. 12, 313-335 26. Bertini, R., Allegretti, M., Bizzarri, C., Moriconi, A., Locati, M., Zampella, G., Cervellera, M.N., Di Cioccio, V., Cesta, M.C., Galliera, E., Martinez, F.O., Di Bitondo, R., Troiani, G., Sabbatini, V., D'Anniballe, G., Anacardio, R., Cutrin, J.C., Cavalieri, B., Mainiero, F., Strippoli, R., Villa, P., Di Girolamo, M., Martin, F., Gentile, M., Santoni, A., Corda, D., Poli, G., Mantovani, A., Ghezzi, P., and Colotta, F. (2004) Proc. Natl. Acad. Sci. U S A. 101, 11791-11796 27. Levac, B.A., O'Dowd, B.F., and George, S.R. (2002) Curr. Opin. Pharmacol. 2, 76-81 28. Breit, A., Lagace, M., and Bouvier, M. (2004) J. Biol. Chem. 279, 28756-28765 29. Lee, S.P., So, C.H., Rashid, A.J., Varghese, G., Cheng, R., Lanca, A.J., O'Dowd, B.F., and George, S.R. (2004) J. Biol. Chem. 279, 35671-35678

16

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

30. Rodriguez-Frade, J.M., Mellado, M., and Martinez-A, C. (2001) Trends Immunol. 22, 612-617 31. Cheng, Y., and Prusoff, W.H. (1973) Biochem. Pharmacol. 22, 3099-3108 32. Eglen, R.M. (2002) Assay Drug Dev Technol. 1, 97-104 33. Hall, D.A., Beresford, I.J.M., Browning, C. and Giles, H. (1999) Br. J. Pharmacol. 126, 810-818 34. Carrillo, J.J., López-Gimenez, J.F., and Milligan, G. (2004) Mol. Pharmacol. 66, 1123-1137 35. Milligan, G. (2004) Eur. J.Pharm. Sci. 21, 397-405 36. Mercier, J.F., Salahpour, A., Angers, S., Breit, A. and Bouvier, M. (2002) J. Biol. Chem. 277, 44925-44931 37. Zerangue, N., Schwappach, B., Jan, Y.N., and Jan, L.Y. (1999) Neuron 22, 537-548 38. Margeta-Mitrovic, M. (2002) Methods 27, 311-317 39. Baneres, J.L. and Parello, J. (2003) J. Mol. Biol. 329, 815-829 40. Yesilaltay, A., and Jenness, D.D. (2000) Mol. Biol. Cell 11, 2873-2884. 41. Cao, T., Brelot, A., and von Zastrow, M. (2004) Mol. Pharmacol. 67, 288-297 42. Vila-Coro, A.J., Mellado, M., Martin de Ana, A., Lucas, P., del Real, G., Martinez-A, C., and Rodriguez-Frade, J.M. (2000) Proc. Natl. Acad. Sci. U S A 97,3388-3393 43. Ballesteros, J.A., Shi, L., and Javitch, J.A. (2001) Mol. Pharmacol. 60, 1-19 44. Ramsay, D., Kellett, E., McVey, M., Rees, S. and Milligan, G. (2002) Biochem. J. 365, 429-440 45. Wang, D., Sun, X., Bohn, L.M. and Sadee, W. (2005) Mol Pharmacol. 67, 2173-2184 46. Jordan, B.A. and Devi, L.A. (1999) Nature 399, 697-700 47. Ayoub, M.A., Couturier, C., Lucas-Meunier, E., Angers, S., Fossier, P., Bouvier, M., and Jockers, R. (2002) J. Biol. Chem. 277, 21522-21528 48. Berglund, M.M., Schober, D.A., Esterman, M.A., and Gehlert, D.R. (2003) J. Pharmacol. Exp. Ther. 307, 1120-1126 49. Latif, R., Graves, P. and Davies, T.F. (2001) J. Biol. Chem. 277, 45059-45067 50. Cheng, Z.J. and Miller, L.J. (2001) J. Biol. Chem. 276, 48040-48047 51. Cornea, A., Janovick, J.A., Maya-Nunez, G., and Conn, P.M. (2001) J. Biol. Chem. 276, 2153-2158 52. Tao, Y.X., Johnson, N.B. and Segaloff, D.L. (2004) J. Biol. Chem. 279, 5904-5914 53. El-Asmar, L., Springael, J.Y., Ballet, S., Andrieu, E.U., Vassart, G. and, Parmentier, M. (2005) Mol. Pharmacol. 67, 460-469 54. Suzuki, S., Chuang, L.F., Yau, P., Doi, R.H., and Chuang, R.Y. (2002) Exp. Cell Res. 280, 192-200 55. AbdAlla, S., Lother, H., Abdel-tawab, A.M., and Quitterer, U. (2001) J. Biol. Chem. 276, 39721-39726 56. AbdAlla, S., Lother, H., el Massiery, A., and Quitterer, U. (2001) Nature Med. 7,1003-1009 57. Kostenis, E., Milligan, G., Christopoulos, A., Sanchez-Ferrer, C.F., Heringer-Walther, S., Sexton, P., Gembardt, F., Kellett, E., Martini, L., Vanderheyden, P., Schultheiss, H.P., and Walther, T. (2005) Circulation 111, 1806-1813

17

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

Tables Table 1 N-and C-terminal modifications of the CXCR1 and CXCR2 receptors have

little effect on affinity for IL8 Construct IL8 (pKi) (M)

CXCR1 -9.5 ± 0.3

CXCR1-GFP2 -9.4 ± 0.2

CXCR1-RLuc -9.1 ± 0.1

FLAG-CXCR1 -9.6 ± 0.1

c-myc-CXCR1 -9.7 ± 0.1

CXCR1-YFP -9.4 ± 0.1

CXCR1-CFP -9.5 ± 0.2

HA-CXCR1 -9.2 ± 0.1

CXCR2 -9.5 ± 0.1

CXCR2-GFP2 -9.5 ± 0.05

CXCR2-RLuc -9.5 ± 0.1

FLAG-CXCR2 -9.9 ± 0.3

c-myc-CXCR2 -9.5 ± 0.5

CXCR2-YFP -9.7 ± 0.01

CXCR2-CFP -9.4 ± 0.05

VSV-CXCR2 -9.6 ± 0.05

18

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

The affinity of IL8 to bind to wild type and each of the modified forms of the CXCR1 and CXCR2 receptors was assessed from analysis of the capacity of IL8 to compete with a single concentration (100pM) of [125I]IL8 for binding. All experiments were performed on a minimum of 3 occasions using separate membrane preparations. No significant differences were noted for each of the modified constructs compared to the wild type (1-way ANOVA with Dunnetts’ post test).

19

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

Figure legends Figure 1 IL8 induces internalization of N-terminally and C-terminally modified forms of the CXCR1 and CXCR2 receptors N-terminally c-myc (a, b), or C-terminally GFP2 (c, d) tagged forms of the CXCR1 (a, c) or CXCR2 (b, d) receptors were expressed in HEK293T cells. Cells were challenged with vehicle (left hand panels) or IL8 (50nM, 30 min, 37°C) (right hand panels) fixed and stained with anti-c-myc (a, b) or directly visualized (c, d). In (e) CXCR1 C-terminally tagged with Renilla luciferase was expressed transiently in HEK293 cells. Detection employed an anti-human CXCR1 antibody (R&D Systems, Abingdon, UK) directed to an extracellular epitope. Figure 2 Constitutive homo- and hetero-dimeric/oligomeric interactions between co-expressed forms CXCR1 and CXCR2 receptors revealed by co-immunoprecipitation A. HEK293T cells were mock transfected (1) or transfected to transiently express FLAG-CXCR1 (2, 4) and/or c-myc-CXCR1 (3,4). A mix control (5) was included representing cells individually expressing the constructs and mixed prior to immunoprecipitation. Upper panel: Cell lysates were immunoprecipitated with anti-FLAG antibody, samples resolved by SDS-PAGE and then immunoblotted with anti-c-myc antibody. Lower panels: Western blot analysis of cell lysates using anti-FLAG and anti-c-myc antibodies was also performed to confirm the anticipated pattern of protein expression. B. Experiments akin to A were performed except that CXCR1 was replaced with CXCR2. C. HEK293T cells were mock transfected (1) or transfected to transiently express FLAG-CXCR1 (2, 4) and/or c-myc-CXCR2 (3, 4). A mix control (5) was included representing cells individually expressing the constructs and mixed prior to immunoprecipitation. Upper panel: Cell lysates were immunoprecipitated with anti-FLAG antibody, samples resolved by SDS-PAGE and then immunoblotted with anti-c-myc antibody. Lower panels: Western blot analysis of cell lysates using anti-FLAG and anti-c-myc antibodies was also performed to confirm the anticipated pattern of protein expression. Figure 3 FRET imaging of constitutive CXCR1 homo-meric interactions in single cells. CXCR1-CFP and CXCR1-YFP were co-expressed (a) or expressed individually (b, CXCR1-CFP, c, CXCR1-YFP). α1a-adrenoceptor-YFP was also expressed individually (d) or co-expressed with CXCR1-CFP (e) in HEK293T cells. Individual cells and cell groups were imaged. Left hand panels, CFP; centre panels, YFP; right hand panels, corrected FRET. The FRET signals from such images were then quantitated as in Methods (f). Data are means +/-S.E.M. from three experiments. Figure 4 FRET imaging of selective and constitutive CXCR2 homo-meric interactions in single cells CXCR2-CFP and CXCR2-YFP were co-expressed (a) or expressed individually (b, CXCR2-CFP, c, CXCR2-YFP). α1a-adrenoceptor-YFP was also expressed individually (d) or co-expressed with CXCR2-CFP (e) in HEK293T cells. Individual cells and cell groups were imaged. Left hand panels, CFP; centre panels, YFP; right hand panels, corrected FRET. The FRET signals from such images were then quantitated as in Methods (f). Data are means +/-S.E.M. from three experiments. Figure 5 FRET imaging of constitutive CXCR1-CXCR2 hetero-meric interactions in single cells

20

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

CXCR1-CFP (a, b) was expressed individually (b) or co-expressed with and CXCR2-YFP (a). CXCR2-CFP was also co-expressed with CXCR1-YFP (c) or expressed individually (d) in HEK293T cells. Individual cells and cell groups were imaged. Left hand panels, CFP; centre panels, YFP; right hand panels, corrected FRET. The FRET signals from such images were then quantitated as in Methods (e). Data are means +/-S.E.M. from three experiments. Figure 6 BRET2 analysis of the selectivity and specificity of CXCR1 and CXCR2 interactions A. CXCR1 (filled circles) and CXCR2 (open circles) homo-dimeric interaction were assessed following transient expression of various ratios of receptor-GFP2 (acceptor) and receptor-Renilla luciferase (donor) in HEK293T cells and addition of Deep Blue C as a luciferase substrate. CXCR1-CXCR2 hetero-meric interactions (filled squares) were assessed following expression of various ratios of CXCR1-Renilla luciferase and CXCR2-GFP2. The ratio on the x-axis reflects direct measures of fluorescence (GFP2) and luminescence (luciferase) (see reference 12 for details). The specificity of CXCR2 receptor interactions (filled diamonds) was assessed by co-expression of CXCR2-Renilla luciferase and α1A-adrenoceptor-GFP2. Lack of inherent interactions between Renilla luciferase and GFP2 (open squares) was obtained by co-expression of CXCR1-Renilla luciferase and GFP2. Data from a single experiment are displayed and the error bars represent mean ± S.E.M. of measurements of BRET2 in three separate wells. Two further experiments produced similar results. B. Intact HEK293T cells transiently transfected with increasing amounts of CXCR1-Renilla luciferase (donor) and CXCR1-GFP2 (acceptor) cDNAs in a 1:1 ratio. The amount of each cDNA used is noted. B(i). Donor and acceptor receptor conjugate relative expression levels were monitored by measuring luminescence (triangles) and fluorescence (squares) (see Results). Data from triplicate assays in a single experiment are displayed. Two further experiments produced similar results. B(ii) BRET2 experiments were performed on these samples. Figure 7 IL8 does not modulate CXCR1 or CXCR2 homo- or hetero-meric interactions BRET2 experiments were performed in the absence (open bars) or presence (filled bars) of 50nM IL8 following co-expression of CXCR1-Renilla luciferase and CXCR1-GFP2, CXCR2-Renilla luciferase and CXCR2-GFP2 or CXCR2-Renilla luciferase and CXCR1-GFP2. Data from triplicate assays in a single experiment are displayed. Two further experiments produced similar results. Figure 8 Tr-FRET detects constitutive CXCR1 and CXCR2 homo-meric complexes at the surface of HEK293 cells Combinations of N-terminally FLAG and c-myc forms of CXCR1 and CXCR2 were transfected individually and the cells mixed (Mix) or co-expressed (Co) in HEK293 cells. Cells were treated with vehicle (open bars) or 100nM IL8 (filled bars). Tr-FRET was then measured in these cells, as in Methods, to monitor homo- and hetero-meric interactions at the cell surface. Data represent means +/- S.E.M from three independent experiments. Figure 9 An endoplasmic reticulum trapping strategy to identify CXCR1 receptor interacting GPCRs A. CXCR1 (a), HA-α2C-adrenoceptor (b) and HA-CXCR1-ER retained (c) constructs were expressed transiently in HEK293T cells grown on coverslips. Immunostaining with anti-CXCR1 antibody was performed in non-permeabilized (left hand panels) and permeabilized (right hand panels) cells.

21

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

B. HEK293T cells were transiently co-transfected with either HA-CXCR1 (open bars) or the ER-retained version of this construct (HA-CXCR1-ER) (filled bars) along with N-terminally FLAG-tagged forms of CXCR1, CXCR2 or the α1A-adrenoceptor. Equal numbers of cells were incubated with anti-FLAG antibody conjugated to APC for 2 h prior to washing. Fluorescence corresponding to the APC-labelled antibody was then quantified. Background fluorescence of HEK293T cells was measured and removed from the values obtained. *** , p< 0.001. Data are means +/- S.E.M. from three experiments. Figure 10 The potency of GRO-α to inhibit cAMP production is unaffected by co-expression of CXCR1 and CXCR2 The potency of GRO-α to inhibit cAMP generation stimulated by forskolin (10µM) was measured in cells individually expressing CXCR1 and CXCR2 (closed squares and closed circles respectively), cells individually expressing the two receptors and mixed prior to the assay (open circles) and cells co-expressing CXCR1 and CXCR2 (open diamonds). pIC50 values represent the potency of GRO-α to inhibit cAMP production. Results represent means ± S.E.M. of five individual experiments. No significant difference in potency was found between mixed CXCR1 and CXCR2 expressing cells and those co-expressing CXCR1 and CXCR2 (Students T-test).

22

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

vehicle +IL8 (50nM) Figure 1

a

b

c

d

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

e

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

Figure 2A

Mr (x10-3)

30

35

35

35

αmyc

αmyc

αFlag

50

75 105

250 160

IP

αFlag

cell lysates

1 2 3 4 5

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

Figure 2b

Mr,(x10-3)

30

35

35

35

αmyc

αmyc

αFlag

50

75 105

250 160

IP αFlag

cell lysates

1 2 3 4 5

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

Figure 2C

Mr (x10-3)

250 160

105

75

50

αFlag

αmyc

αmyc

35

35

35

1 2 3 4 5

30

IP αFlag

cell

lysates

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

CFP YFP FRET Figure 3

a b c d e

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

f

CXCR1CFP

CXCR1CFP +

CXCR1YFP

YFP

1Aα

CXCR1CFP +

0.00

0.25

0.50

0.75

1.00

FRET

NO

RM

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

CFP YFP FRET

Figure 4

a b c d e

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

f

CXCR2CFP

CXCR2CFP +

CXCR2YFP

YFP

1Aα

CXCR2C

FP +

0.00

0.25

0.50

0.75

1.00

FRET

NO

RM

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

Figure 5 CFP YFP FRET a b c d

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

e

CXCR1CFP

CXCR1CFP +

CXCR2YFP

CXCR2CFP

CXCR2CFP +

CXCR1YFP

0.00

0.25

0.50

0.75

1.00

FRET

NO

RM

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

Figure 6A

2.5 5.0 7.5 10.0 12.5

-0.1

0.0

0.1

0.2

0.3

0.4

CXCR1RLuc + CXCR2GFP2

CXCR1RLuc + GFP2

CXCR1RLuc + CXCR1GFP2

CXCR2RLuc + CXCR2GFP2

CXCR2RLuc + α1A-GFP2

[Fluorescence/Luminescence]

BR

ET2 R

atio

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

Figure 6B

0.00 0.25 0.50 0.75 1.00 1.250

50000

100000

150000

200000

250000

300000

350000

400000

450000

Each cDNA transfected (µg)

Lum

ines

cenc

e or

Flu

ores

cenc

e (a

rbita

ry u

nits

) (i)

0.2 0.4 0.6 0.8 10.000.050.100.150.200.250.300.350.400.45

Each cDNA transfected (µg)

BRET

2 Rat

io

(ii)

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

Figure 7

2

CXCR1RLu

c + C

XCR1GFP

2

CXCR2RLu

c + C

XCR2GFP

2

CXCR2RLu

c + C

XCR1GFP

0.0

0.1

0.2

0.3

0.4Basal+ 50nM IL-8

BRET

2 Rat

io

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

Figure 8

Flag

CXC

R1/

Myc

-CXC

R1

Mix

Flag

CXC

R1/

Myc

-CXC

R1

Co

Flag

CXC

R2/

Myc

-CXC

R2

Mix

Flag

CXC

R2/

Myc

-CXC

R2

Co

Flag

CXC

R2/

Myc

-CXC

R1

Mix

Flag

CXC

R2/

Myc

-CXC

R1

Co

Flag

CXC

R1/

Myc

-CXC

R2

Mix

Flag

CXC

R1/

Myc

-CXC

R2

Co

-0.02

-0.01

0.00

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

Basal+ 100nM IL-8

Tr-F

RET

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

Figure 9A

Non-Permeabilized Permeabilized

a b

c

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

Figure 9B

P = 0.38

Flag-CXCR1 Flag-CXCR2 Flag-α1A0

2500

5000

7500

10000

12500

+ HA-CXCR1+ HA-CXCR1-ER

Fluo

resc

ence

*** ***

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

Figure 10

-10.5 -9.5 -8.5 -7.5 -6.5 -5.5-25

0

25

50

75

100

125

Log [GRO-α] (M)

% in

hibi

tion

of F

orsk

olin

stim

ulat

ion

Receptor(s) expressed pIC50 GRO-α (M)

CXCR1 -6.8 ± 0.2 CXCR2 -8.1 ± 0.1

CXCR1 and CXCR2 mixed -7.5 ± 0.2 CXCR1 and CXCR2 co-expressed -7.7 ± 0.2

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from

Shirley Wilson, Graeme Wilkinson and Graeme Milliganselectively and with equivalent affinity

The CXCR1 and CXCR2 receptors form constitutive homo and hetero-dimers

published online June 9, 2005J. Biol. Chem. 

  10.1074/jbc.M413475200Access the most updated version of this article at doi:

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

by guest on February 17, 2018http://w

ww

.jbc.org/D

ownloaded from