Importance of calcium assay parameters in drug discovery

Actelion Pharmaceutical Allschwill Hamamatsu 10th FDSS User Meeting June 2014 Importance of calcium assay parameters in drug discovery Sabine Rouanet Dr. Isabelle Bertrand Directeur: Pr. Jean-Charles Schwartz Dr. Stéphane Krief Dr. Thierry Calmels Success is the ability to go from one failure to another with no loss of enthusiasm (Sir Winston Churchill)

GPCRs signaling Gαq Gαs Gαi Gαq Gαs Gαi Functional selectivity Several ligand-specific receptor conformations can be associated to biased functionnal signaling

Precise affinity required for GPCR antagonism Advance SAR analysis Studying drug specificity Accurate affinity values for pre-development compounds Identification of biased signaling Correlation calcium & CRE-MRE reporter assays Bias plot for histamine H2 agonists Gαs / camp Agonism % of Max reference agonist response % of 10 µm HA response on GTPγ on GTPγ 35 S binding assay 35 S assay 100 80 60 40 20 0 Histamine Amthamine A B 0 20 40 60 80 100 120 % of 10 µm HA response on calcium assay % of Max reference agonist response on calcium assay Inverse Agonism Inverse Agonism Agonism β-arrestin β-blocker carvedilol Cardioprotective effects Thanawala VJ et al, Curr Opin Pharmacol. 2014 Mar 26;16C:50-57 Agonism Antagonism

Calcium mobilization assays at Bioprojet: HTS (Identify and classify hits) Evaluate agonism efficacy and affinity Evaluate type of antagonism (Schild regression analysis) Identification of biased ligands

GPCR antagonism and Calcium assay in drug discovery Need to obtain precise affinity by Kb determination in Calcium assays Kb is applicable at equilibrium conditions that are not encountered with functional calcium assays (incubation exceeds 4 times the dissociation t 1/2 of ligand/receptor)

Use of the pa2 as a universal determinant of antagonist potency pa2 = pkb + log ( 1+ 2 [ A ] / Ka ) At low [agonist] occupancy [ A ] < < < Ka pa2 ~ pkb + log (1) pa2 tend towards the pkb Calculation of pa2 at low agonist responses Overcome the potential bias associated with non equilibrium conditions Concentration response curve dextral displacement Max response reduction Estimate insurmountable antagonists affinity Steven J Charlton and Georges Vauquelin, 2010, British J Pharmacol 161:1250 1265 Terry Kenakin, 2009, A pharmacology Primer: Theory, Application and Methods, Chapter 11, Academic Press Terry Kenakin et al, 2006, JPET 319:710 723 Arthur Christopoulos et al, 1999, Euro J Pharmacol, 382:217 227 Non equilibrium

pa2 = - log [M] of antagonist producing a 2 fold shift of the agonist concentration response curve Use of Dose Ratio (DR) values as surrogate parameter for calculation of pa2 pa2 = log ( DR 1 ) log B At DR = 2 pa2 = log B Competitive surmountable Antagonism at equilibrium DR at EC50 Non competitive (Insurmountable) Antagonism at Hemi-equilibrium DR at low agonist response

Calcium assay parameters and GPCRs-ligand accessibility 1. Adherent vs suspension cells 2. Receptor functionality at the cell membrane 3. Ligand diffusion

GPCRs and ligand accessibility 1. Adherent vs suspension cells 2. Receptor functionality at the cell membrane 3. Ligand diffusion

Settings: 10µl/sec, height 9.6 mm, sensitivity 200ms, gain 1 Calcium Flux on HEK293 cell suspension Calcium Flux on MSR1-HEK293 adherent cells Max-Min (Fluorescence Arbitrary Units) 14000 12000 10000 8000 6000 4000 2000 0 1µM + BP1 antagonist 1µM + BP2 antagonist Ki= 230 nm EC50=330nM Max-Min (Fluorescence Arbitrary Units) 6000 5000 4000 3000 2000 1000 0 1µM + BP1 1µM + BP2 MSR1: macrophage scavenger receptor 1 Ki> 5 µm Ki= 57.4 nm EC50= 500 nm 1E-9 1E-8 1E-7 1E-6 1E-5 1E-4 1E-3 0.01 0.1 1 10 100 100010000100000 Concentrations (µm) 1E-9 1E-8 1E-7 1E-6 1E-5 1E-4 1E-3 0.01 0.1 1 10 100 100010000100000 Concentrations (µm) BP2 antagonist Adherent Ki (nm) 57 nm Suspension Ki (nm) 230 nm BP1 antagonist > 5 µm Inactive

GPCRs and ligand accessibility 1. Adherent vs suspension cells 2. Receptor functionality at the cell membrane 3. Ligand diffusion

Calcium assay on recombinant-gpcr1 expressing HEK293 cells : Arbitrary Fluorescence units (A.F.U) EC50 = 300 nm Time

Calcium assay on native-gpcr1 in HUVEC cells : EC50 = 1.1 µm Arbitrary Fluorescence units (A.F.U) Time Involvement of receptor reserve, agonist-induced structural modifications.? Importance of GPCR expressing cells when looking at the calcium response

Y = A + B * X Parameter Value Error A -0,23066 0,29386 B 1,57048E8 4,85289E7 R SD N P 0,95543 0,32433 3 0,19079 ---------------------------------------------------- Parameter Value Error A -0.72622 0.37705 B 2.09302E8 6.22684E7 R SD N P 0.95848 0.41616 3 0.18409 Calcium assay on native-gpcr1 expressing cells : Evaluation of BPx antagonist (from 1nM to 100 nm) against 3µM reference agonist 14000 12000, EC50= 1.2 µm + 1 nm Antagonist BPx + 3 nm Antagonist BPx + 10 nm Antagonist BPx + 30 nm Antagonist BPx + 100 nm Antagonist BPx, EC50 = 1.6 µm 3200000 3000000, EC50= 1.2 µm + 1 nm Antagonist BPx + 3 nm Antagonist BPx + 10 nm Antagonist BPx + 30 nm Antagonist BPx + 100 nm Antagonist BPx, EC50 = 1.2 µm Max-Min (F.A.U) 10000 8000 6000 4000 2000 log (DR-1) 1,4 1,2 1,0 0,8 0,6 0,4 0,2 pa2=1.5 nm 0,0-0,2-0,4 0,00E+000 2,00E-009 4,00E-009 6,00E-009 8,00E-009 1,00E-008 log [BP1.7577] Log[agonist] pa2 = 8.82 AUC (Integrale) 2800000 2600000 2400000 2200000 log (DR-1) 2.5 2.0 1.5 1.0 0.5 0.0-0.5-1.0 pa2= 8.46 0.00E+000 5.00E-009 1.00E-008 log [agonist] pa2 = 8.46 0 2000000 1E-30.01 10 100 1000 10000 100000 (nm) 1800000 1 10 100 1000 10000 100000 (nm) No major difference observed when calculating pa2 using Max-Min or A.U.C data

GPCRs and ligand accessibility 1. Adherent vs suspension cells 2. Receptor functionality at the cell membrane 3. Ligand diffusion

Calcium assay: Rapid and transient signaling system under non equilibrium condition Influenced by the diffusion characteristics of the injected agonist

Calcium assay: Rapid and transient signaling system under non equilibrium condition Influenced by the diffusion characteristics of the injected agonist Diffusion Movement of a fluid from higher concentration to lower concentration The particles will mix until they are evenly distributed This phenomenon of particles distribution is governed by the first and second laws of Fick

The diffusion phenomenom for the agonist may be of importance regarding : Depth and rate of agonist injection Nature and size of considered agonists (aminergic, lipidic, peptidic ligands) Viscosity of the assay buffer (basic methodology vs NW kits) Volume and surface area of the assay well 96 well plate (full or ½ size wells)

The diffusion phenomenom for the agonist may be of importance regarding : Depth and rate of agonist injection (small molecule ligand) For antagonism charaterization injection: 10 µl / sec at 9.6 mm height

0,2 0,0-0,2-0,4-0,6-0,8-1,0-1,2-1,4 yscale(y) = A + B * xscale(x) where scale() is the current axis scale function. Parameter Value Error A 2,45896 0,98345 B 0,53107 0,1622 R SD N P 0,91802 0,25649 4 0,08198-7,0-6,5-6,0-5,5-5,0 1,2 1,0 0,8 0,6 0,4 0,2 0,0-0,2-0,4-0,6-0,8 yscale(y) = A + B * xscale(x) where scale() is the current axis scale function. Parameter Value Error A 4,63522 0,85702 B 0,74365 0,14135 R SD N P 0,96572 0,22351 4 0,03428-7,0-6,5-6,0-5,5-5,0 1,0 A 2,93449 0,20493 B 0,41618 0,03387 0,8 R SD N P 0,6 0,99021 0,05356 5 0,00116 0,4 0,2 0,0-0,2 yscale(y) = A + B * xscale(x) where scale() is the current axis scale function. Parameter Value Error -7,0-6,5-6,0-5,5-5,0 Compound BPx antagonism using CHO expressing recombinant hu-gpcr RFU (Max-Min) 11000 Ago Ago + X 100nM 10000 Ago + X 300nM Ago + X 1 µm 9000 Ago + X 3 µm Ago + X 10 µm 8000 7000 6000 5000 4000 3000 2000 1000 0 log(dr-1) log [BS6.890] pa2 = 4.63 (23 µm) Inactive EC50= 5 µm RFU (Max-Min) 11000 10000 9000 8000 7000 6000 5000 4000 3000 2000 1000 0 log (DR-1) Ago Ago + X 100nM Ago + X 300nM Ago + X 1 µm Ago + X 3 µm Ago + X 10 µm log [BS6.890] pa2 = 6.23 (589 nm) pa2 = 6.23 EC50=835nM RFU (Max-Min) 12000 11000 10000 9000 8000 7000 6000 5000 4000 3000 2000 1000 0 log (DR-1) Ago Ago + X 100nM Ago + X 300nM Ago + X 1 µm Ago + X 3 µm Ago + X 10 µm log [BS6-890] Expected pa2 = 7 pa2= 7.05 (89.1 nm) pa2 = 7.05 EC50=2µM 10 100 1000 10000 Concentration (nm) 10 100 1000 10000 Concentration (nm) 10 100 1000 10000 Concentration (nm) 100 µl height 180 µl height 240 µl height Expected pa2 = 7 (FlexStation) injection height (related volume) 4 mm (100 µl) 7.2 mm (100 µl 9.6 mm (240 µl) FDSS µcell Determined pa2 Inactive 6.23 7.05

CHO-GPCR cells in 96 well plate Antagonism study using large peptidic endogenous agonist Settings: 10µl/sec, height 3 mm sensitivity 200ms, gain 2 No signal Determined Ki for antagonist and EC50 for agonist far from expected ~ 1 nm and 30 nm, respectively 20000 EC50=310nM 15000 100nM + antagonist BPx, Ki = 17 nm Settings: 80µl/sec, height 3 mm sensitivity 200ms, gain 2 Max-Min (F.A.U) 10000 5000 0 0.01 0.1 1 10 100 1000 10000 100000 Concentration (nm)

Antagonism study on CHO-GPCR cells with large peptidic agonist Settings: 200µl/sec, height 3 mm, sensitivity 200ms, gain 2 0.03 nm to 3 µm + 1 nm Antagonist + 10 nm Antagonist + 100 nm Antagonist + 1 µm Antagonist + 10 µm Antagonist Schild regression analysis yscale(y) = A + B * xscale(x) where scale() is the current axis scale function. 22500 20000 17500 Y Axis Title 4 3 2 1 0 Parameter Value Error A 9,21877 0,70212 B 1,09386 0,09832 R SD N P 0,9881 0,31091 5 0,00156 pa2 = 8.4 (3.7 nm) EC50 = 27 nm Max-Min (F.A.U) 15000 12500 10000 7500 5000-1 1E-9 1E-8 1E-7 1E-6 1E-5 X Axis Title, EC50 = 27 nm + Antagonist BPx 1nM + Antagonist BPx 10nM + Antagonist BPx 100nM + Antagonist BPx 1µM + Antagonist BPx 10µM 2500 0 [Antagonist] 1E-4 1E-3 0.01 0.1 1 10 100 1000 10000 100000 (nm) With large peptidic ligand, fast agonist injection is required to study antagonism

HEK293 cell in 96 well plate Settings: 10µl/sec, height 3 mm, sensitivity 200ms, gain 1 No Wash BD kit assay buffer Regular protocol Fluo4 in HBSS assay buffer HEK293 cell suspension in 96 well plate none 0.3 nm 1 nm 3 nm 10 nm Max-Min (Fluorescence Arbitrary Units) 8000 7000 6000 5000 4000 3000 2000 1000 0 Protocol with NO WASH BD assay buffer Regular protocol with Fluo4 in HBSS bufer EC50 = 3.8 nm EC50 = 2.5 nm 30 nm 0.01 0.1 1 10 100 1000 µcell settings (vit 10µl/sec, height 3 mm, sensitivity 200ms, gain 1) (nm) 100 nm 300 nm Working window is too narrow to study antagonism at 10 µl/sec agonist injection

HEK293 cell in 96 well plate Antagonism study using No Wash BD kit assay buffer Settings: 200µl/sec, height 3 mm, sensitivity 200ms, gain 1 From 0.3 nm to 300 nm + 3 nm Antagonist + 10 nm Antagonist + 30 nm Antagonist + 100 nm Antagonist + 300 nm Antagonist Schild regression analysis Max-Min (Fluorescence Arbitrary Units) 8000 7000 6000 5000 4000 3000 2000 1000 0 + Antagonist 3nM + Antagonist 10nM + Antagonist 30nM + Antagonist 100nM + Antagonist 300nM EC50= 3.6 nm 0.1 1 10 100 1000 (nm) µcell settings (vit 200µl/sec, height 3mm, sensitivity 200ms, gain 1) In No Wash buffer, At 200 µl/sec agonist injection, the Kb (pa2) can be determined for an antagonism

HEK-GPCR cell suspension in 96 well plate Settings: 10µl/sec, height 9.6 mm, sensitivity 200ms, gain 1 1/2 size well Full size well 0.01 nm to 10 µm 1 µm + BP1 1 µm + BP2 0.01 nm to 10 µm 1 µm + BP1 1 µm + BP2 8000 7000 HEK-GPCR cell suspension, full size well (96 well plate) EC50=35nM none 0.01 nm 0.1 nm 1 nm Max-Min (F.A.U) 6000 5000 4000 3000 2000 1000 0 1µM + Antagonist BP1 1µM + Antagonist BP2 0.01 0.1 1 10 100 1000 10000 µcell settings: vit 10µl/sec, height 3mm, sensitivity 200ms, gain 1 Ki=13nM (nm) inactive 10 nm 100 nm 1 µm Settings for full size well are not compatible with ½ size well 10 µm

Antagonism study using ½ size well Settings: 200 µl/sec, height 3 mm sensitivity 200ms, gain 1 HEK-GPCR, cell suspension 1/2 size well Settings: 10 µl/sec, height 3 mm sensitivity 200ms, gain 1 HEK-GPCR, cell suspension full size well 12000 10000, EC50 = 69 nm 1 µm + BP1, Ki= 52.3 nm 1 µm + BP2, Ki= ki= 207 nm EC50 = 69 nm 96 Well plate 1/2 size Full size 12000 10000, EC50 = 94 nm 1 µm + BP1, ki= 52.8 nm 1 µm + BP2, Ki= 258 nm EC50 = 94 nm FAU (max-min) 8000 6000 4000 2000 0 EC50 (nm) Antagonist BP1 Ki (nm) Antagonist BP2 Ki (nm) 69 94 52.3 52.8 207 258 FAU (max-min) 8000 6000 4000 2000 0 1E-3 0.01 0.1 1 10 100 1000 10000 100000 Concentration (nm) 1E-4 1E-3 0.01 0.1 1 10 100 1000 10000 1000001000000 Concentration (nm) There is no differences when Ki is calculated from experiments done in ½ or full size well, at 3mm height but at different rate agonist injection

SUMMARY Use of the pa2 for antagonist potency to overcome the potential bias associated with non equilibrium conditions Precise and defined agonist parameters needed for any given GPCR when implementing calcium assay - Receptor functionality at cell membrane - GPCR expressing cells - Ligand diffusion Other important parameters to consider - Binding kinetics - Receptor trafficking - Receptor homo/hetero oligomerization

10000 Correlation calcium & CRE-MRE reporter assays Correlation between K Β obtained on calcium assay & MRE-CREluc whole cell assay Correlation calcium & GTPγ 35 S binding assays Correlation between K Β obtained on GTPγ35 S & calcium assays 10000 Correlation CRE-MRE reporter & GTPγ 35 S binding assays Correlation between K Β obtained on GTPγ35 S assays & MRE-CREluc whole cell assays K Β with whole cell CRE-MRE-luciferase 1000 100 10 from BP library chemical serie A from BP library chemical serie B reference antagonists 1 1 10 100 1000 10000 K Β with whole cell Calcium assay K Β with membrane GTPgS 1000 100 10 from BP library chemical serie A from BP library chemical serie B reference antagonists 1 1 10 100 1000 10000 K Β with whole cell Calcium assay K Β with membrane GTPgS 1000 100 10 from BP library chemical serie A from BP library chemical serie B reference antagonists 1 1 10 100 1000 10000 K Β with whole cell CRE-luciferase In fine, we have to keep in mind that what really matters is in vivo therapeutic efficacy Terry Kenekin, JPET 336:296 302, 2011