Automation and Miniaturization of Immunoassays for Drug Discovery: AlphaLISA™, a Sensitive No-Wash Assay Sophie Dahan, Nancy Gauthier, Claire Normand, Marjolaine Roy, Véronique Brechler and Stéphane Parent PerkinElmer BioSignal, Inc., 1744 William Street, Suite 600, Montreal, Quebec H3J 1R4, Canada 5 Analyte 20 µL 5 Incubation (RT, dark) 6 Reading 25 µL PKI Cat# Color Assay volume Protocol OptiPlate™-96 6005290 white 50 µL A ½ AreaPlate-96 6005560 white 50 µL A ProxiPlate™-96 6006290 white 50 µL A OptiPlate™-384 6007290 white 50 µL A AlphaPlate™-384 6005350 light-gray 50 µL A ProxiPlate™-384 Plus 6008280 white 20 µL B custom light-gray 20 µL B 6004350 light-gray 10 µL C ProxiPlate™-384 (light-gray) AlphaPlate™-1536 * * White OptiPlates-1536 are not recommended in AlphaLISA assays with high counts due to higher crosstalk compared to light-gray AlphaPlate-1536 2.5 µL 1/2 AreaPlate-96 (white) EC50 = 198 µU/mL LDL = 0.1 µU/mL 10,000,000 1,000,000 AlphaPlate-384 (light-gray) EC50 = 172 µU/mL LDL = 0.7 µU/mL 100,000 10,000 1,000 -∞ -8 -7 -6 -5 -4 -3 8 -2 OptiPlate-96 (white) EC50 = 184 µU/mL LDL = 0.8 µU/mL OptiPlate-384 (white) Max counts: 4.8 M 5,000,000 1/2 AreaPlate-96 (white) Max counts: 4.6 M 4,000,000 3,000,000 AlphaPlate-384 (light-gray) Max counts: 3.7 M 2,000,000 OptiPlate-96 (white) Max counts: 2.8 M 1,000,000 0 -∞ -8 -7 -6 -5 -4 -3 -2 Log [Insulin] U/mL linear scale ProxiPlate-96 (white) Max counts: 1.6 M 10,000,000 ProxiPlate-384 Plus (white) EC50 = 58 µU/mL LDL = 0.14 µU/mL 100,000 1,000 ProxiPlate-384 (light-gray) EC50 = 58 µU/mL LDL = 0.12 µU/mL -∞ -8 -7 -6 -5 -4 -3 Log [Insulin] U/mL -2 1,000 -∞ -8 -7 -6 -5 -4 -3 Assay Miniaturization: Z’-Factor 10,000,000 Fully automated assay LDL = 0.51 µU/mL Manual assay LDL = 0.62 µU/mL 1,000,000 100,000 10,000 1,000 -∞ -8 -7 -6 400,000 ProxiPlate-384 Plus (white) Max counts: 5.4 M 4,000,000 3,000,000 ProxiPlate-384 (light-gray) Max counts: 4.5 M 2,000,000 1,000,000 0 linear scale -∞ -8 -7 -6 -5 -4 -3 -2 100,000 10,000 1,000 -∞ -5 -4 Manual assay LDL = 33 pg/mL 100,000 10,000 1,000 -∞ -12 -11 -10 -9 -8 -7 Log [hIgG] (g/mL) -3 -2 -1 0 1 2 Log [hEPO] (U/mL) Fully automated assay LDL = 71 pg/mL -6 10,000,000 Fully automated assay LDL = 3.5 pg/mL Manual assay LDL = 3.7 pg/mL 1,000,000 100,000 10,000 1,000 -∞ -12 -11 -10 -9 -8 -7 -6 Log [hVEGF] (g/mL) ► similar LDL obtained for all assays in automatic versus manual mode 11 max min 300,000 200,000 Z' = 0.90 100,000 0 10 20 30 40 50 60 70 80 90 100 Automation: Z’-Factor JANUS® Workstation, OptiPlate-384 White Optiplate-384 4,000,000 3,000,000 max min 2,000,000 Z' = 0.90 1,000,000 0 0 20 40 60 80 100 Well number CV% (min) CV% (max) S/B 7.1% 3.5% 252 Automation: JANUS® Workstation → liquid transfers from any combination of laboratory container including 384-well, 1536-well plates possible → suited for dispensing into 96-, 384- and 1536-well plates 8.4% 3.4% 788 Light-gray AlphaPlate-384 3,000,000 max min 2,000,000 1,000,000 Z' = 0.89 0 0 20 40 60 80 100 Well number CV% (min) CV% (max) S/B 7.9% 3.7% 791 ► low CV (≤ 8 %) and high Z’-value (0.9) obtained for both plates ► lower counts obtained with light-gray AlphaPlate, but similar S/B ratios ► both plates, white OptiPlate-384 and light-gray AlphaPlate-384 are suited for insulin AlphaLISA assay ► low CV (≤ 7 %) and high Z’ value (0.9) obtained ► light-gray AlphaPlate-1536 is recommended for insulin AlphaLISA assay The JANUS® is a robotic liquid handling system designed for the efficient automation of sample preparation procedures: 5,000,000 -3 1,000,000 Well number 6,000,000 -4 Fully automated assay LDL = 166 µU/mL Manual assay LDL = 284 µU/mL 1,000,000 500,000 CV% (min) CV% (max) S/B 9 -5 10,000,000 Log [hInsulin] (U/mL) -2 Insulin AlphaLISA Assay, AlphaPlate-1536 0 Plate Comparison: Insulin Assay 10,000 AlphaPlate-1536 (protocol C2) EC50 = 33 µU/mL LDL = 0.3 µU/mL 10,000 Log [Insulin] U/mL Protocol B: 20 µL 1,000,000 100,000 ► similar maximum counts obtained for both protocols ► as expected, protocol C2 (5 µL insulin standard) results in a better sensitivity compared to protocol C1 (1 µL) OptiPlate-384 (white) EC50 = 163 µU/mL LDL = 0.4 µU/mL AlphaLISA Signal (counts) Plate 5 µL ► similar EC50 and LDL for all plates except ProxiPlate-96 (less sensitive for AlphaLISA, not recommended) ► counts proportional to liquid height: → similar high count levels for Optiplate-384, ½ Areaplate-96 → lowest counts for Optiplate-96, Proxiplate-96 → lower counts for AlphaPlate-384 compared to OptiPlate-384 due to light-gray color, but similar LDL, EC50 (note: light-gray color reduces cross-talk) 6 Plate Comparison: Plates PKI Plates and Assay setup 10 µL Plate Comparison: Insulin Assay Protocol A: 50 µL ProxiPlate-96 (white) EC50 = 236 µU/mL LDL = 5 µU/mL AlphaLISA Signal (counts) 3 2.5 µL AlphaPlate-1536 (protocol C1) EC50 = 166 µU/mL LDL = 0.5 µU/mL 1,000,000 30 min Log [Insulin] U/mL The biotinylated anti-analyte antibody binds to the Streptavidin-coated donor beads while another anti-analyte antibody is conjugated to AlphaLISA acceptor beads. In the presence of the analyte, the beads come into close proximity. The excitation of the donor beads provokes the release of singlet oxygen molecules that triggers a cascade of energy transfer to the acceptor beads resulting in a sharp peak of light emission at 615 nm. 4 µL EnVision® instrument Anti-Analyte conjugated AlphaLISA Acceptor Beads Streptavidin-coated Donor Beads 5 µL 10,000,000 AlphaLISA Signal (counts) C2 5 µL AlphaLISA Signal (counts) C1 1 µL 60 min Streptavidin donor beads AlphaLISA Signal (counts) Emission 615 nm biotinylated anti-analyte B 5 µL Incubation (RT) 4 1 A 5 µL AlphaLISA Signal (counts) 3 Δ O2 Excitation 680 nm Mix: anti-insulin acceptor beads (10 µg/mL final) + biotinylated anti-insulin antibody (1 nM final) Automation: Manual vs. JANUS® hInsulin, hEPO, hIgG, hVEGF (Optiplate-384, 50 µL) AlphaLISA Signal (counts) 2 Quantitative detection of insulin in various 96-, 384- and 1536-well microplates was performed with final assay volumes ranging from 10 – 50 µL. Lower detection limits (LDL), EC50 values and maximum counts were compared. Also, Insulin, EPO, VEGF and IgG assays were fully automated using the JANUS® Automated Workstation with very low variation and assay parameters similar to the manual mode. Excellent sensitivities were obtained demonstrating how easy and simple miniaturization and automation are using AlphaLISA. AlphaLISA: Principle Insulin standard dilutions AlphaLISA Signal (counts) 1 10 AlphaLISA Signal (counts) Step Action Assay Miniaturization: Insulin Assay Protocol C: 10 µL assays with 1 or 5 µL of sample AlphaLISA Signal (counts) AlphaLISA™, an immunoassay for the detection of various analytes and biomarkers, offers many advantages over ELISA-like technologies. AlphaLISA is homogeneous, thus not requiring a single wash step. Samples can be tested without dilution due to the large dynamic range (approximately 3 – 5 logs). AlphaLISA assays are simple to develop, rapid to perform as well as sensitive. The technology allows working with small volumes (1 – 5 µL) and is therefore easy to miniaturize and automate enabling High Throughput Screening (HTS). 2 7 Plate Comparison: Protocols Insulin AlphaLISA Assay - Protocols AlphaLISA Signal (counts) 4 Introduction AlphaLISA Signal (counts) 1 12 Conclusions ¾ AlphaLISA is the easiest ELISA technology ever and offers the following advantages: * Detection of molecules of interest in a highly sensitive, quantitative, reproducible and user-friendly manner. * No wash step required, low sample volumes: easy to miniaturize and automate using the JANUS® workstation as demonstrated with four different AlphaLISA assays. -2 ¾ AlphaLISA Log [Insulin] U/mL ► similar EC50 and LDL for white and light-gray ProxiPlate-384 ► more sensitive than 50-µL assay due to higher final analyte concentration ► lower counts for light-gray (custom) compared to white ProxiPlate-384 due to color, but similar LDL, EC50 (note: light-gray color reduces cross-talk) AlphaLISA assay automation: → JANUS® workstation controlled using the WinPREP® 4.1 software → all serial dilutions of the standard curve and reagent dispensing in the 384-well OptiPlate carried out using the Varispan™ (8-tip dispensing arm) → JANUS® deck protected from light while dispensing the donor beads → all steps of the manual assay mimicked during automation (e.g. tip pre-wetting, tip change) and same protocol and reagents used for the automated and manual assays PerkinElmer Life and Analytical Sciences, 940 Winter Street, Waltham, MA USA (800) 762-4000 (+1) 203 925-4602 www.perkinelmer.com can be performed using various plate formats: * 96-well format: * 384-well format: ½AreaPlate-96 recommended all plates perform well, light-gray AlphaPlate-384 recommended for high counts due to cross-talk reduction * 1536-well format: light-gray AlphaPlate-1536 recommended