Sage Journals: Discover world-class research

Abstract

Cellular dielectric spectroscopy (CDS) provides realtime, label-free, universal measurements, enabling comprehensive pharmacological evaluation of cell surface receptors in living cells. The sensitivity of the measurement allows monitoring of ligand-mediated activation of endogenous receptors, therefore generating physiologically relevant data. Activation of receptors results in CDS response profiles that are characteristic of main subsets of G-protein coupled receptors (GPCRs) within a cell line. This allows cluster analysis of response profiles that may be used in several important applications, which include identification of the G-protein coupling of orphan GPCRs and the cataloging of active endogenous receptors in cells. In this study, CDS technology is used in the pharmacological evaluation of multiple receptors in many cell types, including primary cells. Specifically, data is presented demonstrating hit confirmation, receptor selectivity analysis, ligand potency, and Schild analysis of receptor-selective antagonists. CDS results compare favorably to other cell-based assays, and the robustness and reproducibility of CDS assays are reflected by low assay coefficient of variation (CVs) and reliable Z'-scores of the data. Because CDS requires no stable or transiently transfected cells or special reagents, assay development and data acquisition is simple and fast. The ease of use, universality, and label-free nature of the CDS-based platform make it well suited to secondary screening applications in drug discovery.

Keywords

label-free cell-based assay endogenous receptors GPCR cellular dielectric spectroscopy CDS impedance CellKeyTm system signal transduction

Introduction

Cellular dielectric spectroscopy (CDS) is a novel technology that enables comprehensive pharmacological evaluation of endogenous cell surface receptors using a label-free, real-time, kinetic cell-based assay approach. In addition, CDS is universal, allowing measurement of multiple types of receptors, including G protein-coupled receptors (GPCRs), tyrosine kinase receptors (TKRs), and nuclear receptors using the same platform, without the need for any modification of the cell. The unique information generated by CDS allows identification of receptor-mediated signal transduction, which can be used for further characterization (e.g., deconvolution of orphan GPCRs).

CDS technology is based on monitoring the impedance (Z) of an applied electrical current as it flows through a cell monolayer. Previously published work established the use of impedance measurements for investigating the material properties of cells.^{1

–10} For example, impedance measurements have been successfully applied to monitoring cell membrane conductivity,^{11

–15} monolayer permeability,^{16

–19} morphology,^{20
–22} migration,^23,24 and cellular micromotion.²⁵ Work has also been done that describes the use of a single frequency to measure the activation of cell surface receptors.^{26
–28} The commercially available CellKey™ system²⁹ (Fig. 1) builds upon and extends this body of work by enabling impedance measurements and fluid delivery to the cells to be made simultaneously by incorporating fluidics consisting of an onboard 96-channel dispense head with adjustable dispense speed, pipette height, and mixing capability. In addition, impedance measurements are monitored across a large range of frequencies (1 kHz-110 MHz), yielding more complex information that is characteristic of receptor-mediated signaling pathways being activated.

Figure 1.

Overview of cellular dielectric spectroscopy (CDS). (A) Picture of CellKey™ system. Data in this study were generated on breadboard version of instrument (picture not shown). (B) Component diagram pertaining to both the CellKey™ system and breadboard versions. Each well of a 96-well microplate contains custom electrodes that are connected to an impedance-analyzing device. The system contains a 96-well fluidics delivery device and a thermal system for temperature regulation. All aspects of assay set up, execution, and data acquisition and analysis are controlled by custom applications software. (C) Diagram of a single well of the custom 96-well CDS plate. The CDS instrument supplies constant voltage to cells plated on the electrode producing current flow around and between cells (extracellular current, Iec) and through cells (transcellular current, Itc). (D) Graph illustrating the response when a stimulus such as a receptor ligand is applied. The maroon trace is the impedance change due to changes in extracellular current (ΔZ_Iec), and the blue trace is the impedance change due to changes in transcellular current (ΔZ_Itc). Measurements taken prior to ligand addition (indicated by the arrow) are used as a baseline to quantitate the CDS response.

CDS measures changes in impedance (ΔZ) that occur upon stimulation of different cell surface and nuclear receptors. Impedance is defined by the ratio of voltage to current as described by Ohm's law (Z = V/I). In order to measure receptor-mediated impedance changes, mammalian cells are seeded onto a custom 96-well microplate containing electrodes at the bottom of the wells. The CDS-based instrument supplies constant voltage, which produces a current that flows around and between cells (extracellular current or Iec) and through cells (transcellular current or Itc) (Fig. 1C). Contributors to the impedance measurements are (1) changes in cell adherence to their substrate, (2) changes in cell shape and volume, and (3) changes in cell-cell interactions. These individually, or together, affect the flow of extracellular and transcellular current, and hence, the magnitude and characteristics of the signal measured. Each of these physiological changes can be linked to receptor-mediated signal transduction that results, for example, in changes in cytoskeletal organization.^{30
–32} In addition, non-cytoskeletal-based contributors to the CDS measurement have been identified.

Stimulation of differently coupled GPCRs generates characteristic CDS responses primarily, but not exclusively, due to their differential effects on the actin cytoskeleton, which affect the cellular parameters measured by impedance.^{30
–32} For instance, it has been shown that activation of Gq-^{33
–35} and Gi-GPCRs^36,37 and some tyrosine kinase receptors (TKRs)^38,39 leads to increased actin polymerization, while stimulation of Gs-GPCRs leads to actin depolymerization.^{40

–43} These changes happen within 30 s after receptor activation.^30,31 In addition, direct modulation of intracellular levels of second messengers involved in classical GPCR signaling pathways also induce changes in actin stress fiber formation. Treating cells with agents that elevate the cAMP level caused marked alterations in cell morphology, including a rounded phenotype and a complete loss of actin microfilament bundles^{44
–46} while decreasing cAMP results in increased actin stress fibers.⁴⁷ Phorbol ester-activated protein kinase C induced the early appearance of a great number of actin stress fibers.⁴⁸ As CDS is very sensitive to changes in impedance resulting from both cytoskeletal and noncytoskeletal changes following receptor activation, this platform is uniquely positioned to enable measurement of endogenous receptors in a label-free manner without the need for genetic or chemical modification of the cell or target receptor.

Results and Discussion

Enabling Comprehensive Measurement of Endogenous Receptors

Cataloging endogenous active receptors in different cell types. A key application for CDS is the characterization of functional endogenous receptors within different cell types. The procedure involves treating cells of interest with a panel of ligands targeting multiple receptors of interest. CDS responses induced by these ligands suggest the presence of a functional receptor, which can then be verified with receptor-specific and selective tools (such as agonists, antagonists, and siRNA). An example of this application is provided in Figure 2, which shows HeLa cells treated with a variety of ligands targeting GPCRs, PTKRs, and nuclear receptors. A positive response is detected upon addition of many ligands that activate different endogenous receptors (e.g., dopamine, histamine, epidermal growth factor (EGF), endothelin-1, uridine triphosphate (UTP), stromal cell-derived factor 1 alpha (SDF-1), bradykinin, thrombin, hepatocyte growth factor (HGF), insulin-like growth factor (IGF), UK 14,304, vitamin D, retinoic acid; Fig. 2). Subsequent studies focused on GPCRs and using receptor selective and specific agonists and antagonists have been performed to verify that these responses are receptor-specific (data not shown). Thus, in a single assay, using only one 96-well plate, CDS was used to identify and measure activation of many differently coupled GPCRs, as well as TKRs and nuclear receptors in HeLa cells.

Figure 2.

Measured CDS responses to ligands targeting endogenous receptors. HeLa were treated with compounds representing Gq-coupled, Gs-coupled, and Gi-coupled GPCR agonists (e.g., UK14,304, dopamine, histamine, 5HT), growth factors (e.g., EGF, fibroblast growth factor (FGF), IGF), nuclear receptor stimulators (e.g., vitamin D, retinoic acid), and direct modulators (e.g., ionomycin, thapsigargin, PMA, forskolin) at 28 °C. The CDS response for each ligand was calculated based on the maximum response-induced post-ligand addition. A CDS response was considered a hit if the signal was greater than the buffer control mean +3 standard deviations (SD). Data are mean ±SD, n = 3.

CDS has been used to perform a wide range of studies analyzing the effects of many ligands on endogenous and transfected GPCRs in adherent and nonadherent cells of different origins, including the following cell lines commonly used in drug discovery: CHO, HEK 293, U2-0S, COS-7, HeLa, U937, Jurkat, and THP-1 cells. Some of the endogenous receptors analyzed by CDS include the following GPCRs: prostanoid, calcitonin, calcitonin-GRP, adenosine A_2B, β-adrenoceptor, histamine H₁, histamine H₂ serotonin (5HT_1B), α₂-adrenoceptor, somatostatin, dopamine, opiod, bradykinin, P2Y, endothelin, muscarinic m₃, α1-AR, bombesin, CXCR2, and CXCR4. In addition to GPCR activation, CDS can measure activation of endogenous TKRs (EGF, IGF, FGF, HGF, vascular endothelial growth factor (VEGF) and nuclear receptors (vitamin D, estrogen, progesterone, androgen, retinoic acid, and glucocorticoid).

The ability of the CDS technology to functionally measure activity of endogenous receptors in a wide variety of cell types enables more physiologically relevant selectivity and specificity screens with the goal of providing more successful drug discovery efforts.

Identifying the G-protein coupling mechanism of GPCRs. The data presented thus far demonstrate that CDS is a powerful tool for measuring activation of endogenous receptors. A change in impedance induced upon receptor activation enables screening for the presence of functional receptors in a cell. However, as impedance measurements are collected over a wide range of frequencies in a kinetic manner, data generated using CDS technology can be further mined to yield additional information that is characteristic of the receptor-mediated signaling pathways being activated. Focusing specifically on GPCRs, the unique information generated by CDS technology enables a variety of novel applications including characterizing the G-protein coupling of known and unknown GPCRs. In addition, further characterization of the receptor subtypes identified in the application described above is simplified by the ability to identify the G-protein coupling mechanism involved in receptor signaling using CDS response profiles.

Proprietary data analysis methods generate response profiles from CDS measurements that are characteristic of the downstream signaling initiated by the receptor that is activated. For instance, stimulation of differently coupled GPCRs in CHO cells leads to unique kinetic response profiles for each type of receptor shown in Figure 3. Side-by-side comparisons of the kinetic plots of transcellular impedance (Z_Itc) and extracellular impedance (Z_Iec) for groups of receptors illustrate how stimulation of similarly coupled GPCRs generates similar response profiles that are distinguishable from differently coupled GPCRs (Fig. 3).

Figure 3.

CDS response profiles in CHO cells. Stimulation of differently coupled GPCRs leads to unique kinetic response profiles for each type of receptor, such as the ones shown for the responses in a CHO cell background. Cells were treated with (A) carbachol to activate the transfected muscarinic m₁ receptor (CHO_m1), (B) 5HT to activate the endogenous serotonergic 5HT_1B receptor (CHO_m1), (C) dopamine to activate the transfected dopamine D2L receptor (CHO-D2L), (D) PGE₂ to activate the endogenous prostanoid EP₄ receptor (CHO_m1), (E) calcitonin to activate the endogenous calcitonin receptor (CHO_m1), and (F) dopamine to activate the transfected dopamine D1 receptor (CHO-D1) at 28 °C. The initial drop in impedance that distinguishes the Gq-coupled muscarinic m₁ response from the Gi-coupled 5HT_1B, and dopamine D2L response is circled.

An early, small, and fast decrease in transcellular impedance followed by a later increase in transcellular and extracellular impedance is typical of activation of Gq-protein coupled receptors (Gq-GPCRs) such as the muscarinic m₁ receptor transfected in CHO_m1 cells (Fig. 3A).

Transcellular and extracellular impedance increase following the activation of Gi-protein coupled receptors (Gi-GPCRs) such as the endogenous serotonergic 5HT_1B and transfected dopaminergic D2L receptors (Fig. 3B,C).

Transcellular and extracellular impedance decrease following activation of Gs-protein coupled receptors (Gs-GPCRs) such as the endogenous prostanoid EP4 and calcitonin receptors, and the transfected dopaminergic D1 receptor (Fig. 3D-F).

Dissection of the receptor responses with biochemical modulators revealed that the initial transient decrease in transcellular impedance for the muscarinic m₁ receptor and other Gq-coupled GPCR responses is linked to noncytoskeletal effects induced by intracellular calcium. The later and larger changes in both transcellular and extracellular impedance induced upon receptor activation are primarily due to changes in actin cytoskeleton rearrangement via activation of PKC for Gq-GPCRs and increasing and decreasing intracellular cAMP for Gs- and Gi-GPCRs, respectively (manuscript in preparation).

Single-time snapshots of a response can be displayed in two dimensions with each point representing the characteristic CDS response, as shown in Figure 4. The contribution of both extracellular impedance (Z_Iec) and transcellular impedance (Z_Itc) for a CDS response at a given time point can be plotted on a dot plot. The distance from the origin indicates the magnitude of the response that is both efficacy-and potency-dependent. The angle from the origin is indicative of the relative ratio of Z_Iec and Z_Itc that forms the response profile. This type of graph allows representation of multiple receptor responses on a single dot plot for comparison purposes.

Figure 4.

Two dimensional dot plots. The contribution of Z_Iec is plotted on the x-axis, and Z_Itc is plotted on the y-axis for the CDS response at a given time point. In this example, the response measured 10 min post-ligand addition is indicated by the arrow.

Figure 5A,B shows responses for replicates of three differently coupled receptors measured at early (30 s) and late (2.5 min) time points, respectively, after ligand addition in the context of CHO_m1 cells. At the early time point, the Gq-coupled muscarinic m₁ response is visible in the lower left hand quadrant of the dot plot, indicating the decrease in CDS signal (Fig. 5A). As time progresses, the Gq-GPCR response moves to the upper right-hand quadrant, along with the Gi-GPCR response, indicating an increase in CDS signal (Fig. 5B). In contrast, the Gs-GPCR mediated response (prostanoid EP4) decreases over time and stays in the lower left-hand quadrant (Fig. 5A, B). Consequently, analysis of the CDS response profiles at a later time point after receptor stimulation distinguishes the Gs-GPCR response from the Gi- and Gq-coupled responses, while the Gi- and Gq-coupled responses are primarily distinguished by their differential response at an early time point.

Figure 5.

CDS responses plotted in dot-plot format based on the contribution of components Z_Iec and Z_Itc after stimulation of CHO_m1 cells with ligands. The points on this plot represent the changes in Z_Iec and Z_Itc induced by each ligand at (A) 30 s and (B) 2.5 min post-ligand addition. Gray dots indicate negative controls that clustered near the origin and responses to the other compounds plotted in different regions of the graph distinct from the origin, with their distances corresponding to response potency. Black, red, and blue circles indicate the CDS responses to 5HT, carbachol, and PGE2, respectively; n = 8 replicates for each ligand.

These data suggest that comparing the CDS response profile, or dot plot localization of a receptor with unknown G-protein coupling, to the CDS responses of receptors with known G-protein coupling in a specific cell type will allow identification of the signaling pathway of the unknown GPCR.

Endogenous receptors in primary human cells. As primary cells are not amenable to transfection, the ability to measure endogenous receptor responses within primary isolates provides a clear advantage for the use of CDS analysis. Previous studies have demonstrated that stimulation of microvascular endothelial cells with interleukin (IL)-8 leads to cytoskeletal reorganization mediated by combined activation of the CXCR1 and the CXCR2 chemokine receptors.³⁶ To demonstrate that CDS can detect activation of endogenous receptors in primary cells, human neutrophils were treated with IL-8 to activate the CXCR1 and CXCR2 receptors.⁴⁹ EC⁵⁰ values measuring the functional effect of IL-8 were generated in minutes using CDS (Fig. 6A) and were in agreement with EC50 values obtained using chemotaxis assays that typically require 2–4 h to assay and process the data.^50,51

Figure 6.

Cellular responses to activation of chemokine receptors as measured by CDS. (A) Concentration response curve for IL-8, a CXCR agonist, on primary human neutrophils. The CDS-derived EC₅₀ for IL-8 was 6.2 × 10⁻¹⁰ M. CDS data were buffer-subtracted and normalized to 10 nM IL-8. Each data point represents mean ±SD, n = 4. Experiments were performed at 28 °C. (B) Concentration response curve for GROα, a CXCR2-selective agonist on differentiated U937 cells. The CDS derived EC₅₀ value for GROα was 1.2 × 10⁻⁹ M. CDS data were normalized to 4 nM GROα. Each data point represents mean ±SD, n = 3. Experiments were performed at 28 °C. (C) Concentration response curve for SB 225002, an antagonist to the CXCR-2 receptor. Differentiated U-937 cells were preincubated with SB 225002 for 30 min at 28 °C before stimulating with 4 nM GROα. The CDS derived IC₅₀ for SB 225002 was 9.2 × 10⁻⁸ M. CDS responses were normalized to 4 nM GROα. Each data point represents mean ±SD, n 3.

Robustness and reproducibility. The magnitude of the CDS response is dependent primarily upon how tightly the receptor is coupled to the signaling mechanism upon ligand stimulation, and less dependent upon receptor expression levels. The magnitudes of transfected receptor responses (dopaminergic D1 and D2L, and muscarinic m₁) are not significantly different from those pertaining to the endogenous receptors (serotonin 5HT_1B, calcitonin, and prostanoid EP4) shown in Figure 3. Expression of the 5HT_1B receptor in CHO cells is too low to be detected by radioactive ligand binding assays.⁵² However, functional assays such as intracellular cAMP and CDS can be used to detect activation of this low-density receptor that is very efficiently coupled to its effector.⁵² The sensitivity of CDS to receptor-mediated cytoskeletal and noncytoskeletal changes enables robust and reliable measurement of endogenous receptors.

Activation of the endogenous CXCR2 receptors in U937 cells,⁵³ which is typically measured with chemotaxis assays and not detectable using second messenger assays, is easily and quickly detected with CDS. DMSO-differentiated U937 cells yield a robust concentration-dependent response measured by CDS upon addition of GROα, a CXCR2-selective ligand⁵⁴ (Fig. 6B). The response to GROα is inhibited in cells preincubated with SB 225002, a selective antagonist for the CXCR2 receptor,⁵⁵ demonstrating that the CDS measurement is receptor-specific (Fig. 6C). The CDS response is robust and sensitive as cells treated with an EC80 dose (4 nM) of GROα yielded a response with an interplate % of coefficient of variation (%CV) of 6.7% and a Z'-score of 0.71 (n = 30 wells across 3 plates.)

Pharmacological Profiling of Agonists and Antagonists

The data presented thus far demonstrate that CDS is a powerful tool for detecting and characterizing activation of endogenous receptors. While the unique nature of the CDS information enables novel applications, it is the universality and ease of use of the platform that provides a powerful tool for generating pharmacologically relevant data with little assay development required.

Potency ranking. CDS generates pharmacological data for adherent and nonadherent cells that are comparable to data generated from assays designed to measure changes in intracellular second messenger levels. For example, U937 cells and CHO_m1 cells were treated with varying concentrations of agonists targeting the endogenous purinergic P2Y receptors and antagonists targeting the transfected muscarinic m₁ receptor, respectively. EC₅₀ and IC₅₀ values were determined using both CDS and calcium mobilization, and concentration response curves are shown in Figure 7. These results demonstrate that CDS generates the same potency rankings as other cell-based assays without the need for chemical modification of the cells.

Figure 7.

Comparison of EC₅₀ and IC₅₀ values derived from CDS and calcium assays. Concentration response curves were produced for a series of (A) agonists targeting the endogenous purinergic receptor in U-937 cells at 28 °C, and (B) antagonists targeting the transfected muscarinic mi receptor in CHO_m1 cells. Cells were pretreated with antagonists for 20 min at 28 °C prior to stimulation with 10 μM carbachol. These ligands were assessed using both CDS platform and Ca²⁺ (fluo-4, AM) fluorescence on the Molecular Devices Analyst. The two assays demonstrate comparable potency rankings with good correlation.

Selectivity analysis. CDS is a universal approach that allows measurement of Gi- and Gs-coupled GPCRs with the same assay. To demonstrate the use of CDS for receptor subtype selectivity analysis within a single assay in the context of Gi- and Gs-GPCRs, CHO cells stably transfected individually with four different dopamine receptor subtypes were treated with selective and nonselective ligands. CDS measurements demonstrated that quinpirole (D2L-selective agonist), but not dopamine (nonselective ligand), selectively activated cells expressing the D2L receptor (Gi-GPCR) compared to the D4.4 (Gi-GPCR), D1 (Gs-GPCR), or D5 (Gs-GPCR) receptors (Fig. 8A,B). One advantage of using CDS is that it is a single assay platform that enables pharmacologically relevant selectivity studies, which do not require comparison across different assays or platforms for differently coupled receptors.

Figure 8.

Concentration response curves of ligands tested across four cell lines at 37 °C. CHO cells expressing the D1 (Gs), D2L (Gi), D4.4 (Gi), or D5 (Gs) receptors were treated with (A) varying concentrations of quinpirole, a D2L selective agonist, and (B) dopamine, a nonselective agonist. CDS responses were normalized to 10 μM quinpirole or 10 μM dopamine on the D2L cells. Data represent mean ±SD, n = 3. (C) Schild analysis with dopamine agonist and 2-chloro-l 1-(4-methylpiperazino)dibenz[b,f]oxepin (2-CMP-D) antagonist on CHO-D1 (Gs) cells. Cells were pretreated with different concentrations of antagonist at 37 °C for 20 min prior to stimulation with varying concentrations of dopamine. As concentration of antagonist was increased, a rightward shift in the EC₅₀ of the agonist was observed. CDS responses were normalized to 10 μM dopamine. Data represent mean ±SD, n = 3. (D) From the relationship between the shifted EC₅₀ values and the concentration of antagonist, the pA₂ value was determined to be −8.25.

Schild analysis. In addition to ranking compounds and performing selectivity studies, CDS can also be used to generate pA2 values via Schild analysis. CHO cells transfected with the Gs-coupled D1 dopamine receptor were pretreated with various concentrations of the antagonist, 2-chloro-11-(4-methylpiperazino)dibenz[b,f]oxepin (2-CMP-D), prior to the generation of dopamine concentration response curves (Fig. 8C). From the relationship between the shifted EC₅₀ values and the concentration of antagonist, the pA₂ value for 2-CMP-D was determined to be −8.25 (Fig. 8D).

Materials and Methods

Cell Lines

Transfected CHO cells expressing various dopamine receptor subtypes⁵⁶ were kindly provided by MDS Pharma Services. Cell lines CHO_m1 (M1WT3; CRL-1985), HeLa (CCL-2), and U937 (CRL-1593.2)⁵³ were obtained from ATCC (Manassas, VA). Primary neutrophils were purchased from AllCells, LLC (Berkeley, CA). HeLa cells were grown in minimal essential medium (MEM); U937 cells were grown in RPMI-1640; CHO_m1 cells were grown in Ham's F12K; CHO cells expressing transfected dopamine receptor subtypes were grown in Ham's F12. All media were supplemented with 10% fetal bovine serum (FBS), 50 μg/mL streptomycin, and 50 units/mL penicillin. CHO cells expressing transfected receptors (CHO-D1, D2L, D4.4, D5 and CHO_m1) were maintained under drug selection with 100 μg/mL G418. U937 cells were differentiated by exposure to culture media containing 1.5% DMSO for 96 h.

Reagents

The following were purchased from Sigma (St. Louis, MO): bovine serum albumin (BSA), endothelin-1 (ET-1), isoproterenol, stromal cell-derived factor 1 alpha (SDF-1), epidermal growth factor (EGF), insulin-like growth factor (IGF), 5-hydroxytryptamine (5HT), calcitonin, forskolin, uridine triphosphate (UTP), uridine diphosphate (UDP), adenosine diphosphate (ADP), adenosine triphosphate (ATP), ATP-γ-S, prostaglandin E₂ (PGE2), bradykinin, UK 14,304, carbachol, atropine, ipratropium Br, trihexyphenidyl, telenzepine, clozapine, p-fluoro-hexahydro-sila-difenidol (p-FHHSiD), histamine, thapsigargin, bombesin, melatonin, β-estradiol, NECA, neuropeptide Y, thrombin, B-cell attracting chemokine 1 (BCA-1), r-phenylephrine, sphingosine-1-phosphate, platelet activating factor (PAF), MIP-1α, 1-oleoyl-LPA, MIP-3β, retinoic acid, vasopressin, apelin 13, and platelet derived growth factor BB (PDGF BB). Somatostatin, ADP-β-S,3 Li and SB 225002 were purchased from Calbiochem (San Diego, CA); and GROα, IL-8, IL-3, MCP-3, hepatocyte growth factor (HGF), keratinocyte growth factor (KGF), and tumor necrosis factor α (TNF-α) were purchased from R&D Systems (Minneapolis, MN). A76636, quinpirole, and 2-Chloro-11-(4-methylpiperazino)dibenz[b,f]oxepin were obtained from Biomol (Plymouth Meeting, PA).

Hanks balanced salt solution and HEPES buffer solution were purchased from Invitrogen (Grand Island, NY). FBS was purchased from Invitrogen (Grand Island, NY) and Gemini Bio-Products (Woodland, CA). Ham's F12, Ham's F12K, MEM, RPMI-1640, penicillin, streptomycin, and G418 were purchased from Cellgro (Herndon, VA).

Instrument Design

The data in this study were generated on an engineering breadboard version of the CellKey™ system. The principal component of the CDS breadboard system used for these studies included a custom 96-well integrated with a measurement system connected to Agilent (Palo Alto, CA) impedance analyzers. Constant voltage over a range of 30 frequencies, from 1 kHz to 110 MHz, was applied to a monolayer of cells, and the resultant current was measured at an update rate of 20 s. The system was thermally regulated and experiments were run between ambient and 37°C. Onboard fluid additions and exchanges were handled by a 96-head fluid delivery device with a range of 5–200 μL. In addition to the cell plate location, the instrument included two additional stations for 96-well compound plates. All aspects of assay set up, execution, and data acquisition and analysis were controlled by custom applications software.

General Cell Protocols

Adherent cells were plated into the wells of CDS micro-plates in standard tissue culture media containing serum, and the plates were incubated overnight under standard conditions (37°C, 95% humidity, 5% CO₂). During the overnight incubation, the cells adhered to, and formed a monolayer over, the electrodes at the bottom of the custom 96-well microplate, thus becoming part of an electric circuit. Typically, 3.5–4.5 × 10⁴ cells/well were used to form the monolayer, depending on the cell type. After the plating period, the plates were placed on the instrument, and the cell monolayers were washed two times with Hanks balanced salt solution containing 20 mM HEPES and 0.1% BSA (HH). Cells were then allowed to equilibrate at the assay temperature for 15 min. At the end of this period, readings were taken every 20 s for 5 min to establish a baseline. At the end of the baseline period, ligands were added, and the change in the magnitude of the measured impedance (ΔZ) was recorded over time. Typically, 10 min of post-ligand addition data were acquired for any experiment.

For experiments with nonadherent cells, the following protocol was utilized. Cells were harvested from stock tissue culture flasks and washed by centrifugation three times with HH. Typically, 1.0–1.2 × 10⁵ cells were then added to each well and allowed to settle for 30 min at room temperature. This resulted in a single layer of cells on top of the electrodes. Though the nonadherent cell monolayers were not attached to the electrodes, their proximity to the electrodes was sufficient for impedance measurements by CDS. After the cells settled, the plate was placed on the instrument and receptor activation experiments were initiated in the manner described above for adherent cell assays.

Agonist and Antagonist Protocols

At the completion of the wash steps described above, 135 μL of HH were dispensed to the wells of the cell plate. For receptor agonist studies, 15 μL of a 10x concentration of agonist were added to the 135 μL and mixed three times using the onboard fluidics head. For nonadherent cells, the pipetting speed and pipette tip height of the fluidics head were adjusted so as not to disturb the monolayer of settled cells. Readings were taken continuously during the ligand addition procedure and for an additional 10 min post-ligand addition. In the studies reported in this paper, experiments were conducted at either 28°C or 37°C.

Antagonist studies required one more fluid addition, as described here. For adherent cells, the final 135 μL wash volume was removed from the wells and replaced with 135 μl HH containing a 1x concentration of antagonist. Cells were allowed to incubate in the presence of antagonist for various times (see figure legends). At the end of the pretreatment period, 15 μL of a 10x concentration of agonist was added as described in the agonist protocol above.

Calcium Assays

Calcium fluorescence data were generated on the Molecular Devices Analyst using fluo-4, AM.⁵⁷

Summary

CDS is a technology that enables comprehensive pharmacological evaluation of cell surface receptors in real time. It is a universal assay that allows measurement of multiple types of receptors, including G protein-coupled receptors (GPCRs), tyrosine kinase receptors (TKRs), and steroid receptors using the same platform, without the need for transfections or special reagents. The CDS-based platform is sensitive enough to consistently monitor ligand-mediated activation of endogenously expressed receptors, and for this reason, has the ability to generate more physiologically relevant data than obtained with genetically and chemically manipulated cells.

The label-free CDS measurement is a measure of cell function as opposed to binding events that are typical of many current label-free approaches: classical (BIAcore's SPR technology) and more recent (Corning's EPIC, SRU Biosystems' BIND, CSEM's WIOS, and Akubio's RAP technology). A detailed comparison of these and other impedance-based measurement systems is available in a recent review of label-free technologies.⁵⁸

To illustrate the power of the CDS technology, data has been presented demonstrating the use of this label-free approach in pharmacological evaluation of ligands for a number of receptors in a variety of cell types, including primary cells. Specifically, the CDS technology has been applied to hit confirmation, receptor selectivity analysis, ligand potency, and Schild analysis of receptor-selective antagonists. CDS measurements quantitatively align with results from other cell-based assays in determining the potency and ranking of agonists and antagonists. The robust and reproducible nature of CDS assays is reflected in low coefficients of variation, high signal-to-background, and excellent Z'-scores of the data. The data generated using CDS technology are unique in that the response profiles reflect receptor-mediated signaling pathways and are characteristic of main subsets of GPCRs within a cell line. Clustering analysis has been performed using response profiles from receptors that lead to potential novel applications including identification of the G-protein coupling of unknown GPCRs and the cataloging of active endogenous receptors in cell lines. Because the CDS assay requires no stable or transiently transfected cells or special reagents, assay development and data acquisition are simple and fast. These studies demonstrate that the CDS platform provides an easy-to-use, label-free, universal assay with applications in secondary screening, hit identification and lead optimization areas of drug discovery. The CellKey™ system, based on CDS technology, will be commercially available in late 2005.²⁹

References

Foster

Schwan

Dielectric Properties of Tissues and Biological Materials: A Critical Review. Crit. Rev. Biomed. Eng. 1989, 17, 25–104.

Gheorghiu

Characterizing Cellular Systems by Means of Dielectric Spectroscopy. Bioelectromagnetics 1996, 17, 475–482.

Gheorghiu

Measuring Living Cells Using Dielectric Spectroscopy. Bioelectrochem. Bioenerg. 1996, 40, 133–139.

Schwan

Mechanisms Responsible for Electrical Properties of Tissues and Cell Suspensions. Med. Prog. Technol. 1993, 19, 163–165.

Asami

Yonezawa

Dielectric Behavior of Wild-Type Yeast and Vacuole-Deficient Mutant Over a Frequency Range of 10 KHz to 10 Ghz. Biophys. J. 1996, 71, 2192–2200.

Bordi

Cametti

Gili

Reduction of the Contribution of Electrode Polarization Effects in the Radiowave Dielectric Measurements of Highly Conductive Biological Cell Suspensions. Bioelectrochemistry 2001, 54, 53–61.

Bordi

Cametti

Gili

Dielectric Spectroscopy of Erythrocyte Cell Suspensions: A Comparison Between Looyenga and Mawell- Wagner-Hanai Effective Medium Theory Formulations. J. Non-Cryst. Solids 2002, 305, 278–284.

Coster

Chilcott

Coster

Impedance Spectroscopy of Interfaces, Membranes and Ultrastructures. Bioelectrochem. Bioenerg. 1996, 40, 79–98.

Smith

Duffy

Shen

Oliff

Dielectric Relaxation Spectroscopy and Some Applications in the Pharmaceutical Sciences. J. Pharm. Sci. 1995, 84, 1029–1044.

10.

Wegener

Sieber

Galla

Impedance Analysis of Epithelial and Endothelial Cell Monolayers Cultured on Gold Surfaces. J. Biochem. Biophys. Methods 1996, 32, 151–170.

11.

Paunescu

T. G.

Helman

S. I.

PGE2 Activation of Apical Membrane Cl- Channels in A6 Epithelia: Impedance Analysis. Biophys. J. 2001, 81, 852–866.

12.

Paunescu

T. G.

Helman

S. I.

Camp Activation of Apical Membrane Cl- Channels: Theoretical Considerations for Impedance Analysis. Biophys. J. 2001, 81, 838–851.

13.

Awayda

M. S.

Van Driessche

Helman

S. I.

Frequency-Dependent Capacitance of the Apical Membrane of Frog Skin: Dielectric Relaxation Processes. Biophys. J. 1999, 76, 219–232.

14.

Devor

D. C.

Singh

A. K.

Lambert

L. C.

DeLuca

Frizzell

R. A.

Bridges

R. J.

Bicarbonate and Chloride Secretion in Calu-3 Human Airway Epithelial Cells. J. Gen. Physiol 1999, 113, 743–760.

15.

Tamada

Hug

Frizzell

Bridges

Microelectrode and impedance analysis of anion secretion in Calu-3 cells. JOP 2001, 2, 219–228.

16.

Moy

A. B.

Van Engelenhoven

Bodmer

Kamath

Keese

Giaever

Shasby

D. M.

Histamine and Thrombin Modulate Endothelial Focal Adhesion Through Centripetal and Centrifugal Forces. J. Clin. Invest 1996, 97, 1020–1027.

17.

Moy

A. B.

Winter

Kamath

Blackwell

Reyes

Giaever

Keese

Shasby

D. M.

Histamine alters endothelial barrier function at cell-cell and cell-matrix sites. Am J Physiol Lung Cell Mol Physiol 2000, 278, 888–898.

18.

Sandoval

Malik

A. B.

Minshall

R. D.

Kouklis

Ellis

C. A.

Tiruppathi

Ca2+ signalling and PK{α} activate increased endothelial permeability by disassembly of VE-cadherin junctions. J. Physiol 2001, 533, 433–445.

19.

Mehta

Tiruppathi

Sandoval

Minshall

R. D.

Holinstat

Malik

A. B.

Modulatory role of focal adhesion kinase in regulating human pulmonary arterial endothelial barrier function. J. Physiol 2002, 539, 779–789.

20.

Ehret

Baumann

Brischwein

Schwinde

Wolf

Online control of cellular adhesion with impedance measurements using interdigitated electrode structures. Med Biol Eng Comput 1998, 36, 365–370.

21.

Ehret

Baumann

Brischwein

Schwinde

Stegbauer

Wolf

Monitoring of cellular behaviour by impedance measurements on interdigitated electrode structures. Biosens Bioelectron 1997, 12, 29–41.

22.

Asami

Irimajiri

Dielectrospectroscopic Monitoring of Early Embryogenesis in Single Frog Embryos. Phys. Med. Biol. 2000, 45, 3285–3297.

23.

Luong

Habibi-Rezaei

Meghrous

Xiao

Male

Kamen

Monitoring motility, spreading, and mortality of adherent insect cells using an impedance sensor. Anal Chem 2001, 73, 1844–1848.

24.

Wegener

Keese

Giaever

Electric Cell-Substrate Impedance Sensing (ECIS) as a Noninvasive Means To Monitor the Kinetics of Cell Spreading to Artificial Surfaces. Exp. Cell Res. 2000, 259, 158–166.

25.

Giaever

Keese

Micromotion of Mammalian Cells Measured Electrically. PNAS 1991, 88, 7896–7900.

26.

Santini

Cametti

Indovina

Peterson

Membrane electrical properties associated with insulin receptor downregulation in human erythrocytes. Exp Hematol 1994, 22, 40–44.

27.

Steinem

Janshoff

Wegener

Ulrich

Willenbrink

Sieber

Galla

Impedance and shear wave resonance analysis of ligand-receptor interactions at functionalized surfaces and of cell monolayers. Biosens Bioelectron 1997, 12, 787–808.

28.

Wegener

Zink

Rosen

Galla

Use of electrochemical impedance measurements to monitor beta-adrenergic stimulation of bovine aortic endothelial cells. Pflugers Arch 1999, 437, 925–934.

29.

MDS Sciex website. www.mdssciex.com. Last accessed July 20, 2005.

30.

Dittmar

Husemann

Schewe

Nofer

J. -R.

Niggemann

Zanker

K. S.

Brandt

B. H.

Induction of cancer cell migration by epidermal growth factor is initiated by specific phosphorylation of tyrosine 1248 of c-erbB-2 receptor via epidermal growth factor receptor. FASEB J 2002, 16, 1823–1825.

31.

Kim

E. S.

Bearer

E. L.

Arp2/3 Complex Is Required for Actin Polymerization During Platelet Shape Change. Blood 2002, 99, 4466–4474.

32.

Vanhauwe

J. F.

Thomas

T. O.

Minshall

R. D.

Tiruppathi

Gilchrist

Yoon

E. J.

Malik

A. B.

Hamm

H. E.

Thrombin Receptors Activate Go Proteins in Endothelial Cells to Regulate Intracellular Calcium and Cell Shape Changes. J. Biol. Chem 2002, 277, 34143–34149.

33.

Kawanabe

Okamoto

Nozaki

Hashimoto

Miwa

Masaki

Molecular Mechanism for Endothelin-1-induced Stress-Fiber Formation: Analysis of G Proteins Using a Mutant EndothelinA Receptor. Mol. Pharmacol 2002, 61, 277–284.

34.

Lesh

R. E.

Emala

C. W.

Lee

H. T.

Zhu

Panettieri

R. A.

Hirshman

C. A.

Inhibition of geranylgeranylation blocks agoinst-induced actin reorganization in human airway smooth muscle cells. Am J Physiol Lung Cell Mol Physiol 2001, 281, L824–L831.

35.

Pierce

K. L.

Fujino

Srinivasan

Regan

J. W.

Activation of FP Prostanoid Receptor Isoforms Leads to Rho-mediated Changes in Cell Morphology and in the cell Cytoskeleton. J. Biol. Chem 1999, 274, 35944–35949.

36.

Schraufstatter

I. U.

Chung

Burger

IL-8 Activates Endothelial Cell CXCR1 and CXCR2 Through Rho and Rac Signaling Pathways. Am J. Physiol. (London) Lung Cell Mol. Physiol. 2001, 280, L1094–L1103.

37.

Saltarelli

Heterotrimetric Gi/O Proteins Control Cyclic AMP Oscillations and Cytoskeletal Structure Assembly in Primary Human Granulosa-Lutein Cells. Cell. Signal. 1999, 11, 415–433.

38.

Maddala

Reddy

Epstein

Rao

Growth factor induced activation of Rho and Rac GTPases and actin cytoskeletal reorganization in human lens epithelial cells. Mol Vis 2003, 9, 329–336.

39.

Lorenz

DesMarais

Macaluso

Singer

Condeelis

Measurement of barbed ends, actin polymeriztion, and motility in live carcinoma cells after growth factor stimulation. Cell Motil Cytoskeleton 2004, 57, 207–217.

40.

Suzuki

Nakamura

Takahashi

Ikuhara

Matsuzaki

Isogai

Hori

Suda

Calcitonin-induced changes in the cytoskeleton are mediated by a signal pathway associated with protein kinase A in osteoclasts. Endocrinology 1996, 137, 4685–4690.

41.

Martineau

L. C.

McVeigh

L. I.

Jasmin

B. J.

Kennedy

C. R. J.

P38 MAP kinase mediates mechanically induced COX-2 and PG EP4 receptor expression in podocytes: Implications for the actin cytoskeleton. Am J Physiol Renal Physiol 2004, 286, 693–701.

42.

Perez

Bouschet

Fernandez

Bockaert

Journot

Dynamic reorganization of the astrocyte actin cytoskeleton elicited by cAMP and PACAP: A role for phosphatidyllnositol 3-kinase inhibition. Eur J Neurosci 2005, 21, 26–32.

43.

Hirshman

C. A.

Zhu

Panettieri

R. A.

Emala

C. W.

Actin depolymerization via the beta-adrenoceptor in airway smooth muscle cells: A novel PKA-independent pathway. Am J Physiol Cell Physiol 2001, 281, C1468–C1476.

44.

Szaszi

Kurashima

Kaibuchi

Grinstein

Orlowski

Role of the Cytoskeleton in Mediating cAMP-dependent Protein Kinase Inhibition of the Epithelial Na+/H+ Exchanger NHE3. J. Biol. Chem 2001, 276, 40761–40768.

45.

Kreisberg

J. I.

Ghosh-Choudhury

Radnik

R. A.

Schwartz

M. A.

Role of Rho and myosin phosphorylation in actin stress fiber assembly in mesangial cells. Am J Physiol Renal Physiol 1997, 273, F283–F288.

46.

Lamb

Fernandez

Conti

Adelstein

Glass

Welch

Feramisco

Regulation of actin microfilament integrity in living nonmuscle cells by the cAMP-dependent protein kinase and the myosin light chain kinase. J. Cell Biol 1988, 106, 1955–1971.

47.

Patterson

Lum

Schaphorst

Verin

Garcia

Regulation of endothelial barrier function by the cAMP-dependent protein kinase. Endothelium 2000, 7, 287–308.

48.

Gujdar

Sipeki

Bander

Buday

Farago

Phorbol ester-induced migration of HepG2 cells is accompanied by intensive stress fibre formation, enhanced integrin expression and transient downregulation of p21-activated kinase 1. Cell Signal 2003, 15, 307–318.

49.

Chuntharapai

Lee

Hebert

Kim

Monoclonal antibodies detect different distribution patterns of IL-8 receptor A and IL-8 receptor B on human peripheral blood leukocytes. J Immunol 1994, 153, 5682–5688.

50.

McGee

Shah

Watson

Parsons

Jepras

Wilson

M. J.

Reith

Development of a Robust, Automated System To Perform Chemotaxis Assays Using Chemotx Migration Chambers (TP051). LabAutomation2005, San Jose, CA, 2005.

51.

Ludwig

Petersen

Zahn

Gotze

Schroder

J. -M.

Flad

H. -D.

Brandt

The CXC-Chemokine Neutrophil-Activating Peptide-2 Induces Two Distinct Optima of Neutrophil Chemotaxis by Differential Interaction With Interleukin-8 Receptors CXCR-1 and CXCR-2. Blood 1997, 90, 4588–4597.

52.

Giles

Landell

Bolofo

Wilson

Martin

Characterization of a 5-HT1B receptor on CHO cells: Functional responses in the absence of radioligand binding. Br. J. Pharmacol 1996, 117, 1119–1126.

53.

Sundstrom

Nilsson

Establishment and Characterization of a Human Histiocytic Lymphoma Cell Line (U-937). Int. J. Cancer 1976, 17, 565–577.

54.

Hebert

Vitangcol

Baker

Scanning Mutagenesis of Interleukin-8 Identifies a Cluster of Residues Required for Receptor Binding. J. Biol. Chem. 1991, 266, 18989–18994.

55.

White

J. R.

Lee

J. M.

Young

P. R.

Hertzberg

R. P.

Jurewicz

A. J.

Chaikin

M. A.

Widdowson

Foley

J. J.

Martin

L. D.

Griswold

D. E.

Sarau

H. M.

Identification of a Potent, Selective Non-peptide CXCR2 Antagonist That Inhibits Interleukin-8-induced Neutrophil Migration. J. Biol. Chem 1998, 273, 10095–10098.

56.

Hill

Lai

Hnilo

Lin

Karla

Bounds

Herz

Mitchell

Cloning, expression, and comparison of the binding characteristics of the known human dopamine receptors. Adv Neurol 1996, 69, 41–52.

57.

Ciambrone

G. J.

Liu

V. F.

Lin

D. C.

McGuinness

R. P.

Leung

G. K.

Pitchford

Cellular Dielectric Spectroscopy: A Powerful New Approach to Label-Free Cellular Analysis. J Biomol Screen 2004, 9, 467–480.

58.

Comley

Label-Free Detection. Drug Discovery World 2005, 6, 63–74.

Cellular Dielectric Spectroscopy: A Label-Free Technology for Drug Discovery

Abstract

Keywords

Introduction

Results and Discussion

Enabling Comprehensive Measurement of Endogenous Receptors

Pharmacological Profiling of Agonists and Antagonists

Materials and Methods

Cell Lines

Reagents

Instrument Design

General Cell Protocols

Agonist and Antagonist Protocols

Calcium Assays

Summary

References