Calibration Curves in Quantitative Ligand Binding Assays ... · CV CVs
Method Development Strategies for Automation of Ligand ... · In the presented work, a strategy to...
1
Results are expressed as signal-to-noise values obtained after 12 min incubation with different concentrations of HRP-conjugated antibody. Incubation was extended to 30 min with TMB A due to the slow development observed. The observations demonstrated: Commercially available preparations with slow or fast development have limitations of use when targeting wide analysis ranges TMB enzymatic reaction is almost completed after 30 min To avoid signal saturation, fine titration of capture/detection reagents should be preferred to TMB substrate dilution Further investigations showed that the TMB substrate developed at Celerion could be stored up to 12 hr at RT and light protected (data not shown). Moreover, the kinetic of signal development observed with our preparation is suitable for wide analysis ranges when associated to optimized capture/ detection reagent concentrations as well as conditions of incubations. Stability Assessments for Automation - Immune Complexes in Read Buffer (ECLA) A key challenge encountered when developing high throughput ECLA methods is the stability of immune complexes in Read Buffer. Captured complexes lead to light emission when voltage is applied. With a SECTOR ® Imager 2400 (Meso Scale Discovery), the stepwise reading of the 24 sectors takes around 2 min per plate (Figure 4). Figure 4. Reading Sequence with SECTOR Imager 2400. In case of instability, immune complexes dissociate quickly, leading to a signal gradient observed from the left to the right side of the plate (Figure 5A). To face this issue, optimized protocols are developed to assure long term stability of immune complexes (Figure 5B). Figure 5. Stability Assessment of Immune Complexes in Read Buffer. Stability improvement showed positive impacts on: Smooth scheduling without bottleneck Data reliability Plate homogeneity Easiest selection of plate lots Case Study: Comparison of Manual vs Automated Methods (ELISA) A newly automated method for an existing manual immunogenicity assay had been developed which introduced an analytical delay of up to 3 hr from the time samples were introduced to the coated plate until they were incubated with detection reagent. This delay needed to be assessed and the stability of the samples on the coated plate needed to be confirmed, as compared to the existing, faster but less efficient manual method. Introduction Regulated laboratories proposing Ligand Binding Assays (LBAs) in support of biotherapeutics development face increasing demand to support projects with improved quality and efficiency. In comparison to small molecule analysis, large molecule bioanalysis encounters many unique challenges: Large numbers of study samples to be analyzed: screening/confirmatory/ titer assays for quasi-quantitative ADA assays or long half-life large molecules such as mAb PK assays Limited dynamic range of calibration curves Standard format used: 96-well plates, duplicate analysis Automated assays cover a wide range of applications such as: Enzymatic Immunoassays (EIA) Electrochemiluminescence Assays (ECLA) Radioimmuno Assays (RIA) Benefits of automation include improvement in: Throughput Traceability Reliability Reproducibility Robustness However, one of the greatest challenges of moving towards an automated method is stability. Even with dedicated on-board incubators, washers and readers the high throughput of samples can lead to delays in analytical steps, which all need to be stress-tested and refined. The work presented here describes: Strategies used for automation of LBAs in Celerion’s facility Examples of stability assessments made with respect to automation A stability comparison of manual and automated methods in terms of precision (case study) Method Strategies for Automation of LBAs Strategies adopted at Celerion Switzerland AG for the development of automated LBAs comprise the following key steps (Figure 1): Definition (aims, resources, time schedule) Establishment of the manual method Program development Qualification Validation Archiving After a planning phase, methods are first developed manually with automation in mind. Various incubation conditions are tested to identify critical steps and assay bottlenecks are mitigated by developing robust methods with standardized processes which utilize reagents with proven stability. Lastly, the archiving of validated automated processes allows development of similar assays with optimized conditions. Transfer of existing manual methods to automation is also proposed and is illustrated hereafter as a case study. Figure 1. Strategy for LBA Method Automation. Stability Assessments for Automation - 3,3’,5,5’-tetramethylbenzidine (TMB) Substrates (EIA) For the development of automated EIA, different commercially available TMB substrates were compared with our in-house preparation. Criteria of comparison were: Kinetic curves Stability Kinetics were assessed at two concentrations of coated HRP-conjugated antibody to mimic high and low positive samples (0.02 and 0.008 µg/mL, Figure 2A and 2B, respectively). Stability was evaluated by exposing TMB substrates on the benchtop of an automated platform before incubation on the assay plates coated with HRP-conjugated antibody (t 0 and 5 hr benchtop exposition, Figure 3A and 3B, respectively). Method Development Strategies for Automation of Ligand Binding Assays M. Montjovent, H. Williams, A. Markus, M. Groeschl, H. Faust, S. Wood and P. Struwe Celerion Switzerland AG, 8320 Fehraltorf, Switzerland Poster presentation at EBF 2015 8 th Open Meeting Barcelona, Nov 18-20, 2015 Control samples were used for the comparison. However, as assay response varies from run to run direct comparison of the OD response of control samples was inappropriate. Instead, the response of each control sample was normalized against the run-specific cut-point (rCP) and the mean values from a number of runs were compared. This comparison was repeated for each of the assay steps, screening, confirmatory and titer. An acceptance criteria of +/-30% mean bias between normalized results was set (Table 1). Table 1. Comparison of Manual vs Automated Liquid Handling for Immunogenicity. In all cases the acceptance criteria was met, showing that the response from the automated method was comparable to the manual method, and that stability of the samples and the coated plates had not been affected. Additionally, the %CV of the normalized responses showed a marked increase in precision for the automated method. The ratio between the low and high positive controls also gave a more consistent result with significantly lower %CV values, showing there was less variation between low and high control responses per plate. Discussion and Conclusion In the presented work, a strategy to develop new ligand binding assays for automation was discussed as well as practical assessments regarding stability. A comparison of quasi-quantitative data obtained in an immunogenicity assay with manual and automated methods was used to illustrate the advantages of automation. The conclusions are: Benefits of automation require well-defined strategies to develop new methods Smooth transition from method development to validation is assured when automation is clearly in perspective Stability of reagents as well as method robustness are critical parameters to assess in the perspective of automation For immunogenicity assays, the comparison of response ratio offers a suitable tool for manual vs automated method evaluation References 1. Automate it: ligand-binding assay productivity in a discovery bioanalytical setting. S.S. Leung and E.A. Dreher Bioanalysis 2013 5(14): 1775–1782 2. Laboratory automation of high-quality and efficient ligand-binding assays for biotherapeutic drug development. J. Wang, V. Patel et al. Bioanalysis 2013 5(13): 1635–1648 3. Ligand Binding Assays in the 21st century laboratory: automation. A.B. Ahene, C. Morrow et al. AAPS J. 2012 14(1): 142–153 Manual Screening Automated Screening LPC/rCP LPC/HPC LPC/rCP HPC/rCP HPC/rCP LPC/HPC Mean 2.88 4.87 0. 59 3.10 5.77 0.54 SD 0.536 1.368 0.113 0.268 0.563 0.017 %CV 18.6 28.1 18.3 8.6 9.8 3.1 n 26 26 26 24 24 24 %Bias 7.7 18.4 Manual Confirmatory Automated Confirmatory LPC/rCP HPC/rCP HPC/rCP LPC/HPC LPC/rCP LPC/HPC Mean SD %CV n %Bias -21.4 -22.0 0.54 0.031 5.7 60 4.29 0.418 9.7 60 2.34 0.258 11.1 60 2.97 5.50 0.54 0.506 0.822 0.067 17.0 15.0 12.3 55 55 55 Manual Titer Automated Titer LPC/rCP LPC/HPC LPC/rCP HPC/rCP HPC/rCP LPC/HPC Mean SD %CV n %Bias -19.8 -20.9 0.53 0.028 5.4 31 4.86 0.441 9.1 32 2.57 0.242 9.4 31 3.20 6.14 0.52 0.415 0.686 0.036 13.0 11.2 7.0 146 146 146 SOP Writing Method Limitations Program description Analytical run information Recovery method Program Qualification (PROD Environment) Automated sample transfer PK assays ADA assays 1 run for carryover (min 12 NC and 12 ULOQ or HPC) 3 P&A runs with 5 QC levels in duplicate 3 runs with PC (2 sets at each level); if confirmatory assay, 2 sets of drug-inhibited LPC and HPC in 3 runs) Automated dilution PK assays ADA assays 1 run for carryover (min 12 NC and 12 ULOQ or HPC) 3 P&A runs and 2 runs with 6 diluted DQC in duplicate 2 runs with 6 diluted PC Archived Automated Methods Master method Liquid classe(s) Labware (DiTis, carriers, racks) N Criteria Fullfilled Establishment of the Method Manually, Keeping Automation in Mind Experimental conditions, limitations STD curve, QCs, PCs To avoid bottleneck: Test the robustness of incubation(s) and the possibility to store diluted samples Check the stability of reagents at RT Problem Isolation Technical (dilutor, o’ring) Settings (liquid classes, vectors) Program adaptation Aims Reached Y N Validation Define Aims Time schedule, availability of the instrument(s) Step(s) to be automated Sample volumes Platform(s) to be used Sample tubes format Program Development (DEV Environment) Optimize the deck layout with method specific labware Create and test method specific liquid classes by gravimetry Create the core of the program Test the system with electronic data file input Perform stress tests Y A B 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 0 10 20 30 40 50 OD 450 nm Duration of incubation in TMB Substrate (min) TMB A TMB B TMB C TMB D TMB E TMB Celerion 0 0.5 1 1.5 0 10 20 30 40 50 OD 450 nm Duration of incubation in TMB Substrate (min) TMB A TMB B TMB C TMB D TMB E TMB Celerion High concentration of HRP conjugate Low concentration of HRP conjugate A B 0 10 20 30 40 50 0.0008 0.004 0.02 0.1 Signal to noise ratio Ab-HRP (ug/mL) TMB A TMB B TMB C TMB D TMB E TMB Celerion 0 10 20 30 40 50 0.0008 0.004 0.02 0.1 Signal to noise ratio Ab-HRP (ug/mL) TMB A TMB B TMB C TMB D TMB E TMB Celerion Before benchtop exposition (t 0 ) After benchtop exposition (5 hr) 1 2 3 4 5 6 7 8 9 10 11 12 A B C D E F G H Sector 1 Sector 24 0 20 40 60 80 100 120 0 10 20 30 40 50 60 Mean sector RLU values (% of mean observed at to) Duration of incubation in Read Buffer (min) 0 20 40 60 80 100 120 0 8 16 24 Mean sector RLU values (% of mean observed at to) Plate sector A B Without stability improvement With stability improvement The 24 sectors were read in 2 min One sector was read each 10 or 20 min Figure 2. Comparison of Kinetic Curves in TMB Substrates. Figure 3. Stability of TMB Substrates.
Transcript of Method Development Strategies for Automation of Ligand ... · In the presented work, a strategy to...
Results are expressed as signal-to-noise values obtained after 12 min incubation with different concentrations of HRP-conjugated antibody. Incubation was extended to 30 min with TMB A due to the slow development observed.
The observations demonstrated: Commercially available preparations with slow or fast development have limitations of use when targeting wide analysis ranges
TMB enzymatic reaction is almost completed after 30 min To avoid signal saturation, fine titration of capture/detection reagents should be preferred to TMB substrate dilution
Further investigations showed that the TMB substrate developed at Celerion could be stored up to 12 hr at RT and light protected (data not shown). Moreover, the kinetic of signal development observed with our preparation is suitable for wide analysis ranges when associated to optimized capture/detection reagent concentrations as well as conditions of incubations.
Stability Assessments for Automation - Immune Complexes in Read Buffer (ECLA)
A key challenge encountered when developing high throughput ECLA methods is the stability of immune complexes in Read Buffer. Captured complexes lead to light emission when voltage is applied.
With a SECTOR® Imager 2400 (Meso Scale Discovery), the stepwise reading of the 24 sectors takes around 2 min per plate (Figure 4).
Figure 4. Reading Sequence with SECTOR Imager 2400.
In case of instability, immune complexes dissociate quickly, leading to a signal gradient observed from the left to the right side of the plate (Figure 5A). To face this issue, optimized protocols are developed to assure long term stability of immune complexes (Figure 5B).
Figure 5. Stability Assessment of Immune Complexes in Read Buffer.
Stability improvement showed positive impacts on: Smooth scheduling without bottleneck Data reliability Plate homogeneity Easiest selection of plate lots
Case Study: Comparison of Manual vs Automated Methods (ELISA)
A newly automated method for an existing manual immunogenicity assay had been developed which introduced an analytical delay of up to 3 hr from the time samples were introduced to the coated plate until they were incubated with detection reagent. This delay needed to be assessed and the stability of the samples on the coated plate needed to be confirmed, as compared to the existing, faster but less efficient manual method.
IntroductionRegulated laboratories proposing Ligand Binding Assays (LBAs) in support of biotherapeutics development face increasing demand to support projects with improved quality and efficiency.
In comparison to small molecule analysis, large molecule bioanalysis encounters many unique challenges: Large numbers of study samples to be analyzed: screening/confirmatory/titer assays for quasi-quantitative ADA assays or long half-life large molecules such as mAb PK assays
Limited dynamic range of calibration curves Standard format used: 96-well plates, duplicate analysis
Automated assays cover a wide range of applications such as: Enzymatic Immunoassays (EIA) Electrochemiluminescence Assays (ECLA) Radioimmuno Assays (RIA)
Benefits of automation include improvement in: Throughput Traceability Reliability Reproducibility Robustness
However, one of the greatest challenges of moving towards an automated method is stability. Even with dedicated on-board incubators, washers and readers the high throughput of samples can lead to delays in analytical steps, which all need to be stress-tested and refined.
The work presented here describes: Strategies used for automation of LBAs in Celerion’s facility Examples of stability assessments made with respect to automation A stability comparison of manual and automated methods in terms of precision (case study)
MethodStrategies for Automation of LBAs
Strategies adopted at Celerion Switzerland AG for the development of automated LBAs comprise the following key steps (Figure 1): Definition (aims, resources, time schedule) Establishment of the manual method Program development Qualification Validation Archiving
After a planning phase, methods are first developed manually with automation in mind. Various incubation conditions are tested to identify critical steps and assay bottlenecks are mitigated by developing robust methods with standardized processes which utilize reagents with proven stability. Lastly, the archiving of validated automated processes allows development of similar assays with optimized conditions. Transfer of existing manual methods to automation is also proposed and is illustrated hereafter as a case study.
Figure 1. Strategy for LBA Method Automation.
Stability Assessments for Automation - 3,3’,5,5’-tetramethylbenzidine (TMB) Substrates (EIA)
For the development of automated EIA, different commercially available TMB substrates were compared with our in-house preparation. Criteria of comparison were: Kinetic curves Stability
Kinetics were assessed at two concentrations of coated HRP-conjugated antibody to mimic high and low positive samples (0.02 and 0.008 µg/mL, Figure 2A and 2B, respectively). Stability was evaluated by exposing TMB substrates on the benchtop of an automated platform before incubation on the assay plates coated with HRP-conjugated antibody (t0 and 5 hr benchtop exposition, Figure 3A and 3B, respectively).
Method Development Strategies for Automation of Ligand Binding AssaysM. Montjovent, H. Williams, A. Markus, M. Groeschl,H. Faust, S. Wood and P. StruweCelerion Switzerland AG, 8320 Fehraltorf, Switzerland
Poster presentation at EBF 2015 8th Open Meeting Barcelona, Nov 18-20, 2015
Control samples were used for the comparison. However, as assay response varies from run to run direct comparison of the OD response of control samples was inappropriate. Instead, the response of each control sample was normalized against the run-specific cut-point (rCP) and the mean values from a number of runs were compared. This comparison was repeated for each of the assay steps, screening, confirmatory and titer. An acceptance criteria of +/-30% mean bias between normalized results was set (Table 1).
Table 1. Comparison of Manual vs Automated Liquid Handling for Immunogenicity.
In all cases the acceptance criteria was met, showing that the response from the automated method was comparable to the manual method, and that stability of the samples and the coated plates had not been affected. Additionally, the %CV of the normalized responses showed a marked increase in precision for the automated method. The ratio between the low and high positive controls also gave a more consistent result with significantly lower %CV values, showing there was less variation between low and high control responses per plate.
Discussion and ConclusionIn the presented work, a strategy to develop new ligand binding assays for automation was discussed as well as practical assessments regarding stability. A comparison of quasi-quantitative data obtained in an immunogenicity assay with manual and automated methods was used to illustrate the advantages of automation.
The conclusions are: Benefits of automation require well-defined strategies to develop new methods Smooth transition from method development to validation is assured when automation is clearly in perspective
Stability of reagents as well as method robustness are critical parameters to assess in the perspective of automation
For immunogenicity assays, the comparison of response ratio offers a suitable tool for manual vs automated method evaluation
References1. Automate it: ligand-binding assay productivity in a discovery bioanalytical
setting. S.S. Leung and E.A. Dreher Bioanalysis 2013 5(14): 1775–1782
2. Laboratory automation of high-quality and efficient ligand-binding assays for biotherapeutic drug development. J. Wang, V. Patel et al. Bioanalysis 2013 5(13): 1635–1648
3. Ligand Binding Assays in the 21st century laboratory: automation. A.B. Ahene, C. Morrow et al. AAPS J. 2012 14(1): 142–153
Manual Screening Automated Screening
LPC/rCP LPC/HPCLPC/rCP HPC/rCP HPC/rCP LPC/HPC
Mean 2.88 4.87 0. 59 3.10 5.77 0.54
SD 0.536 1.368 0.113 0.268 0.563 0.017
%CV 18.6 28.1 18.3 8.6 9.8 3.1
n 26 26 26 24 24 24
%Bias 7.7 18.4
Manual Confirmatory Automated Confirmatory
LPC/rCPHPC/rCP HPC/rCP LPC/HPCLPC/rCP LPC/HPC
Mean
SD
%CV
n
%Bias -21.4 -22.0
0.54
0.031
5.7
60
4.29
0.418
9.7
60
2.34
0.258
11.1
60
2.97 5.50 0.54
0.506 0.822 0.067
17.0 15.0 12.3
55 55 55
Manual Titer Automated Titer
LPC/rCP LPC/HPCLPC/rCP HPC/rCP HPC/rCPLPC/HPC
Mean
SD
%CV
n
%Bias -19.8 -20.9
0.53
0.028
5.4
31
4.86
0.441
9.1
32
2.57
0.242
9.4
31
3.20 6.14 0.52
0.415 0.686 0.036
13.0 11.2 7.0
146 146 146
SOP WritingMethodLimitationsProgram descriptionAnalytical run informationRecovery method
Program Qualification (PROD Environment)Automated sample transfer
PK assays ADA assays1 run for carryover (min 12 NC and 12 ULOQ or HPC)
3 P&A runs with 5 QC levels in duplicate
3 runs with PC (2 sets at each level); if confirmatory assay, 2 sets of drug-inhibited LPC and HPC in
3 runs)Automated dilution
PK assays ADA assays1 run for carryover (min 12 NC and 12 ULOQ or HPC)
3 P&A runs and 2 runs with 6 diluted DQC in duplicate
2 runs with 6 diluted PC
Archived Automated Methods
Master methodLiquid classe(s)Labware (DiTis, carriers, racks)
N
Criteria Fullfilled
Establishment of the Method Manually, Keeping Automation in Mind
Experimental conditions, limitationsSTD curve, QCs, PCs
To avoid bottleneck:Test the robustness of incubation(s) and the possibility to store diluted samplesCheck the stability of reagents at RT
Problem Isolation
Technical (dilutor, o’ring)Settings (liquid classes, vectors)Program adaptation
Aims Reached
Y
N
Validation
Define
Aims Time schedule, availability of the instrument(s)Step(s) to be automated Sample volumesPlatform(s) to be used Sample tubes format
Program Development (DEV Environment)
Optimize the deck layout with method specific labwareCreate and test method specific liquid classes by gravimetryCreate the core of the programTest the system with electronic data file inputPerform stress tests
Y
A B
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
0 10 20 30 40 50
OD
450
nm
Duration of incubation in TMB Substrate (min)
TMB A
TMB B
TMB C
TMB D
TMB E
TMB Celerion
0
0.5
1
1.5
0 10 20 30 40 50
OD
450
nm
Duration of incubation in TMB Substrate (min)
TMB A
TMB B
TMB C
TMB D
TMB E
TMB Celerion
High concentration of HRP conjugate Low concentration of HRP conjugate
A B
0
10
20
30
40
50
0.0008 0.004 0.02 0.1
Sign
al to
noi
se ra
tio
Ab-HRP (ug/mL)
TMB A
TMB B
TMB C
TMB D
TMB E
TMB Celerion
0
10
20
30
40
50
0.0008 0.004 0.02 0.1
Sign
al to
noi
se ra
tio
Ab-HRP (ug/mL)
TMB A
TMB B
TMB C
TMB D
TMB E
TMB Celerion
Before benchtop exposition (t0) After benchtop exposition (5 hr)
A B
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
0 10 20 30 40 50
OD
450
nm
Duration of incubation in TMB Substrate (min)
TMB A
TMB B
TMB C
TMB D
TMB E
TMB Celerion
0
0.5
1
1.5
0 10 20 30 40 50
OD
450
nm
Duration of incubation in TMB Substrate (min)
TMB A
TMB B
TMB C
TMB D
TMB E
TMB Celerion
High concentration of HRP conjugate Low concentration of HRP conjugate
A B
0
10
20
30
40
50
0.0008 0.004 0.02 0.1
Sign
al to
noi
se ra
tio
Ab-HRP (ug/mL)
TMB A
TMB B
TMB C
TMB D
TMB E
TMB Celerion
0
10
20
30
40
50
0.0008 0.004 0.02 0.1
Sign
al to
noi
se ra
tio
Ab-HRP (ug/mL)
TMB A
TMB B
TMB C
TMB D
TMB E
TMB Celerion
Before benchtop exposition (t0) After benchtop exposition (5 hr)
1 2 3 4 5 6 7 8 9 10 11 12
A
B
C
D
E
F
G
HSector 1
Sector 24
0
20
40
60
80
100
120
0 10 20 30 40 50 60
Mea
n se
ctor
RLU
val
ues
(% o
f mea
n ob
serv
ed a
t to)
Duration of incubation in Read Buffer (min)
0
20
40
60
80
100
120
0 8 16 24
Mea
n se
ctor
RLU
val
ues
(% o
f mea
n ob
serv
ed a
t to)
Plate sector
A BWithout stability improvement With stability improvement
The 24 sectors were read in 2 min One sector was read each 10 or 20 min
1 2 3 4 5 6 7 8 9 10 11 12
A
B
C
D
E
F
G
HSector 1
Sector 24
0
20
40
60
80
100
120
0 10 20 30 40 50 60
Mea
n se
ctor
RLU
val
ues
(% o
f mea
n ob
serv
ed a
t to)
Duration of incubation in Read Buffer (min)
0
20
40
60
80
100
120
0 8 16 24
Mea
n se
ctor
RLU
val
ues
(% o
f mea
n ob
serv
ed a
t to)
Plate sector
A BWithout stability improvement With stability improvement
The 24 sectors were read in 2 min One sector was read each 10 or 20 min
Figure 2. Comparison of Kinetic Curves in TMB Substrates.
Figure 3. Stability of TMB Substrates.
