Experiences with Fluorescence Methods in Ultra-High Throughput ...

Experiences with Fluorescence Methods in
Ultra-High Throughput Screening and
Drug Discovery Using the Analyst
Michael Helms, Ph.D.
New Lead Discovery
Exelixis, Inc.
South San Francisco, CA

Outline
• Introduction to drug discovery at Exelixis
• FP assay on Analyst
• Fluorogenic assay on Analyst

New Lead Discovery at Exelixis
Assay
Development
High-Throughput
Screening
Confirmation
of “hits”
Determination
of potency (IC 50 )
Resynthesis
of most potent
compounds
Selectivity assays
Determination
of potency (IC 50 )
Determination
of cytotoxicity
Lead optimization and pharmacology:
determine drug properties
(ADMET and PK) in vivo and in vitro

Justification of Large Compound Library
• Currently near 2 million compounds
• Automated combinatorial chemistry produces 2 million
compounds per year
• Screening must be rapid, reliable and inexpensive
Random screen of
large, diverse
library
Focused screen of
selected set of
compounds
predicted to be
active
Advantage:
Yields diverse core
structures
Advantage:
Faster and lower
cost
Disadvantage:
More screening,
therefore slower and
higher cost
Disadvantage:
Yields less diversity,
requires
assumptions

Screening Facility
Capacity: over 20 screens per year

Requirements of Assays
• Z’-factor 0.5
• Signal:background 10
• DMSO tolerance up to 2%
• Consistent across plates and days
• Stable reagents
• Predictive: assay corresponds to events in vivo
• Identifies active compounds (IC 50 as predicted)
• Resistant to interference from test compounds
• Acceptable cost (less than 10 cents/well)

Source: Zhang et al. 1999, J. Biomolecular Screening 4: 67-73.
Quality Control of an Assay: The Z Factor
Z' 1
Good range: Z’ = 0.5 to 1
Standard deviations are
small compared to change
3
SD
3
SD
1
mean
1
mean
2
2
Bad range: Z’ < 0.5
Standard deviations are
large compared to change
Z'-Factor = 0.67
Z'-Factor = 0
150
100
Value
100
50
Value
75
50
25
0
100% Activity 0% Activity
0
100% Activity 0% Activity

Example of Assay Data
Plate 933 from Kinase Screen.
Z' Factor = 0.84, S/B = 2.5
Luminescence (cps)
400000
300000
200000
100000
100% Inhibition
Sample Wells
0% Inhibition
0
0 100 200 300 400
Well Number

Advantages of Different Assay Formats
Advantages:
Optical
Methods
Non-Optical
Methods
Fluorescence Luminescence Radiometric
Fast,
sensitive,
homogeneous
Less prone to
interference
Direct,
robust,
sensitive
Disadvantages
Prone to
Interference
Complex, slow
Hazardous,
complex,
slow,
expensive
disposal

Types of Fluorescence Assays Amenable to High-
Throughput Screening
•Fluorescence Intensity: change in intensity
• Fluorescence Resonance Energy Transfer (FRET):
change in intensity based on distance
•Fluorescence Polarization (FP): change in polarization
based on size
•Fluorescence Lifetime: based on lifetime change
•Fluorescence Correlation Spectroscopy (FCS): based on
translational diffusion (size)

Small molecule
Rotates rapidly
Low Polarization Value
(P near 0)
Large molecule
Rotates slowly
High Polarization Value
(P up to 0.5, or 500mP)
Fluorescence Polarization (FP)
I
P
I
P α V
II
II
I
I

Advantages of FP Assays in High-Throughput
Screening
• Only one fluorophore required
• Homogeneous (mix and read, no washing)
Fast
Simple
• Good for small ligands binding to proteins,
but not for protein-protein interactions

Analyst
Analyst Instrument
from LJL BioSystems
Light Source
Filter
Dichroic
Mirror
Polarizer
or
PMT
Plate

Serine/Threonine Kinase FP Assay (STX Assay)
Anti-P-Ser Ab
Kinase + Mg-ATP +
S
+
P
S
Fluor
P
S
Fluor
P
S
Fluor bound to Ab
High Polarization
Activity
Fluor free from Ab
Low Polarization

Optimization of STX Assay with a Target Kinase
Linear activity over time: Substrate concentration near K m :
Km determination for Kinase 1.
Varying concentrations of peptide
with 10 nM Kinase 1 and 40 M ATP.
Km of peptide for Kinase 1.
Km = 2.5 0.6 M, R 2 = 0.98.
Data from 1-9-01.
Polarization (mP)
250
200
150
100
0uM
0.3uM
1uM
3uM
10uM
30uM
Rate (mP/minute)
4
3
2
1
50
0 25 50 75 100 125 150
Time (minutes)
0
0 10 20 30 40
[Peptide] (M)

Comparison of Potency Assays:
FP versus Radiometric Assays
Staurosporine, a general kinase inhibitor:
Staurosporine inhibition curve.
20 nM Kinase 2, 3 M peptide
substrate, 50 M ATP.
IC 50 from fit = 121 2 nM, R 2 =0.83.
10000
Staurosporine inhibition of Kinase 2.
IC 50 = 135 1 nM, R 2 = 0.996.
Polarization (mP)
150
100
50
Uncorrected cpm
7500
5000
2500
0
0 1 2 3 4
log (Staurosporine (nM))
0
0 1 2 3 4
Log [Staurosporine] (nM)

Radiometric Assay Concept
33
P-ATP ADP
Kinase
33
P
Protein or Peptide Substrate
384-Well Microplate
Advantages:
Direct, sensitive, robust
Disadvantages:
Expensive, hazardous,
heterogeneous

Sensitivity of FP Assay to Interference from Library
Compounds
30000000
Intensity
20000000
10000000
0
384-Well Plate

Reduction of Fluorescent Interference at Rhodamine
Wavelengths
For 8000 test compounds, rhodamine wavelengths give a 4-fold reduction in
the number of interfering compounds (intensity > 2-fold over background)
Plate # Compound Fold improvement
1 1 uM 4.0
2 1 uM 3.1
3 1 uM 2.9
4 1 uM 2.4
5 1 uM 4.0
1 10 uM 9.2
2 10 uM 2.8
3 10 uM 3.2
4 10 uM 2.8
5 10 uM 5.5
4.0 AVG

In Vitro ADME in Drug Discovery
Confirmed hits from HTS
High potency in vitro
and in cells
CYP Assays
Liver Microsomal Assays

Cytochrome P 450 Monooxygenase Assays
• CYPs are the main drug metabolizing enzymes
• Six important isozymes: 1A2, 2C8, 2C9, 2C19, 2D6, and
3A4.
• Net oxidative reaction:
NADPH + O 2 + R-H → NADP + + H 2 O + R-OH
• Homogenous format, high-throughput (up to 1000
compounds per week)
• Enzymes and substrates from BD Gentest

CYP Assay Concept
Dibenzylfluorescein (DBF)
Non-fluorescent
CYP 3A4
X
Inhibitor or
Substrate, e.g.
Ketoconazole
Fluorescein
Fluorescent
Other substrates: coumarins

Example Data from CYP Assay
Inhibition of CYP 1A2 and 3A4
by Ketoconazole
Inhibition (%)
100
CYP 3A4
50
0
CYP 1A2
-1 0 1 2 3 4
Log (Concentration (nM))
CYP 3A4:
IC 50 = 2 1 nM
Hill Slope = 1.1
R 2 = 0.98
CYP 1A2:
IC 50 > 1 M
Hill Slope = 1.6
R 2 = 0.96
Ketoconazole
Mass = 531.4305
CA 65277-42-1
N
O
N
O
N
N
O
H
O
Cl
Cl

Conclusions
• Fluorescence and luminescence methods can be
applied to a variety of drug discovery applications
• Persisting needs: fast, robust and inexpensive
technologies
• Interesting technologies:
microarrays
microfluidics
single-molecule methods

Exelixis, Inc.
www.exelixis.com Nasdaq: EXEL
Acknowledgments:
Stefan Engst and Stephen Quashnick (ADME)
Shaun Nguyen and Pengguang Wu (HTS)
Kirk McMillan, Alison Joly, Scott Robertson, Wentao Zhang, Jing
Wang, Herman Ng, Teecia Kimura, Chris Jaeger, Lillian Mock (NLD)