Label-free impedance responses of endogenous and synthetic chemokine receptor CXCR3 agonists correlate with Gi-protein pathway activation

Biochemical and Biophysical Research Communications 419 (2012) 412–418

Contents lists available at SciVerse ScienceDirect

Biochemical and Biophysical Research Communications journal homepage: www.elsevier.com/locate/ybbrc

Label-free impedance responses of endogenous and synthetic chemokine receptor CXCR3 agonists correlate with Gi-protein pathway activation Anne O. Watts a, Danny J. Scholten a, Laura H. Heitman b, Henry F. Vischer a,⇑, Rob Leurs a a Leiden/Amsterdam Center for Drug Research (LACDR), Division of Medicinal Chemistry, Faculty of Sciences, VU University Amsterdam, De Boelelaan 1083, 1081 HV Amsterdam, The Netherlands b Leiden/Amsterdam Center for Drug Research (LACDR), Division of Medicinal Chemistry, Leiden University, P.O. Box 9502, 2300 RA Leiden, The Netherlands

a r t i c l e

i n f o

Article history: Received 18 January 2012 Available online 13 February 2012 Keywords: Label-free impedance measurement CXCR3 Chemokine b-Arrestin G-protein Biased ligands

a b s t r a c t The chemokine receptor CXCR3 is a G-protein-coupled receptor that signals through the Gai class of heterotrimeric G-proteins. CXCR3 is highly expressed on activated T cells and has been proposed to be a therapeutic target in autoimmune disease. CXCR3 is activated by the chemokines CXCL9, CXCL10 and CXCL11. CXCR3 signaling properties in response to CXCL10, CXCL11 and the synthetic agonist VUF10661 have previously been evaluated using conventional endpoint assays. In the present study, label-free impedance measurements were used to characterize holistic responses of CXCR3-expressing cells to stimulation with chemokines and VUF10661 in real time and to compare these responses with both G-protein and non-G-protein (b-arrestin2) mediated responses. Differences in response kinetics were apparent between the chemokines and VUF10661. Moreover, CXCR3-independent effects could be distinguished from CXCR3-specific responses with the use of the selective CXCR3 antagonist NBI-74330 and the Gai inhibitor pertussis toxin. By comparing the various responses, we observed that CXCL9 is a biased CXCR3 agonist, stimulating solely G-protein-dependent pathways. Moreover, CXCR3-mediated changes in cellular impedance correlated with G-protein signaling, but not b-arrestin2 recruitment. Ó 2012 Elsevier Inc. All rights reserved.

1. Introduction The chemokine receptor CXCR3 is one of the 19 known human chemokine receptors, a subfamily of G-protein-coupled receptors (GPCRs) involved in the migration of immune cells along chemotactic gradients to sites of infection and inflammation [1]. CXCR3 is expressed on many cell types, including T lymphocytes, dendritic cells, endothelial cells and cancer cells. CXCR3 expression is enhanced during inflammation by interferon-c and plays a role in wound healing, autoimmune disease, immune responses to various pathogens, and has an angiostatic effect [2–5]. CXCR3 is activated by the chemokines CXCL9, CXCL10 and CXCL11, which are produced and released by monocytes and macrophages at sites of inflammation [6]. CXCR3 mediates chemotaxis towards these chemokines in a Gai2-dependent manner [3–5]. Additionally, it was recently shown that CXCR3 recruits b-arrestins in response to CXCL10 and CXCL11 [7]. CXCL9, CXCL10, and CXCL11 utilize different extracellular and intracellular CXCR3 regions for binding and signaling, respectively, and induce CXCR3 internalization through different mechanisms [8–11]. The interest

⇑ Corresponding author. Fax: +31 20 5987610. E-mail address: [email protected] (H.F. Vischer). 0006-291X/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2012.02.036

in CXCR3 as a pharmacological target has resulted in the generation of a range of pharmacological tools to study CXCR3 function [6,12,13]. CXCR3 antagonists may have therapeutic potential in the treatment of autoimmune diseases [14] and cancer [15], whereas CXCR3 agonists may aid wound healing [16], respectively. Evolving insights in GPCR pharmacology have unmasked the complex and multidimensional signaling capabilities of these receptors in response to distinct agonists. Importantly, in the last decade, numerous ligands have been identified and/or reclassified that appear to activate one pathway more efficiently than the other while binding to the same GPCR. Identification and full characterization of these so-called ‘‘biased’’ ligands requires analysis of all signaling pathways that are potentially activated by the GPCR of interest. The great majority of biochemical and biophysical assays to detect the activation of a selected pathway requires the introduction of chemical or biological biosensors to detect signaling and/or accumulation of second messengers. These manipulations can change the stoichiometry between receptor and intracellular effector molecules to such an extent that the coupling efficiency of the pathway is unintentionally altered, resulting in an incorrect estimation of potency and/or efficacy of ligands [17]. In recent years, labelfree technologies based on impedance and optical sensors have been developed and applied to allow detection of GPCR-mediated signaling [18,19]. Great advantages of these label-free approaches

A.O. Watts et al. / Biochemical and Biophysical Research Communications 419 (2012) 412–418

are their holistic and non-invasive nature, allowing recording of integrated cellular responses upon receptor activation without preselecting and/or interfering with signaling pathways. A good example of these advantages – for studying chemokine receptor signaling in particular – is the sensitive detection of Gai functional responses. With other ‘‘classical’’ assays, Gai-mediated effects can be detected indirectly either by inhibition of forskolin-induced functional responses or with the transfection of chimeric G-proteins [20]. In addition, Gi-mediated stimulation of phospholipase Cb via the release of Gbc subunits is inefficient as compared to Gq-coupled GPCRs, or even undetectable [21,22]. Recent publications demonstrate the use of this label-free technology for measuring ligand potency as well as determining target selectivity and ligand-biased function [23–26]. In the present study, integrated responses of CXCR3-expressing cells were measured as real-time changes in electrical impedance upon stimulation with CXCL9, CXCL10, CXCL11, and the synthetic agonist VUF10661. Use of the selective CXCR3 antagonist NBI74330 and the Gai inhibitor pertussis toxin allowed identification of CXCR3-dependent and -independent effects of these ligands. The label-free responses were correlated with the activities of the CXCR3 ligands in both G-protein and non-G-protein-coupled responses. 2. Material and methods 2.1. Materials Parental HEK293 cells and HEK293 cells stably expressing human CXCR3 (HEK293-CXCR3) were a gift from Dr. K. Biber (University Medical Center Groningen, the Netherlands). HEK293T cells were purchased from ATCC-LGC Standards (Wesel, Germany). Cell culture media, penicillin/streptomycin and bovine serum albumin (BSA) were purchased from PAA Laboratories GmbH (Paschen, Austria). Fetal bovine serum (FBS) was obtained from Integro B.V. (Dieren, the Netherlands). Linear 25-kDa polyethylenimine was purchased from Polysciences (Warrington, PA). G418, poly-L-lysine, pertussis toxin, Triton X-100, disodium pyrophosphate, adenosine 50 -triphosphate disodium salt, Tris base, glycerol, phosphoric acid and carbamylcholine chloride (carbachol) were purchased from Sigma–Aldrich (St. Louis, MO). Beetle luciferin and coelenterazine h were purchased from Promega (Madison, WI). Chemokines were purchased from Peprotech (Rocky Hill, NJ). VUF10661 and NBI-74330 were synthesized as described previously [7,27]. [125I]CXCL10 was purchased from PerkinElmer (Waltham, MA). Dithiothreitol was purchased from Duchefa Biochemie (Haarlem, the Netherlands). 2.2. DNA constructs The cyclic AMP response element-luciferase (CRE-Luc) reporter gene plasmid was a kind gift from Dr. W. Born (National Jewish Medical and Research Center, Denver, CO., USA). CXCR3-Renilla luciferase (RLuc) and b-arrestin2-enhanced yellow fluorescent protein (eYFP) fusion constructs have been described previously [7]. 2.3. Cell culture and transfection Cells were maintained in 5% CO2 at 37 °C in Dulbecco’s Modified Eagle Medium (DMEM) containing 10% FBS, penicillin, and streptomycin. To maintain stable expression of CXCR3 (HEK293-CXCR3) the medium was supplemented with 250 lg/ml G418. HEK293 and HEK293-CXCR3 cells were transiently transfected with 500 ng/ 1  106 cells CRE-Luc reporter plasmid using linear polyethylenimine, as described previously [13]. For b-arrestin2 recruitment

413

assays, HEK293T cells were transfected with cDNA encoding CXCR3-RLuc and b-arrestin2-eYFP, as described in [7]. 2.4. HEK293 cell membrane preparation and [125I]-CXCL10 binding Preparation of HEK293 cell membrane fractions and [125I]CXCL10 competition binding were performed as described previously [13]. 2.5. CRE-Luc reporter gene One day after transfection, cells were transferred to poly-Llysine-coated white 96-well plates. The following day, culture medium was replaced with serum-free DMEM (penicillin/streptomycin) containing 0.05% BSA and the indicated ligands. After 5 h, stimulation medium was replaced by 25 ll substrate solution (39 mM TrisH3PO4 pH 7.8, 39% glycerol, 2.6% Triton X-100, 860 lM dithiothreitol, 18 mM MgCl2, 825 lM ATP, 77 lM disodium pyrophosphate, 230 lg/mL beetle luciferin) and luminescence was measured using a Victor3 multi label plate reader (PerkinElmer, Waltham, MA). 2.6. BRET-based b-arrestin2 recruitment One day after transfection, cells were transferred to poly-Llysine-coated white 96-well plates. The following day, medium was replaced with Hank’s Balanced Salt Solution and fluorescence was measured on a Victor3 multi label plate reader (ex. 485 nm; em. 535 nm). Ten minutes after coelenterazine h (5 lM final concentration) addition, ligand solutions in Hank’s Balanced Salt Solution supplemented with 0.05% BSA were added in the stated concentrations and incubated for a further 5 min. Bioluminescence resonance energy transfer (BRET) (em. 535 nm) and RLuc expression (em. 460 nm) were measured with a Victor3 multi label plate reader. Baseline-corrected BRET ratios were calculated by first dividing BRET by RLuc emission values, followed by subtraction of the BRET ratio of cells expressing CXCR3-RLuc alone. 2.7. Impedance measurement The xCELLigence system (Roche Applied Science, Penzberg, Germany) uses E-plates (ACEA Biosciences Inc, San Diego, CA) with electrode arrays integrated into the bottom of the wells, which allow measurement of impedance at the electrode-cell interface. Data is recorded at an electrical current of 10 kHz and automatically converted to Cell Index (CI) by the xCELLigence software. CI = (Rtn  Rt0)/15(X), where Rt0 is the background electrical resistance measured in the absence of cells prior to the start of each experiment; Rtn is the electrical resistance at time point n. Parental HEK293 cells or HEK293-CXCR3 were seeded in DMEM containing 10% FBS, penicillin and streptomycin at 5  104 cells per well of 96well E-plate. The plate was inserted into the xCELLigence station and maintained in 5% CO2 at 37 °C while CI was measured with 15-min intervals. The cells were allowed to grow until a CI of 1.7–2.2 was reached, approximately 20 h after seeding the cells. At this point, medium was replaced with 90 or 95 ll DMEM containing 0.1% FBS, penicillin and streptomycin, and cells were allowed to adjust to low-serum conditions for 3 h. Next, the E-plate lid was replaced with an open lid and ligands were added with the aid of an automated multichannel pipette (5 ll/well), while the E-plate remained in the xCELLigence station. Immediately prior to ligand addition, the measurement frequency was increased to 30-s intervals, followed by 1-min, 5-min and, finally, 15-min intervals. CI values were normalized to CI values just prior to ligand addition (time is 0 min) and were baseline-corrected with CI traces obtained from cells treated with the appropriate vehicle solution.

414

A.O. Watts et al. / Biochemical and Biophysical Research Communications 419 (2012) 412–418

Table 1 Affinities, potencies and intrinsic activities of CXCL9, CXCL10, CXCL11 and the small CXCR3 agonist VUF10661 in pathway-specific and label-free assays. pKi values were determined in [125I]-CXCL 10homologous displacement experiments, pEC50 values for CRE-Luc inhibition, BRET-based b-arrestin2 recruitment and impedance assays are given as averages ± SEM of 2–6 independent experiments. Intrinsic activities (a) were calculated from averaged normalized data. Statistical differences in pEC50 were determined for each assay by one-way ANOVA followed by a Bonferroni’s multiple comparison test (p < 0.05). 125

Ligand CXCL9 CXCL10 CXCL11 VUF10661 a b c d

I-CXCL10

CRE-Luc

pKi 8.1 ± 0.1 9.1 ± 0.2 10.4 ± 0.1 7.3 ± 0.1

BRET b-arrestin2

a

pEC50 8.1 ± 0.1 9.1 ± 0.2c 9.6 ± 0.1c 7.1 ± 0.2

1.0 1.0 1.0 1.0

Impedance

a

pEC50 n.d.a 6.9 ± 0.1 8.2 ± 0.1 5.9 ± 0.2

a

pEC50 6.8 ± 0.1d 7.3 ± 0.1d 7.2 ± 0.2d 6.1 ± 0.1

0a 0.5 1.0 1.9

1.3b 0.6 1.0 1.0

n.d. = not detectable. CXCL9 not included in calculation due to large CXCR3-independent effect. potencies in the CRE-Luc reporter gene assay did not differ significantly from each other. potencies in the electrical impedance assay did not differ significantly from each other.

A 0.5

B

0.4

Baseline Normalized CI

Baseline Normalized CI

0.4

0.3

0.2

0.1

0.3

0.2

0.1

0.0

0.0 20

40

60

80

20

100

40

60

80

100

80

100

time (min)

time (min)

-0.1

-0.1

C 0.5

D

0.5

0.4

Baseline Normalized CI

0.4

Baseline Normalized CI

0.5

0.3

0.2

0.1

0.0

0.3

0.2

0.1

0.0 20

40

60

80

100

20

time (min) -0.1

40

60

time (min) -0.1

Fig. 1. Chemokines and VUF10661 induce transient impedance changes of CXCR3-expressing cells. Cells were treated with increasing concentrations of the chemokines (A) CXCL10, (B) CXCL11, (C) CXCL9, and (D) the small CXCR3 agonist VUF10661. (+)100 pM, (j) 1 nM, (d) 3.2 nM, () 10 nM, (h) 32 nM, (s) 100 nM, (e) 316 nM, () 1 lM, (4) 3.2 lM and (d) 10 lM. Data are presented as averages of 2–4 independent experiments ± SEM.

2.8. Data analysis Nonlinear curve fitting and statistical analyses of the data were performed using Prism 4.03 (GraphPad Software Inc., San Diego, CA). Ki values were calculated using the Cheng–Prusoff equation Ki = IC50/(1 + [125I-CXCL10]/Kd of 125I-CXCL10) [28]. Peak values of the CI traces from 0 to 100 min after ligand addition were determined by the xCELLigence software, and used to construct dose response curves in Prism 4.03. Intrinsic activities (a) were calculated for all tested ligands by dividing their maximum response in the

functional assays by the maximum response of the endogenous full agonist CXCL11.

3. Results and discussion Binding affinities (pKi) of the chemokines CXCL9, CXCL10, CXCL11, and the small synthetic compound VUF10661 for CXCR3 were determined using [125I]-CXCL10 competition assays on membranes from HEK293-CXCR3 cells. All four ligands are able to

415

A.O. Watts et al. / Biochemical and Biophysical Research Communications 419 (2012) 412–418

Baseline Normalized CI

0.5 0.4 0.3 0.2 0.1 0.0 -10

-9

-8

-7

-6

-5

[ligand] (Log10M) Fig. 2. Concentration–response curves of peak CI values induced by CXCL9, CXCL10, CXCL11, and VUF10661 in CXCR3-expressing cells. Cells were treated with increasing concentrations of (e) CXCL9, (h) CXCL10, (s) CXCL11, and (j) VUF10661. Data are presented as averages of 2–4 independent experiments ± SEM.

completely inhibit [125I]-CXCL10 binding to CXCR3 (data not shown). CXCL11 displayed a 20 and 200-fold higher affinity for CXCR3 in comparison to CXCL10 and CXCL9, respectively, whereas the affinity of VUF10661 was approximately 1000-fold lower than CXCL11 (Table 1).

To evaluate G-protein-dependent CXCR3 signaling, HEK293CXCR3 cells were transiently transfected with the CRE-Luc reporter gene plasmid. All four ligands act as full agonists in inhibiting forskolin (1.5 lM)-induced CRE activity with potencies (pEC50) in the same rank order as the CXCR3 binding affinities (Table 1). This CXCR3-induced attenuation of adenylate cyclase activity can be fully inhibited by pertussis toxin pre-treatment (100 ng/ml, 16 h), indicating the involvement of Gi-proteins (data not shown). Full agonism of both CXCL11 and VUF10661 was recently also shown in Gi activation as measured by [35S]-GTPcS binding to CXCR3expressing membranes, in which CXCL11 displayed a 1000-fold higher potency than VUF10661 [7]. As a G-protein-independent readout, ligand-induced b-arrestin2 recruitment to CXCR3 was measured using a BRET-based assay in transiently transfected HEK293T cells. In contrast to the Giprotein-dependent inhibition of CRE activity, the four ligands have different intrinsic activities to induce b-arrestin2 recruitment to CXCR3 (Table 1). CXCL11 and CXCL10 are full and partial agonists, respectively, in recruiting b-arrestin2 to CXCR3, whereas VUF10661 acts as ‘‘superagonist’’ (a = 1.9), as shown previously [7]. However, saturating concentrations (1 lM) of CXCL9 did not result in any significant b-arrestin2 recruitment, revealing for the first time that CXCL9 is a biased CXCR3 agonist with preference for Gi-protein signaling. Indeed, CXCL9-induced CXCR3 internalization was previously suggested to be b-arrestin2 independent [8]. CXCL11, CXCL10, and VUF10661 showed the same rank order in potency in CRE-Luc reporter gene and b-arrestin2 recruitment assays, although their

0.5

0.5

Baseline Normalized CI

B 0.6

Baseline Normalized CI

A 0.6

0.4 0.3 0.2 0.1 0.0

0.4 0.3 0.2 0.1 0.0

20

40

60

80

100

20

time (min)

60

80

100

80

100

time (min)

D 0.6

0.5

0.5

Baseline Normalized CI

C 0.6 Baseline Normalized CI

40

0.4 0.3 0.2 0.1 0.0

0.4 0.3 0.2 0.1 0.0

20

40

60

time (min)

80

100

20

40

60

time (min)

Fig. 3. The CXCR3-specific antagonist NBI-74330 reveals CXCR3-mediated and CXCR3-independent effects of CXCR3 chemokines and VUF10661. HEK293-CXCR3 cells were treated with (h) 0.1% DMSO control, (s) 10 lM NBI-74330 10 min prior to addition of (A) 32 nM CXCL10, (B) 32 nM CXCL11, (C) 316 nM CXCL9 or (D) 10 lM VUF10661. Parental HEK293 cells that do not express CXCR3 were treated with CXCR3 agonists without pre-treatment (j). Data are presented as averages of 2–3 independent experiments ± SEM.

416

A.O. Watts et al. / Biochemical and Biophysical Research Communications 419 (2012) 412–418

A

B

0.5

0.4

Baseline Normalized CI

Baseline Normalized CI

0.4

0.5

0.3

0.2

0.1

0.0

0.3

0.2

0.1

0.0 20

40

60

80

100

20

40

time (min)

C

D

0.5

80

100

0.5

0.4

Baseline Normalized CI

0.4

Baseline Normalized CI

60

time (min)

0.3

0.2

0.1

0.0

0.3

0.2

0.1

0.0 20

40

60

80

100

time (min)

20

40

60

80

100

time (min)

Fig. 4. Changes in CI induced by chemokines and a small CXCR3 agonist are mediated by Gai. HEK293-CXCR3 cells were grown overnight in the (h) absence or (s) presence of 100 ng/mL PTX. Cells were treated with (A) 32 nM CXCL10, (B) 32 nM CXCL11, (C) 316 nM CXCL9 or (D) 10 lM VUF10661 were measured. Data are presented as averages of 2–3 independent experiments ± SEM.

potencies were 1.2 to 2.2 orders of magnitude lower for b-arrestin recruitment than G-protein signaling. To explore CXCR3-induced integrated cellular responses in real time, changes in electrical impedance were measured using the xCELLigence system upon treatment of HEK293-CXCR3 cells with increasing concentrations of CXCL9, CXCL10, CXCL11, and VUF10661 (Fig. 1). The chemokines CXCL9, CXCL10 and CXCL11 resulted in changes of cellular impedance with comparable shapes, showing a peak at 8 min followed by a shoulder at 15–20 min after ligand addition (Fig. 1A–C). These impedance traces of CXCL9, CXCL10 and CXCL11 did not return to baseline CI, even in measurements up to 3 h after ligand addition (data not shown), but stabilized at a higher level, approximately one hour after ligand addition. Interestingly, peak CI values induced by CXCL9 were equally high or higher than those reached in response to equal concentrations of the other two chemokines. This was surprising, given the low potency of CXCL9 in inhibiting CRE activity on one hand, and the lack of efficacy in recruiting b-arrestin2 on the other hand. Moreover, the CXCL9 trace stabilized at higher CI values than CXCL10 and CXCL11. In contrast, VUF10661 induced traces that lack a shoulder and stabilized more quickly, approximately 30 min after ligand addition (Fig. 1D). The high efficacy of b-arrestin2 recruitment by VUF10661 (Table 1 and [7]) can be hypothesized to result in a more rapid CXCR3 desensitization, which could explain both differences in CI response kinetics as compared to the chemokines. Peak CI values in the first 100 min after ligand addition were used to generate concentration–response curves from which ligand potencies were determined (Fig. 2; Table 1).

CXCL9, CXCL10, and CXCL11 have comparable potencies in the electrical impedance assay. This is in contrast to the potency rank order observed in the CRE-Luc reporter gene and b-arrestin2 recruitment assays, in which CXCL11 has the highest and CXCL9 has the lowest or even no potency, respectively. VUF10661 has a significantly lower potency than the chemokines in the electrical impedance assay, which is in agreement with the rank order observed in the CRE activity and b-arrestin2 recruitment assays. To confirm that CI changes by CXCL9, CXCL10, CXCL11, and VUF10661 are induced upon their interaction with CXCR3, the cells were pre-incubated with a saturating concentration (10 lM) of the CXCR3-selective antagonist NBI-74330 [13]. NBI-74330 completely inhibited CXCL10 and CXCL11-induced impedance responses in HEK293-CXCR3 cells (Fig. 3A and B). In contrast, NBI-74330 only partially inhibited responses to CXCL9 and VUF10661 (Fig. 3C and D), suggesting that both these ligands affect impedance of HEK293-CXCR3 cells in both CXCR3-dependent and independent manner. Interestingly, NBI-74330 was previously shown to fully inhibit VUF10661-induced Gi-protein signaling and b-arrestin recruitment in CXCR3-expressing cells [7]. To confirm the CXCR3-independent effects of CXCL9 and VUF10661 on impedance, all four ligands were tested in the parental HEK293 cell line, which does not endogenously express CXCR3 [29]. CXCL10 and CXCL11 did not change the impedance of parental HEK293 cells (Fig. 3A and B). However, CXCL9 and VUF10661 induced CI traces with a similar shape in parental HEK293 cells in comparison to HEK293-CXCR3 cells that were pre-treated with NBI-74330, with higher amplitude for CXCL9 in parental cells (Fig. 3C and D). Under

A.O. Watts et al. / Biochemical and Biophysical Research Communications 419 (2012) 412–418

these conditions the CI values had not stabilized 100 min after CXCL9 stimulation, which is in contrast to the HEK293-CXCR3 cells that were only stimulated with CXCL9. The involvement of Gi-proteins in impedance responses to the four agonists was determined by pre-treating the HEK293-CXCR3 cells with PTX, which has previously been shown not to affect b-arrestin recruitment to CXCR3 [7]. Both CXCL10 and CXCL11 responses were completely inhibited by PTX (Fig. 4A and B), indicating that the CXCR3-induced changes in impedance are fully mediated via Gi-proteins. In contrast, CXCL9 and VUF10661-induced impedance changes were only partially inhibited by PTX, and to a similar extent as when antagonizing CXCR3 with NBI-74330 (Fig. 4C and D). Hence, CXCL9 and VUF10661 appear to activate some cellular responses in a CXCR3-independent and PTX-insensitive manner, which has previously been described for CXCL9 in peripheral blood mononuclear cells and the monocytic cell line THP-1 [30]. These effects were found to be dependent on p38 and ERK1/2 phosphorylation. However, the alternative receptor targeted by CXCL9 remains unidentified. In contrast to changes in impedance mediated by agonist stimulation of the b2 adrenergic receptor, to which both Gi and Gs were shown to contribute [31], electrical impedance changes mediated by CXCR3 are apparently completely dependent on Gi activity. NBI-74330 and PTX did not affect the impedance response to 10 lM carbachol, which is mediated via the endogenously expressed Gaq-coupled muscarinic M3 receptor (Supplementary Fig. 1), indicating that the observed inhibitory effects on CXCR3-mediated impedance are not due to non-specific effects of the treatments. Since PTX completely inhibited CXCR3-dependent responses of all four agonists in the electrical impedance assay, this suggests that the increase in cellular impedance reflects Gai-protein activity and not b-arrestin recruitment. However, b-arrestin may influence the kinetics of the impedance response. Interestingly, agonists on the Gi-coupled niacin receptors that activate both G-protein signaling and b-arrestin recruitment, induced a transient dip in CI in the first six minutes after stimulation [26]. This dip was absent in the CI traces of ligands that only activate niacin receptor-mediated Gprotein signaling but were unable to recruit b-arrestin. Although VUF10661 has been shown to have a higher b-arrestin recruitment efficacy than CXCR3 chemokines [7], no differences were apparent between chemokines and VUF10661 in the first six minutes of the CXCR3-mediated impedance traces. It is possible that the b-arrestin-biased niacin receptor ligands are more efficacious than VUF10661, resulting in a more pronounced signal in impedance measurements. However, it is clear that additional information (e.g. data from cells treated with RNAi targeting b-arrestin) is required to identify a possible role of b-arrestin in the electrical impedance kinetics downstream of GPCR activity. In conclusion, real-time impedance measurements readily allow detection of Gai activity. In addition to the Gai response mediated by CXCR3, CXCR3-independent and PTX-insensitive effects were integrated into the signals of CXCL9 and VUF10661. Such phenotypic information is advantageous for insights in integrated signaling and potentially unwanted effects of ligands, though it also highlights the need for caution in interpreting data obtained from these label-free experiments. Acknowledgement This work was supported by the Dutch Top Institute Pharma (Project Number D1-105: the GPCR forum). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.bbrc.2012.02.036.

417

References [1] D.J. Scholten, M. Canals, D. Maussang, et al., Pharmacological modulation of chemokine receptor function, Br. J. Pharmacol. (2011), doi:10.1111/j.14765381.2011.01551.x. [2] E.J.A. Van Wanrooij, S.C.A. De Jager, T. Van Es, et al., CXCR3 antagonist NBI74330 attenuates atherosclerotic plaque formation in LDL receptor-deficient mice, Arterioscler Thromb. Vasc. Biol. 28 (2) (2008) 251–257. [3] M.J. Smit, P. Verdijk, E.M. Van der Raaij-Helmer, et al., CXCR3-mediated chemotaxis of human T cells is regulated by a Gi- and phospholipase Cdependent pathway and not via activation of MEK/p44/p42 MAPK nor Akt/PI-3 kinase, Blood 102 (6) (2003) 1959–1965. [4] B.D. Thompson, Y. Jin, K.H. Wu, et al., Inhibition of G alpha i2 activation by G alpha i3 in CXCR3-mediated signaling, J. Biol. Chem. 282 (13) (2007) 9547– 9555. [5] Y.Z. Jin, B.D. Thompson, Z.Y. Zhou, et al., Reciprocal function of Galphai2 and Galphai3 in graft-versus-host disease, Eur. J. Immunol. 38 (7) (2008) 1988– 1998. [6] C.E. Heise, A. Pahuja, S.C. Hudson, et al., Pharmacological characterization of CXC chemokine receptor 3 ligands and a small molecule antagonist, J. Pharmacol. Exp. Ther. 313 (2005) 1263–1271. [7] D. Scholten, M. Canals, M. Wijtmans, et al., Pharmacological characterisation of a small-molecule agonist for the chemokine receptor CXCR3, Br. J. Pharmacol. (2011), doi:10.1111/j.1476-5381.2011.01648.x. [8] R.A. Colvin, G.S.V. Campanella, J. Sun, et al., Intracellular domains of CXCR3 that mediate CXCL9, CXCL10, and CXCL11 function, J. Biol. Chem. 279 (29) (2004) 30219–30227. [9] M.A. Cox, C.-H. Jenh, W. Gonsiorek, et al., Human interferon-inducible 10-kDa protein and human interferon-inducible T cell chemoattractant are allotopic ligands for human CXCR3: Differential binding to receptor states, Mol. Pharmacol. 59 (2001) 707–715. [10] G. Xanthou, T.J. Williams, J.E. Pease, Molecular characterization of the chemokine receptor CXCR3: evidence for the involvement of distinct extracellular domains in a multi-step model of ligand binding and receptor activation, Eur. J. Immunol. 33 (2003) 2927–2936. [11] R.A. Colvin, G.S.V. Campanella, L.A. Manice, et al., CXCR3 requires tyrosine sulfation for ligand binding and a second intracellular loop arginine residue for ligand-induced chemotaxis, Mol. Cell Biol. 26 (15) (2006) 5838–5849. [12] I.L. Stroke, A.G. Cole, S. Simhadri, et al., Identification of CXCR3 receptor agonists in combinatorial small-molecule libraries, Biochem. Biophys. Res. Commun. 349 (1) (2006) 221–228. [13] D. Verzijl, S. Storelli, D.J. Scholten, et al., Noncompetitive antagonism and inverse agonism as mechanism of action of nonpeptidergic antagonists at primate and rodent CXCR3 chemokine receptors, J. Pharmacol. Exp. Ther. 325 (2) (2008) 544–555. [14] S. Lacotte, S. Brun, S. Muller, et al., CXCR3, inflammation and auto-immune diseases, Ann. NY Acad. Sci. 1173 (2009) 310–317. [15] M.L. Balestrieri, A. Balestrieri, F.P. Mancini, et al., Understanding the immunoangiostatic CXC chemokine network, Cardio. vasc. Res. 78 (2) (2008) 250–256. [16] C.C. Yates, P. Krishna, D. Whaley, et al., Lack of CXC chemokine receptor 3 signaling leads to hypertrophic and hypercellular scarring, Am. J. Pathol. 176 (4) (2010) 1743–1755. [17] T.P. Kenakin, Cellular assays as portals to seven-transmembrane receptorbased drug discovery, Nat. Rev. Drug Discov. 8 (8) (2009) 617–626. [18] M. Yu, J.M. Atienza, J. Bernard, et al., Real-time monitoring of morphological changes in living cells by electronic sensor array: an approach to study G protein-coupled receptors, Anal. Chem. 78 (1) (2006) 35–43. [19] R. Schröder, J. Schmidt, S. Blättermann, et al., Applying label-free dynamic mass redistribution technology to frame signaling of G protein-coupled receptors noninvasively in living cells, Nat. Protoc. 6 (11) (2011) 1748–1760. [20] M. Hamel, M. Henault, H. Hyjazie, et al., Discovery of P2Y14 agonist and antagonist using conventional and nonconventional methods, J. Biomol. Screen 16 (9) (2011) 1098–1105. [21] T.D. Werry, G.F. Wilkinson, G.B. Willars, Mechanisms of cross-talk between Gprotein-coupled receptors resulting in enhanced release of intracellular Ca2+, Biochem. J. 374 (Pt 2) (2003) 281–296. [22] E.M. Rosethorne, J.R. Leighton-Davies, D. Beer, et al., ATP priming of macrophage-derived chemokine responses in CHO cells expressing the CCR4 receptor, Naunyn Schmiedebergs Arch. Pharmacol. 370 (1) (2004) 64–70. [23] R.P. McGuinness, J.M. Proctor, D.L. Gallant, et al., Enhanced selectivity screening of GPCR ligands using a label-free cell based assay technology, Comb. Chem. High Throughput Screen 12 (8) (2009) 812–823. [24] M.F. Peters, C.W. Scott, Evaluating cellular impedance assays for detection of GPCR pleiotropic signaling and functional selectivity, J. Biomol. Screen 14 (3) (2009) 246–255. [25] A.N. Flynn, D.V. Tillu, M.N. Asiedu, et al., The protease-activated receptor-2 specific agonists 2-aminothiazol-4-yl-LIGRL-NH2 and 6-aminonicotinylLIGRL-NH2 stimulate multiple signaling pathways to induce physiological responses in vitro and in vivo, J. Biol. Chem. 286 (21) (2011) 19076–19088. [26] M. Kammermann, A. Denelavas, A. Imbach, et al., Impedance measurement: a new method to detect ligand-biased receptor signaling, Biochem. Biophys. Res. Commun. 412 (3) (2011) 419–424.

418

A.O. Watts et al. / Biochemical and Biophysical Research Communications 419 (2012) 412–418

[27] S. Storelli, D. Verzijl, J. Al-Badie, et al., Synthesis and structure-activity relationships 3H-quinazolin-4-ones and 3H-pyrido[2,3-d]pyrimidin-4-ones as CXCR3 receptor antagonists, Arch. Pharm, (Weinheim) 340 (6) (2007) 281–291. [28] Y. Cheng, W.H. Prusoff, Relationship between the inhibition constant (K1) and the concentration of inhibitor which causes 50 percent inhibition (I50) of an enzymatic reaction, Biochem. Pharmacol. 22 (23) (1973) 3099–3108. [29] B.K. Atwood, J. Lopez, Wager-Miller, et al., Expression of G protein-coupled receptors and related proteins in HEK293, AtT20, BV2, and N18 cell lines as revealed by microarray analysis, BMC Genomics 12 (2011) 14.

[30] J.H. Gong, E.F. Nicholls, M.R. Elliott, et al., G-protein-coupled receptor independent immunomodulatory properties of chemokine CXCL9, Cell Immunol. 261 (2) (2010) 105–113. [31] W. Stallaert, J.F. Dorn, E. van der Westhuizen, et al., Impedance responses reveal (2)-adrenergic receptor signaling pluri dimensionality and allow classification of ligands with distinct signalling profiles, PLoS One 7 (1) (2012) e29420.