Patrick
An Introduction to Medicinal Chemistry 3/e
Chapter 6
PROTEINS AS DRUG
TARGETS:
RECEPTOR STRUCTURE &
SIGNAL TRANSDUCTION
Part 5: Case Study
©1
Contents
Part 5: Case Study
6. Case Study - Inhibitors of EGF Receptor Kinase
6.1. The target (4 slides)
6.2. Testing procedures
- In vitro tests (3 slides)
- In vivo tests (2 slides)
- Selectivity tests
6.3. Lead compound – Staurosporine
6.4. Simplification of lead compound (2 slides)
6.5. X-Ray crystallographic studies (2 slides)
6.6. Synthesis of analogues
6.7. Structure Activity Relationships (SAR)
6.8. Drug metabolism (2 slides)
6.9. Further modifications (3 slides)
6.10.Modelling studies on ATP binding (4 slides)
6.11.Model binding studies on Dianilinophthalimides (4 slides)
6.12.Selectivity of action (3 slides)
6.13.Pharmacophore for EGF-receptor kinase inhibitors
6.14.Phenylaminopyrrolopyrimidines (3 slides)
6.15.Pyrazolopyrimidines
[43 slides]
©1
6. Case Study - Inhibitors of EGF Receptor Kinase
6.1 The target
- Epidermal growth factor receptor
- Dual receptor / kinase enzyme role
Extracellular
space
Receptor
Binding site
cell
membrane
Cell
Kinase active site
(closed)
©1
6.1 The target
Overexpression
of erbB1 gene
Excess
receptor
-
KINASE INHIBITOR
Excess sensitivity
to EGF
Excess signal
from receptor
Potential
anticancer
agent
Excess cell growth
and division
Tumours
©1
6.1 The target
H
N
H
Protein
N
N
O
O
N
N
H
O
O
O
H
P
O
P
O
O
O
O
HO
O
H
T yrosine
residue
H
OH OH
ATP
H
P
Protein
HN
O
Mg
Tyrosine kinase
H
N
N
N
N
O
O
N
O
O
H
H
P
H
H
OH OH
ADP
Protein
O
O
P
Protein
HN
O
O
O
O
O
P
O
O
Phosphorylated
tyrosine residue
©1
6.1 The target
Inhibitor Design
Possible versus binding site for tyrosine region
Possible versus binding site for ATP
Inhibitors of the ATP binding site
Aims:
To design a potent but selective inhibitor versus EGF receptor
kinase and not other protein kinases.
©1
6.2 Testing procedures
In vitro tests
Enzyme assay
using kinase portion of the EGF receptor produced by recombinant
DNAtechnology. Allows enzyme studies in solution.
EGF-R
cell
membrane
Cell
Recombinant
DNA
Water
soluble
kinase
©1
6.2 Testing procedures
In vitro tests
Enzyme assay
Test inhibitors by ability to inhibit standard enzyme catalysed reaction
ATP
ADP
OH
OP
Angiotensin II
Angiotensin II
kinase
Assay product
to test inhibition
Inhibitors
• Tests inhibitory activity only and not ability to cross cell membrane
• Most potent inhibitor may be inactive in vivo
©1
6.2 Testing procedures
In vitro tests
Cell assays
• Use cancerous human epithelial cells which are sensitive to EGF for growth
• Measure inhibition by measuring effect on cell growth - blocking kinase
activity blocks cell growth.
• Tests inhibitors for their ability to inhibit kinase and to cross cell membrane
• Assumes that enzyme inhibition is responsible for inhibition of cell growth
Checks
• Assay for tyrosine phosphorylation in cells - should fall with inhibition
• Assay for m-RNA produced by signal transduction - should fall with
inhibition
• Assay fast growing mice cells which divide rapidly in presence of EGF
©1
6.2 Testing procedures
In vivo tests
• Use cancerous human epithelial cells grafted onto mice
• Inject inhibitor into mice
• Inhibition should inhibit tumour growth
• Tests for inhibitory activity + favourable pharmacokinetics
©1
6.2 Testing procedures
Selectivity tests
Similar in vitro and in vivo tests carried out on serine-threonine
kinases and other tyrosine kinases
©1
6.3 Lead compound - Staurosporine
H
N
O
N
N
O
H3C
H3C
O
NH
H3C
•
•
•
•
•
Microbial metabolite
Highly potent kinase inhibitor but no selectivity
Competes with ATP for ATP binding site
Complex molecule with several rings and asymmetric centres
Difficult to synthesise
©1
6.4 Simplification of lead compound
H
N
O
N
N
H 3C
H 3C
O
Simplification
Remove asymmetric
ring
H
N
O
*
*
*
O
*
Simplification
Symmetry
NH
H 3C
Staurosporine
N
H
N
H
H
N
O
N
H
O
N
H
Arcyriaflavin A
• Symmetrical molecule
• Active and selective vs
PKC but not EGF-R
©1
6.4 Simplification of lead compound
maleimide ring
H
N
O
O
Bisindolylmaleimides
PKC selective
N
H
N
H
H
N
O
O
indole ring
Phthalimide
indole ring
Simplification
N
H
Aniline
Simplification
N
H
Aniline
Dianilinophthalimide (CGP 52411)
• Selective inhibitor for EGF
receptor and not other kinases
• Reversal of selectivity
H
N
O
N
H
O
N
H
©1
6.5 X-Ray crystallographic studies
Different shapes implicated in different selectivity
Arcyriaflavin
Planar
O
H
N
Bisindolyl-maleimides
Bowl shaped
O
O
H
N
Dianilino-phthalimides
Propellor shaped
asymmetric
O
N
H
N
H
O
O
NH
N
H
H
N
HN
N
H
©1
6.5 X-Ray crystallographic studies
Propeller conformation relieves steric clashes
Steric
clash
O
H
N
O
H
N
O
HH
HH
Steric
clash
H
Twist
H
H
H
NH
NH
O
HN
Planar
HN
Propeller
shape
©1
6.6 Synthesis of analogues
O
H 2C
TMSCl, NEt3
DMF,
O
O
H 3C
Diels Alder
Toluene
Si (CH3) 3
O
CH3
Anilines
O
O
100 oC
H 2C
O
O
Acetic acid, 120 oC
MeO2C
Si (CH3) 3
C
C
(H3C) 3SiO
OSi(CH3) 3
CO2Me
H 3C
O
CH3
O
O
a) LiOH, MeOH
O
O
O
O
NH3 or formamides
R1
NR2
R 2N
O
140-150 oC
b) (Ac)2O, toluene
R1
O
R
N
R1
R1
NR2
2
R N
R1
R1
NR2
R 2N
©1
6.7 Structure Activity Relationships (SAR)
O
R
N
O
R1
R1
NR2 R 2N
•
•
•
•
•
•
R=H
Activity lost if N is substituted
Aniline aromatic rings essential (activity lost if cyclohexane)
R1=H or F (small groups). Activity drops for Me and lost for Et
R2=H Activity drops if N substituted
Aniline N’s essential. Activity lost if replaced with S
Both carbonyl groups important. Activity drops for lactam
H
N
NH
O
HN
©1
6.7 Structure Activity Relationships (SAR)
Parent Structure: R=R1=R2=H chosen for preclinical trials
IC50 = 0.7 mM
H
N
O
NH
O
HN
CGP 52411
©1
6.8 Drug metabolism
Excretion
H
N
O
O
Glucuronylation
Glucose O
HO
H
N
O
NH
O
NH
Drug
HN
Metabolism
(man,mouse,
rat, dog)
HN
CGP 52411
H
N
O
Metabolism
(monkey)
HO
NH
O
HN
Glucuronylation
OH
Glucose O
Drug O Glucose
Excretion
©1
6.8 Drug metabolism
Introduce F at para position as metabolic blocker
H
N
O
F
Metabolic
blocker
NH
O
HN
CGP 53353
F
Metabolic
blocker
©1
6.9 Further modifications
a) Chain extension
H
N
O
Chain extension
NH
O
HN
Chain extension
CGP58109
Activity drops
©1
6.9 Further modifications
b) Ring extension / expansion
extension
ring
expansion
H
N
O
NH
HN
O
HN
CGP 52411 (IC50 0.7mM)
NH
O
O
NH
remove
polar groups
HN
CGP54690 (IC50 0.12mM)
Inactive in cellular assays
due to polarity
(unable to cross cell membrane)
HN
N
O
NH
HN
CGP57198 (IC50 0.18mM)
Active in vitro and in vivo
©1
6.9 Further modifications
c) Simplification
H
N
O
NH
O
HN
CGP52411
Simplification
H
N
O
NH
O
OH
CGP58522
Similar activity in enzyme assay
Inactive in cellular assay
©1
6.10 Modelling studies on ATP binding
• No crystal structure for EGF- receptor available
• Make a model active site based on structure of an
analogous protein which has been crystallised
• Cyclic AMP dependant protein kinase used as template
©1
6.10 Modelling studies on ATP binding
Cyclic AMP dependant protein kinase
+ Mg + ATP + Inhibitor (bound at
substrate site)
Crystallise
Crystals
X-Ray Crystallography
Structure of protein /
inhibitor / ATP complex
Molecular modelling
Identify active site
and binding interactions
for ATP
©1
6.10 Modelling studies on ATP binding
• ATP bound into a cleft in the enzyme with adenine portion
buried deep close to hydrophobic region.
• Ribose and phosphate extend outwards towards opening of
cleft
• Identify binding interactions (measure distances between
atoms of ATP and complementary atoms in binding site to see
if they are correct distance for binding)
• Construct model ATP binding site for EGF-receptor kinase
by replacing amino acid’s of cyclic AMP dependent protein
kinase for those present in EGF receptor kinase
©1
6.10 Modelling studies on ATP binding
Gln767
HN
H-bond interactions
empty
pocket
Thr766
H2NOC
H
N
O
O
Leu768
H3C
Met769
H
H3C
H
O
N
H
H
N
O
N
S
H3C
H
N
1
6
N
N
O
O
N
O
O
O
P
O
O
O
P
O
O
H
1N
P
O
H
H
OH
H
OH
is a H bond acceptor
6-NH2 is a H-bond donor
Ribose forms H-bonds to Glu in ribose pocket
'ribose' pocket
©1
O
6.11 Model binding studies on Dianilinophthalimides
Gln767
HN
empty
pocket
H-bond interaction
Thr766
H2NOC
H
N
O
Leu768
H3C
Met769
H
H3C
O
H
O
N
O
H
N
O
N
S
H
NH
O
H3C
O
HN
'ribose' pocket
©1
6.11 Model binding studies on Dianilinophthalimides
• Both imide carbonyls act as H-bond acceptors (disrupted if
carbonyl reduced)
• Imide NH acts as H bond donor (disrupted if N is substituted)
• Aniline aromatic ring fits small tight ribose pocket
• Substitution on aromatic ring or chain extension prevents
aromatic ring fitting pocket
• Bisindolylmaleimides form H-bond interactions but cannot fit
aromatic ring into ribose pocket.
• Implies ribose pocket interaction is crucial for selectivity
©1
6.11 Model binding studies on Dianilinophthalimides
Gln767
HN
empty
pocket
Thr766
H-bond interaction
H2NOC
H
N
O
O
Leu768
H3C
Met769
H
H3C
H
O
N
O
HN
H
O
N
S
H3C
O
H
N
NH
O
HN
'ribose' pocket
©1
6.11 Model binding studies on Dianilinophthalimides
Gln767
HN
empty
pocket
Thr766
H-bond interaction
H2NOC
H
N
O
O
Leu768
H3C
Met769
H
H3C
H
O
O
N
H
N
O
N
S
H
NH
O
H3C
NH
O
'ribose' pocket
©1
6.12 Selectivity of action
POSERS ?
• Ribose pocket normally accepts a polar ribose so why can it
accept an aromatic ring?
• Why can’t other kinases bind dianilinophthalimides in the same
manner?
©1
6.12 Selectivity of action
Amino Acids present in the ribose pocket
Hydrophobic
Protein Kinase A
EGF Receptor Kinase
Hydrophilic
Leu,Gly,Val,Leu
Glu,Glu,Asn,Thr
Leu,Gly,Val,Leu,Cys
Arg,Asn,Thr
©1
6.12 Selectivity of action
• Ribose pocket is more hydrophobic in EGF-receptor kinase
• Cys can stabilise and bind to aromatic rings (S-Ar interaction)
Gln767
HN
empty
pocket
Thr766
H2NOC
H
N
O
H
H3C
Leu768
H3C
Met769
S
O
H
O
N
O
H
N
O
N
H
NH
O
H3C
O
HN
H
S
'ribose' pocket
• Stabilisation by S-Ar interaction not present in other kinases
• Leads to selectivity of action
©1
6.13 Pharmacophore for EGF-receptor kinase inhibitors
O
HBD
HBD
H
N
HBA
NH
O
HBA
HN
Ar
Pharmacophore
Ar
• Pharmacophore allows identification of other potential inhibitors
• Search databases for structures containing same pharmacophore
• Can rationalise activity of different structural classes of inhibitor
©1
6.14 Phenylaminopyrrolopyrimidines
CGP 59326 - Two possible binding modes for H-bonding
Cl
HBD
H
HBD
HBA
N
H
N
H
N
N
N
HBA
N
Ar
N
N
H
Cl
Mode I
Mode II
Only mode II tallies with pharmacophore and explains activity
and selectivity
©1
6.14 Phenylaminopyrrolopyrimidines
Cl
O
empty
pocket
empty
pocket
O
H
H
N
N
H
N
N
N
CGP59326
H
N
N
N
H
N
H
N
H
CGP59326
H
S
S
Cl
'ribose' pocket
Binding Mode I like ATP
(not favoured)
'ribose' pocket
Binding mode II (favoured)
Illustrates dangers in comparing structures and assuming similar
interactions (e.g. comparing CGP59326 with ATP)
©1
6.14 Phenylaminopyrrolopyrimidines
HBD
H
HBD
N
HBA
H
N
HBA
N
N
Ar
Ar
Cl
©1
6.15 Pyrazolopyrimidines
i) Lead compounds
Cl
NH2
N
NH2
N
N
N
N
N
H 2N
(I) EC50 0.80mM
N
N
H
(II) EC50 0.22mM
• Both structures are selective EGF-receptor kinase inhibitors
• Both structures belong to same class of compounds
• Docking experiments reveal different binding modes to obey
pharmacophore
©1
6.15 Pyrazolopyrimidines
ii) Structure I
empty
pocket
O
H
Extra binding
interactions
HBD
H
HBA
N
H
N
H
N
N
H
Structure
I
N
N
N
N
N
N
N
H
S
Ar
'ribose' pocket
©1
6.15 Pyrazolopyrimidines
ii) Structure I
NH2
NH2
N
N
N
N
N
N
(I) EC50 0.80mM
N
N
(III) EC50 2.7mM
©1
6.15 Pyrazolopyrimidines
iii) Structure II
• Cannot bind in same mode since no fit to ribose pocket
• Binds in similar mode to phenylaminopyrrolopyrimidines
empty
pocket
Cl
O
H
N
N
N
NH
Structure
II
N
H
N
H 2N
H
S
unoccupied
'ribose' pocket
©1
6.15 Pyrazolopyrimidines
Extra
H-bonding
interaction
iv) Drug design on structure II
Cl
Cl
Cl
OH
HBD
HBD
H
N
N
H
H
N
N
NH
N
N
NH
HBA
H
N
N
NH
NH
HBA
N
Simplification
N
H 2N
(II)
EC50 0.22mM
(remove extra
functional group)
Extension
N
N
(add aromatic
ring for ribose
pocket)
(IV)
EC50 0.16mM
Activity increases
N
NH
Extension
N
N
NH
N
Ar
Cl
Cl
(V)
EC50 0.033mM
Activity increases
Ar fits ribose pocket
(VI)
EC50 0.001mM
Activity increases
• Upper binding pocket is larger than ribose pocket allowing greater
variation of substituents on the ‘upper’ aromatic ring
©1