1. technology and computing

Urs Lüthi *, Celia Müller, Thomas Sasse and John Gatfield *
Actelion Pharmaceuticals Ltd., Gewerbestrasse 16, CH-4123 Allschwil, Switzerland
Impedance-based Secondary Screening of OX1 Receptor
Antagonists using the 384-format RTCA HT Instrument
Impedance-based technologies have established themselves as powerful tools in cell biology. Impedance allows a label-free and sensitive kinetic readout of changes in
cell morphology induced by a multitude of different stimuli, thus it is used in many different cell biological contexts. In drug discovery, impedance technologies are
entering the High-Throughput (HTS) domain, especially in the field of GPCR (G-protein-coupled receptor) signaling assays this technology is regarded as a valuable
complementation of already existing classical GPCR assays.
In this study, the recently developed 384-well impedance device RTCA HT (Roche Applied Science) was used as a secondary screening tool in a HTS for orexin type 1
(Ox1) receptor antagonists. Primary hits obtained by the classical calcium flux (FLIPR) technology using CHO-Ox1 cells were validated in newly developed fully-automated
RTCA HT assays using the same cells. Impedance data were analyzed for antagonistic activities of the compounds, specificity of activity as well as robustness of these
results (intra-assay and inter-assay variability, Z’). Finally, the antagonist hit confirmation rate compared to FLIPR data was determined. In conclusion, we find that with the
novel 384-well RTCA HT technology assay robustness is comparable to classical GPCR assay technologies, hit confirmation rates were between 60-70 % with most of the
non-confirmed hits being non-specific blockers of calcium signaling.
Introduction
Results
Impedance technology allows a label-free kinetic readout of cellular morphological changes and is ideally suited to measure GPCR activation or
inhibition independent of the coupling pathway in recombinant or non-recombinant cellular systems. As shown in Figure 1A all major G-protein
pathways, i.e. Gaq, Gas, Gai and Ga12/13, lead to changes in cytoskeletal organization. Such changes in cellular morphology alter the electrical
impedance of an adherent monolayer and can be measured when cells are grown on electrode surfaces. The newly developed 384-well format
RTCA HT Instrument (Roche Applied Science) allows the analysis of such cellular responses to compounds in an HTS mode.
A
B
Hit distribution of RTCA secondary screen
40%
-
RTCA vs FLIPR in CHO-Ox1 cells
RTCA Ox1 (% inhibition)
130%
200
In this study we describe the use of the Roche RTCA HT instrument in a secondary screening within a HTS campaign searching for antagonists of
the orexin type 1 receptor (Ox1). As seen in Figure 1B, 71063 compounds were subjected to HTS using calcium flux assays (FLIPR) with CHO-Ox1
cells. 324 compounds thereof could be confirmed in the same assay. These 324 Ox1 antagonist hits were first subjected to FLIPR specificity assays
using unrelated CHO-S1P3 cells, and then analyzed in 2 fully-automated RTCA HT screenings on CHO-Ox1 and CHO-S1P3 cells, respectively.
= confirmed in RTCA
150
100
50
A
B
Ligand
71063 compounds
FLIPR
0
CHO-Ox1
-50
HTS at 10m M
>50% inhibition
324 confirmed hits
Ga
Ga s
Ga i
50
60
70
80
Ga q
CHO-Ox1
GTP
counter screen: 324 hits tested
Gbg
FLIPR
C
D
Hits not confirmed in RTCA show unspecific FLIPR antagonism
CHO-S1P3
Ga 12/13
Hit confirmation by RTCA
200
FLIPR
324
150
Changes in cell morphology
RTCA HT
RTCA HT
CHO-Ox1
CHO-S1P3
100
active in S1P3
FLIPR>30%
inhibition
=
209 (=65%)
RTCA
209
50
Figure 1. Impedance as universal readout of GPCR signaling (A), and its use as a secondary screening assay in the screening cascade of an Ox1
receptor antagonist HTS (B).
59 out of 115 are
unspecific antagonists in FLIPR
(= S1P3 actives)
-50
Methods
xCELLigence Technology
B
transcellular current
intercellular current
(cell capacitance)
(cell resistance)
C
D
Impedance (Z)
115
0
-100
50
A
100
RTCA Ox1 (% inhibition)
Impedance 2nd screen: 324 hits tested
Impedance
90
FLIPR Ox1 (% inhibition)
FLIPR
GDP
60
70
80
90
100
FLIPR Ox1 (% inhibition)
Figure 5. Analysis of RTCA antagonist data and correlation with FLIPR data. A Hit distribution of RTCA screen (% inhibition of orexinA signal) and cut-off
criteria for hit confirmation (40% - 130% inhibition). B Comparison of activities in the Ox1 FLIPR assay versus Ox1 RTCA assay. 209 out of 324 compounds (in
dark green) fullfill RTCA cut-off criteria as hits (65% confirmation rate). C The 115 non-confirmed hits are colored in dark green if active as antagonists in the
CHO-S1P3 FLIPR assay. D Venn diagram showing that 209 out of 324 FLIPR hits are confirmed by RTCA. Of the 115 non-confirmed hits 59 compounds (51%)
are false positive FLIPR hits (they also block calcium flux in CHO-S1P3 cells).
65% of FLIPR hits originating from an internal screening campaign was confirmed by RTCA (209 out of 324 FLIPR hits). Importantly, 51% of the nonconfirmed hits were found to be unspecific in the FLIPR assay. Therefore RTCA is a powerful secondary screening tool and specificity filter.
Rcells
Quality Metrics
Rmedium
Ccells
A Quality metrics of Screen 1
AC (kHz)
Figure 2. The basics of impedance. A Cells grown on interdigitated electrode surfaces impose an impedance to current flow as resistance (intercellular flow) or
capacitance (transcellular flow). B A cell culture well as a condensor: The current flow through the cell culture is impeded by the resistance of the medium
(Rmedium) and the combined resistance and capacitance of the cell layer (Rcells, Ccells) C Magnification of a RTCA electrode in a 96 well plate. D Gold
electrodes of a 384-well plate.
Integration of RTCA HT Stations into the Agilent Automation System
RTCA HT
plate shuf fling
B EC50 curve of orexinA
0.28
1.36
4.4
2.9
0.35
1.45
3.7
4.1
0.57
1.43
2.8
2.2
0.50
1.46
3.6
2.2
0.56
1.48
2.5
3.4
0.48
1.52
2.9
4.6
0.41
1.42
2.4
4.7
0.54
1.38
2.1
2.9
0.48 +/- 0.11
1.44 +/- 0.05
2.9 +/- 0.8
3.4 +/- 1.0
RTCA Screen 1
EC50 values orexinA across plates
E FLIPR
vs RTCA HT
EC50
orexinA
[nM]
IC50
ref. antagonist
[nm]
IC50 [nM]
C IC50 curve of ref. antagonist
Cytomat 6001
10.0
FLIPR
6.7
3.4
RTCA HT
9.0
13.4
Biotek cell washer
0.1
1
2
F Intra-assay duplicates
Measure 25min
in 30s intervals
C RTCA HT Stations on Automation Platform
 wash cells 3x
 incubate 50min
Correlation coefficient 0.96
G Inter-assay duplicates
4
5
6
7
1
8
4
5
6
7
8
RTCA Screen 2
EC50 values orexinA across plates
IC50 values ref. antagonist across plates
IC50 ref. antagonist screen 2
13.3nM +/- 3.2nM
EC50 orexinA screen 2
9.5nM +/- 4.0nM
100
EC50 [nM]
% effect screening 2
 measure impedance
Compound plates
(1mL 4mM)
10.0
1.0
1
% effect screening 1
10
1
0.1
0.1
Bridge Server
% effect duplicate 1
3
plate no
100.0
 cell index normalization
2
plate no
Correlation coefficient 0.86
% effect duplicate 2
 add medium
Data processing
3
IC50 [nM]
1st addition
compounds
10
1
1.0
0.1
overnight
measurements
Measure 3min
in 30s intervals
IC50 ref. antagonist screen 1
13.4nM +/- 1.7nM
100
 add cells (4000/well)
2nd addition
orexinA
IC50 values ref. antagonist across plates
EC50 orexinA screen 1
8.6nM +/- 2.3nM
 add growth medium
 measure background
D Intra-assay and inter-assay reproducibility of EC50 and IC50 values
CV buffer
100.0
B RTCA HT Station
E-plates (384 wells)
CV 10nM
orexinA
S/B
EC50 [nM]
A Screening workflow
Z'
2
3
4
5
6
7
8
plate no
1
2
3
4
5
6
7
8
plate no
 export data as txt files
 import into internal Actelion sof tware
Figure 3. RTCA HT integration into the Agilent Automation System. For integration a new branch of Agilent’s scheduling software VWorks4.0 together with
a newly developed driver was installed on the system computer. The 2 RTCA HT Stations were placed on the deck of the Automation Platform next to 2 liquid
pipetting workstations, see B and C. Note that the 2 RTCA Stations were also connected to an RTCA HT Analyzer and an RTCA HT Control Unit. The kinetic
files were automatically recorded on the Control Unit, data reduction was performed via Bridge Server of the Control Unit, exported as txt files and evaluated in
a proprietary HTS software (A).
RTCA HT Assay Optimization
The RTCA HT Instruments integrated in the Agilent Automation system delivered data sets with very high intra-assay as well as inter-assay reproducibility
Conclusions
We have integrated 2 xCELLigence RTCA HT Instruments (Roche Applied Science) into our double BioCel1200 Automation Platform from
Agilent. Integration, performance and reliability was assessed in automated screening protocols, by performing a secondary assay on hits that
were previously identified in a calcium flux FLIPR assay to block the Ox1 GPCR receptor.
OrexinA dose response curve
OxA [nM]
0
First Addition (compounds)
Baseline
Figure 6. Quality metrics for the 2 independent Ox1 RTCA HT Secondary Screens. Note that every assay plate contained a reference antagonist dilution
series and an orexinA dilution series. A Quality metrics of the first impedance screen. B, C Representative EC50 curve of orexinA agonist and IC50 curve of a
reference antagonist. D Intra-assay reproducibility was assessed by determining the on-plate reference EC50 and IC50 values for the 8 microplates, both for the
first and second impedance screen. E Impedance EC50/IC50 values compared to calcium flux FLIPR values. F % inhibition values of adjacent duplicates of
Screen 1 are plotted against each other. G Inter-assay reproducibility was determined by repeating the whole impedance screening on a different day and
plotting % inhibition values (averages) from the first screening against those of the second screening.
1.2
Second Addition (orexinA)
3.7
11
use 10nM orexinA for Screening
33
Programme steps
100
Data Export
 The 2 RTCA HT Instruments could be easily integrated into the Agilent Automation System
 Assay development was straight forward
 384-well E-plates were of high quality, single well failure rate was <1 well/plate
 Z’ factors for this assay were around 0.5
 Coefficients of Variation (CV) were < 5%
 Excellent intra-assay as well as inter-assay reproducibility (correlation coefficients were 0.96 and 0.86, respectively)
 EC50 values of orexinA and IC50 values of the reference antagonist showed little variation across plates and across screening runs
 65% of Ox1 FLIPR antagonists could be confirmed by the RTCA HT Instruments
 Impedance could detect unspecific FLIPR hits (51% of the non-confirmed hits)
The xCELLigence RTCA HT System
Figure 4. OrexinA concentration response of the optimized automated RTCA assay. 4000 cells/well were seeded in 384-well E-plates and incubated over
night at 37°C 5% CO2 with measuring impedance at regular intervals. Next day, cell medium was replaced and cells were equilibrated for 50min. A baseline
measurement was taken, compounds were added (First Addition: 13mM final conc.) and impedance changes were measured for 3min. In the Second Addition,
different concentrations of orexinA (in hexaplicates) were added and impedance was recorded for 25min.
The optimized work flow generated highly reproducible impedance traces with small error bars. For screening, 10nM orexin will be optimal, and
impedance will be measured for 25min after orexinA addition.
-
is easy to integrate into automated workflows in Agilent’s Biocel Platforms
is a powerful secondary screening tool
provides information-rich label free kinetic data on compounds
provides additional information on compounds not accessible by classical GPCR technologies
Acknowledgements:
We are grateful to Udo Eichenlaub, Kairat Madin, Stefan Haelg, Markus Scheuermann and Thomas Nikolaus from
Roche Diagnostics, and Lesley Schultz (installation), Bill Rust (driver), Betty Li (software build releases) and Li
Zeng from Agilent Technologies for support and early access to this technology.
* Corresponding authors:
[email protected] and [email protected]