Transcript Docking and Cancer Drug Design

Docking and Cancer
Drug Design
Jennie Bever
Where to start?

Ligand
(analog-based)

Target
(structure-based)
Goals of Docking
1)
2)
3)
Characterize binding site
- make an image of binding site with
interaction points
Orient ligand into binding site
- grid search
- descriptor mapping
- energy search
Evaluate strength of the interaction
DG bind= DG complex – (DGligand +
DGtarget)
Structure-based design of selective and potent
G quadraplex-mediated telomerase inhibitors
Neidle et al. 2001 PNAS 98(9) 4844-4849
NMR-derived G4
structure
PDB entry 143D
Repeat of
d[AG3(TTAG3)3]
Telomerase
Activated in 80-90% human tumors
 Keeps cancer cells’ telomeres stable
length

Telomerase as a Drug Target

Telomerase needs single stranded 3’DNA
primer end
– To hybridize with telomerase RNA template

Drug Target:
– Use higher order structure to prevent
telomeres from unfolding
No ss primer
Block telomerase
G quadraplex
Telomere repeat d[AG3(TTAG3)3]
G repeats form H bonds quadraplex
Stabilize Quadraplex– How?

Extended planar
aromatic
chromophore

Nearly all known
quadraplex ligands
share this structure
Problems with known ligands
Affinity for duplex DNA comparable to
quadraplex binding
 Low potency for telomerase inhibition
IC50=2-5mM
 Level of acute cytotoxicity too high

From Structure to Ligand
1) Energy minimize
structures of target
and ligand
Molecular
dynamics
simulation
 time averaged
structures
3) Create
pseudointercalation
binding site
Docking

Dock ligand into
pseudo-intercalation
site
– Manual, automatic,
and flexible ligand
docking

Energy minimize to
determine DG complex
 Determine DGligand
_=interaction energy
of ligand with
surroundings when
explicitly solvated

Visual inspection of
compound 1
docking
Compounds 3 and 4
Similarly docked
Docking results

Attempts at docking compound 2 were
unsuccessful
– Distorted the quadraplex structure due to
bulky side chains
Ligand/quadraplex interactions in
solution: SPR Rate constant determination
Ligand/quadraplex interactions in solution:
SPR Equilibrium constant determination
Will these molecules work?
Trap Assays and Growth Inhibition (Cytotoxicity)
Docking-based Development of Purine-like
Inhibitors of Cyclin-Dependent Kinase-2
Koca et al. J. Med. Chem. 2000, 43, 2506-2513
Cdk2 structure
complexed with ATP
Cdk2 regulates G1/S
transition
Cdk deregulation in
many tumors and tumor
cell lines

Xray structures of
3 cdk2 complexes
available
– ATP, olomoucine,
roscovitine

All bind in same
cleft, but with
different
orientations
Docking of 2,6,9-purine derivatives
into cdk2

Minimize structures (target and ligands)
 Represent active site as spheres
 Calculate partial atomic charges from force
fields
 Rigid docking
 Pick receptor
structure to use
 Rigid and Flexible
docking with
proposed ligands
Interaction energy and Inhibitory
potency (IC50) for 2,6,9-purine
derivatives on cdk1 and cdk2
They say good correlation for cdk2
interaction energy can be used to predict activity
Interaction energy versus IC50
For cdk1, rank of
activities also closely
bounded to rank
of interaction energies
Searching for New Inhibitors
Role of Electrostatic and van der Waals
contribitions in binding ligand to active site
H bonds predicted by rigid docking
agree with experimental data
How well does docking compare to
crystal structure?
2 binding modes found
 roscovatine-like—agreed very well with
crystal structure
 ATP-like—differences in the orientation
of N9 isopropyl and C2 side chain shape
Fix C2 and isopropyl to be rigid—repeat docking
Paper Conclusions:
Rigid docking useful for activity
predictions
 Relationship between cdk1 activity and
cdk2 interaction energy
 Use these experiments to design new
inhibitors of cdk2 and cdk1

My Conclusions
Docking is now and will be an integral
tool in drug design in the future
 Play with some of the modeling
software—it’s cool!—and helps to
understand how simulations are done

 Thanks
to Dr. Bourne!