Conformational Flexibility and Ligand DesignConformational Flexibility and Ligand Design

Dr. Eamonn F. HealyDr. Eamonn F. HealyProfessor of ChemistryProfessor of Chemistry

St. Edward’s University, Austin, Tx. St. Edward’s University, Austin, Tx.

Our research focuses on the design of structure-activity probes to elucidate enzymatic activity. The interdisciplinary approach includes molecular modeling for the simulation of inhibitor binding , overexpression of wild-type and mutant target proteins and in vitro assays of enzymatic activity and inhibition. Our targets include HIV-1 integrase, the c-Kit and src-abl proteins , and the metalloproteinases associated with CXCL16 shedding. Large flexible ligands and conformationally mobile proteins present two distinct, but related challenges. Lessons learned from investigating these target systems will be presented.

A process for drug design which bases the design of the drug upon the structure of its protein target.

1. Structural mapping of the receptor (protein, P) active site

2. Identification of ligands (L) of complementary shape and appropriate functionality

3. Docking of the ligand to the receptor site - predicting a range of PL complexes with different GPL values

4. Scoring i.e. ranking GPL and correlating with experimentally determined properties such as IC50 values

PROTEIN

1. Structural mapping of the receptor active site

Crystal structure available for Integrase but :I. Limitations of crystal structure:

Often only one domain Membrane or DNA attachment sites

usually not shown Crystal structure vs. physiologically

active structureII. Position of hydrogens undetermined III. Residues missing or ill-definedIV. Protonation of His undeterminedV. SolvationVI.Conformational Dynamics

LIGAND

2. Identification of ligands (L) of complementary shape and appropriate functionality

Crystal structure often available for Inhibitor bound to catalytic core but :

I. Position of hydrogens undeterminedII. Tautomeric structures possibleIII. Influence of pHIV. Need to limit conformational flexibility based

on experimental and theoretical crteria

Fixed and planar

Based on HF/6-31G* calculationsLimited to +/- 45 degrees

DOCKING

3. Docking of the ligand to the receptor site - predicting a range of PL complexes with different GPL values

The prediction of the ligand conformation and orientation within a targeted binding site

involves:I. Positioning ligand and evaluating

quality of bindingII. Refining ligand positionIII. Energy minimization (electrostatic,

steric, strain and h-bond)

SCORING

4. Scoring i.e. ranking GPL and correlating with experimentally determined properties such as IC50 values

The prediction of the optimum ligand conformation

and orientation within a targeted

binding site involves:I. Posing : Determining the fit

of the ligandII. Conformational SearchingIII. Scoring and Ranking

Predicted Gbind

(kcal mol-1)

-12.7

Predicted Gbind

(kJ mol-1)

-53.0

Predicted Ki

(M)5.10-10

pKi 9.3

Expected IC50

< nanomolar

A B

DC

(A) 6.7 degree tilt on the N-lobe towards the C-lobe for the conversion of inactive c-Kit (yellow) to the active form (blue) ; (B) The induced fit of imanitib

mesylate to c-Kit, shown as a superimposition of the inactive form in the absence of inhibitor (yellow) and c-Kit+inhibitor ( gray); (C) , (D) Key

structural elements of inactive (C) and active (D) conformations

FEBS Lett. 2009, 583, 2899-2906.

Surface representation showing Ile653 side chain in red. Electrostatic map of bound inhibitor

Solvent accessible surfaces for Ile653 side chain + Aryl ring of inhibitor.

Trajectory analysis for solvent-exposed intra-helix backbone hydrogen bond distance between Leu644 and

Gly648 of the C-helix for a 1 ns molecular dynamics run for c-Kit with ( )nd without ( )bound ligand .

Final conformation for the 1ns MD run for c-Kit with bound ligand.

Final conformation for the 1ns MD run for c-Kit without bound ligand.

N

NO

H

H

HO

Thr670

O

OGlu640

HN

H

H

Lys623

1

34

56 7

89

10

2

A B

C D

(A) Inhibition of c-Kit through the binding of 9-hydroxy ellipticine. (B) Inhibition of c-Abl through the binding of 9-hydroxy ellipticine. (C) Docked structures for 9-hydroxy ellipticene in blue,9-methoxyellipticine in green and

ADP in gray with 1OPK. (D) Postion of the 9-methoxyellipticine ligand in relation to the backbone hydrogen bond between Phe401 and Leu403.

Alignments

Repositioning of DFG motif through movement of activation loop in going from inactive (yellow)

to active (blue) c-Kit

Dominant Patterns of Drug Resistance for certain Tyrosine Kinases

DCQ acids; DCT acids

DKAs

Quinolone derived

PDP SQL

HIV-1 Integrase Inhibitors

OO

OH

OH

HO

HO

COOHHOOC

OO

L-CA

Cl

NH

HO

NH

NH

NN

O

CITEP ( a DKA)

J Mol. Graph. Model. 2009, 27 , 14.

L-Chicoric Acid and HIV-1 Integrase

Isomer Erel (kcal mol-1) % Population (T=298 K)AM1 HF/6-31G** MP2/6-31G** AM1 HF MP2

cis-cis / syn-syn 0.0 0.0 0.0 75.6 77.3 74.9

cis-trans / syn-syn 1.2 1.2 1.1 24.2 21.6 23.9

trans-trans / syn-syn 3.7 2.6 2.5 0.2 1.1 1.2cis-cis / syn-anti 6.0 5.5 10.9 <1.10-4 <2.10-4 <2.10-6Ligand IN Protein Cluster

Occa.Gbind (kcal mol-1) Ki (M) IC50(relative)b h-bonded residues

L-CA (s-cis /s-cis)

1QS4 44 -8.1 1.1 1 D116,Q148, K156, K159

L-CA (s-cis /s-trans)

1QS4 29 -7.6 2.6 1 T66, H67,D92, Q148, K156, K159

L-CA (s-cis /s-cis)

Q148A 27 -6.2 31.7 20 D116, A148, K156

L-CA tetraacetylated (s-cis /s-cis)

1QS4 33 -6.6 14.0 9 T66, H67, Q148, K156, K159

Conformer energies, and predicted conformer populations, for the various L-CA conformer geometries. Results of 100 independent docking runs for the ligands L-CA and its tetraacetylated derivative with proteins 1QS4 and the mutant Q148A.

OO

OH

OH

HO

HO

COOHHOOC

OO

L-CA

Cl

NH

HO

NH

NH

NN

O

CITEP ( a DKA)

O

R

H

H

X

R

H

H

O

X

s-trans s-cis

RO

OR'

RO

OR'

syn anti

Top-ranked binding modes for s-cis /s-cis L-CA (top left), s-cis /s-trans L-CA (top right), tetracetylated derivative of L-CA (bottom

left) and a surface image of the inhibitor binding pocket with bounds-cis /s-cis L-CA overlaid with the experimentally observed

5CITEP inhibitor (bottom right).

Hydrogen bonded interactions and electrostatic map for the top-ranked s-cis /s-

cis L-CA solution

Druggability and Drug Resistance

ADAM 17 and ADAM 10

A zinc-binding group provides high affinity but low selectivity

Major determinant of potency and selectivity.

Long chains can provide selectivity.

Wide range of substituents tolerated

and are determinants of potency. Steric bulk beneficial for oral

availability.

Wide range of substituents

tolerated. Charged polar

groups influence bilary excretion.

ADAM Ligand Design

S1’-S3

’ Channel

S1'

S1

E

B

S2'

A TAPI-2

C R=CH3 (TMI-1)D R=CH2NH2

S1' S1'

S3'

S3'

S3'

S1'

S3'

OHNH

O

O

NH O

NH

NH

NH2

O OHNH

O

N

S

O

O

O

OHNH

O

N S

R

O

S

O

O

N

NH

OH

OO

SOO

O

SO

S

1'

S1

E

B

S2'

A TAPI-2

C R=CH3 (TMI-1)D R=CH2NH2

S1' S1'

S3'

S3'

S3'

S1'

S3'

OHNH

O

O

NH O

NH

NH

NH2

O OHNH

O

N

S

O

O

O

OHNH

O

N S

R

O

S

O

O

N

NH

OH

OO

SOO

O

SO

Acetylenic Sufonamide Inhibitor Design

B

S2'

A TAPI-2

S1' S1'

S3'S3'

OHNH

O

O

NH O

NH

NH

NH2

O OHNH

O

N

S

O

O

O

OHNH

O

N S

R

O

S

O

O

N

NH

OH

OO

SOO

O

SO

B

S2'

A TAPI-2

S1' S1'

S3'S3'

OHNH

O

O

NH O

NH

NH

NH2

O OHNH

O

N

S

O

O

O

OHNH

O

N S

R

O

S

O

O

N

NH

OH

OO

SOO

O

SO

TACE: Potency

S1'

C R=CH3 (TMI-1)D R=CH2NH2

S3'

OHNH

O

O

NH O

NH

NH

NH2

O OHNH

O

N

S

O

O

O

OHNH

O

N S

R

O

S

O

O

N

NH

OH

OO

SOO

O

SO

S

1'

C R=CH3 (TMI-1)D R=CH2NH2

S3'

OHNH

O

O

NH O

NH

NH

NH2

O OHNH

O

N

S

O

O

O

OHNH

O

N S

R

O

S

O

O

N

NH

OH

OO

SOO

O

SO

TACE: Potency

TACE: Potency

S2'

A TAPI-2

S3'

OHNH

O

O

NH O

NH

NH

NH2

O OHNH

O

N

S

O

O

O

OHNH

O

N S

R

O

S

O

O

N

NH

OH

OO

SOO

O

SO

S2'

A TAPI-2

S3'

OHNH

O

O

NH O

NH

NH

NH2

O OHNH

O

N

S

O

O

O

OHNH

O

N S

R

O

S

O

O

N

NH

OH

OO

SOO

O

SO

ADAM17 vs ADAM10: Selectivity

S1'

C R=CH3 (TMI-1)

S3'

OHNH

O

O

NH O

NH

NH

NH2

O OHNH

O

N

S

O

O

O

OHNH

O

N S

R

O

S

O

O

N

NH

OH

OO

SOO

O

SO

S

1'

C R=CH3 (TMI-1)

S3'

OHNH

O

O

NH O

NH

NH

NH2

O OHNH

O

N

S

O

O

O

OHNH

O

N S

R

O

S

O

O

N

NH

OH

OO

SOO

O

SO

ADAM17 vs ADAM10: Selectivity

S1

S2

S1

E

S3'

S1'

OHNH

O

O

NH O

NH

NH

NH2

O OHNH

O

N

S

O

O

O

OHNH

O

N S

RO

S

O

O

N

NH

OH

OO

SOO

O

SO

S1

E

S3'

S1'

OHNH

O

O

NH O

NH

NH

NH2

O OHNH

O

N

S

O

O

O

OHNH

O

N S

RO

S

O

O

N

NH

OH

OO

SOO

O

SO

Future Ligand Design

S2

S1

S3'

S1'

O

O

OS

O

NO

NSOH

S

O

N

O

NH

OH

S

2

S1

S3'

S1'

O

O

OS

O

NO

NSOH

S

O

N

O

NH

OH

Future Ligand Design: Electrostatic map of S2 binding pocket

S1 S1’ S3

’ S2 S2’

loop

Protein Alignment for ADAM17 and ADAM10