Transcript Drug discovery for neglected tropical diseases

Drug discovery for
neglected tropical diseases
Greg Crowther & Wes Van Voorhis
Department of Medicine
University of Washington
Protein-based
(target-based) projects
1. Prioritization of drug targets
• TDRtargets.org
2. Structural genomics of pathogen proteins
• MSGPP.org
• SSGCID.org
3. “Piggy-back” approach to target-based drug design
• protein farnesyltransferase
• glycogen synthase kinase
4. Identifying targets of cell-active compounds
• thermal melt assays
• enzyme activity assays
Fluorescence
PDB: 1tqx
Temperature
Prioritization of drug targets
TDRtargets.org is a database of protein data relevant to drug discovery research.
Other team leaders: Fernán Agüero (U. of San Martin), Matt Berriman (Sanger Institute),
Stuart Ralph (U. of Melbourne), David Roos (U. of Pennsylvania), Sam Sia (Columbia U.).
TDRtargets.org allows users to
Prioritization
“zoom in” on protein
targets of of drug targets
particular interest …
… and create genome-wide rankings of targets
Prioritization
of drug
based
on the users’
own targets
criteria.
Structural genomics of
pathogen proteins
Medical Structural Genomics of Pathogenic Protozoa
(MSGPP.org)
• selection of potential drug targets
• expression, crystallization, 3D structure determination,
ligand binding
• organisms: Plasmodium, T. brucei, T. cruzi, Leishmania,
T. gondii, E. histolytica, Giardia, Cryptosporidium
• collaborators: Fred Buckner, Erkang Fan, Wim Hol,
Ethan Merritt, Christophe Verlinde (all UW)
Seattle Structural Genomics Center for Infectious Disease
(SSGCID.org)
• like MSGPP, but focused on biodefense-related
pathogens and (re-)emerging diseases
• other team leaders: Peter Myler (SBRI), Lance Stewart
(deCODE), Gabriele Varani (UW), Garry Buchko (Battelle)
“Piggy-back” approach to
target-based drug design
Drug development is difficult and expensive when starting from scratch.
Look for promising drug targets where a lot of development has already
been performed.
See if the existing drugs show promise against “our” diseases, then
piggy-back onto existing efforts.
Image: NASA / Tom Tschida
Collaborators: Fred Buckner, Mike Gelb, K.K. Ojo, Christophe Verlinde (all UW).
“Piggy-back” approach to
target-based drug design
Protein Farnesyltransferase
• adds farnesyl (C15H25) groups to proteins (important for localization,
etc.)
• target for oncology -- in Phase III trials
• inhibitors cure rodent malaria (PubMed ID: 17606674)
• current work: optimizing pharmacokinetics
Glycogen Synthase Kinase
• phosphorylates glycogen synthase and signaling-related proteins
• target for mania, Alzheimer’s, diabetes
• inhibitors kill T. brucei (PubMed ID: 18644955)
• current work: testing in animal models
Identifying targets of
cell-active compounds
• Thousands of antimalaria compounds have been identified in screens of
chemical libraries.
• Their subcellular targets are unknown, making optimization difficult. How do
you improve activity against the parasite without hurting the host?
• Try to identify targets of some of compounds in order to facilitate optimization. Our
(high-throughput) approaches: thermal melts and enzyme activity assays.
• Collaborators: Roger Wiegand et al. (Broad Institute); Kip Guy (St. Jude); Kelli
Kuhen, Richard Glynne, Achim Brinker et al. (Genomics Institute of Novartis
Research Foundation).
Images: trampledunderfoot.co.uk; tulane.edu/~wiser; clipartof.com
Identifying targets of
cell-active compounds
Thermal melt:
Heat protein, watch it unfold.
Solvent-accessible
hydrophobic surface area
(measured with fluorescent dye)
Temperature
Adaptation of a figure by Martin C. Stumpe and Helmut Grubmuller (www.mpibpc.mpg.de).
Identifying targets of
cell-active compounds
Fluorescence
Melting temperature (Tm) reflects protein stability
less stable
(no ligand)
more stable
(with ligand)
DTm
Temperature
A compound that targets a particular protein should bind to it
and stabilize it, shifting the melting curve and Tm to the right.
Identifying targets of
cell-active compounds
Preliminary validation of thermal melt approach
DHODH inhibitor 1
DHODH inhibitor 2
Tm of DHODH (degrees C)
65
63
DHODH inhibitor 3
HSP90 inhibitor 1
61
HSP90 inhibitor 2
59
57
55
53
51
49
47
45
0
10
20
30
40
50
60
70
80
90
Compound concentration (uM)
• DHODH inhibitors cause dose-dependent increases in DHODH’s Tm
• Negative controls: HSP90 inhibitors don’t change DHODH’s Tm
100
Identifying targets of
cell-active compounds
Thermal melt assays for target identification
Advantages:
Limitations:
• can be applied to most Plasmodium
proteins*
• false positives
• a standard buffer (100 mM HEPES,
150 mM KCl, pH 7.5) works well for
many proteins*
ligands bind to protein and
raise its Tm but don’t inhibit it
• false negatives
not all substrates increase Tm;
not all inhibitors do either?
*Crowther et al. (2009), J. Biomol. Screen 14: in press.
Identifying targets of
cell-active compounds
Enzyme activity assays for target identification
Advantages:
Limitations:
• direct readout of target inhibition
• useless for noncatalytic proteins
• published info on Km’s, optimal
buffers, etc. is available for many
enzymes
• radioactivity, absorbance at UV
wavelengths, HPLC, etc. are inappropriate
for high-throughput screening
• substrates may not be available
• each enzyme is different
Identifying targets of
cell-active compounds
Examples of high-throughput enzyme activity assays
dUTPase (PF11_0282)
Reaction: dUTP => dUMP + PPi
Coupling reaction (Pyrophosphatase): PPi => 2Pi
Detect ↑Pi via malachite green kit (absorbance at 620 nm).
Glycerol-3-Phosphate Dehydrogenase (PFL0780w)
Reaction: glycerol-3-phosphate + NAD+ => dihydroxyacetone phosphate + NADH
Coupling reaction (Diaphorase): resazurin + NADH => resorufin + NAD+
Detect ↑resorufin via fluorescence (excite at 560 nm, emit at 590 nm).
OMP Decarboxylase (PF10_0225)
Reaction: OMP => UMP + CO2
Coupling reaction (CMP Kinase): UMP + ATP => UDP + ADP
Detect ↓ATP via Kinase-Glo luminescence, or detect ↑ADP via fluorescence polarization.
S-Adenosylhomocysteine Hydrolase (PFE1050w)
Reaction: S-adenosylhomocysteine => homocysteine + adenosine
Coupling reaction (Adenosine Deaminase): adenosine => inosine
Detect ↑homocysteine –SH via ThioGlo fluorescence (excite at 379 nm, emit at 513 nm).
Identifying targets of
cell-active compounds
Thermal melt and enzyme activity assays
in the context of drug discovery
Anti-parasitic compounds:
inhibit P. falciparum +
promising chemical properties
Test whether the protein is a target of the
compound in parasites:
a) substrate buildup?
b) select resistance = mutations in target?
c) overexpress in Plasmodium = increase
in ED50?
Screen for proteincompound associations:
thermal melts,
enzyme activity assays
High-Throughput Screen:
validated proteins that
need new scaffolds
Hit optimization:
directed by target
Summary
• Target-based drug discovery has not yet led to many new drugs for
neglected tropical diseases.
• Nevertheless there are reasons for optimism.
- New genomic/bioinformatic data (e.g., via TDRtargets.org):
more possible protein targets
better prioritization of targets
- New biochemical methods (e.g., thermal melts):
more “screenable” proteins
- New 3D protein structures (e.g., via MSGPP and SSGCID):
more structure-based drug design
- New private-sector involvement (e.g., Novartis):
better compound libraries
more screening horsepower
more piggy-backing opportunities
Questions? Comments?
If we’re out of time, feel free
to send email to
[email protected]
or talk to me tonight at dinner.