Transcript Library Design for Leadlike Compounds: A Historical

Library Design for Leadlike
Compounds:
A Historical Perspective
Tudor I. Oprea
EST Lead Informatics
What Is A Lead?
• Many compounds are active, but not all actives
are leads
• Leads have to meet project dependent criteria:
biological activity validated, both in primary
and secondary screens, against known
targets, for a series of compounds (when
available)
• must be patentable, and display good initial DMPK
profile
Medicinal Chemistry / AstraZeneca Mölnal
Tudor I. Oprea
Common Sources for Leads in Drug
Discovery
High-affinity leads
affinity << 0.1 M
MW >> 350
CLogP < 3
Leadlike leads
affinity > 0.1 M
MW < 350
CLogP < 3
DRUG
Druglike leads
affinity > 0.1 M
MW > 350
CLogP > 3
• One needs to distinguish “leadlike” leads from other sources of
lead structures, e.g., natural products that are high-affinity
compounds (NPY or taxol are leads!) or from “druglike” leads
that are marketed structures (e.g., propranolol)
Medicinal Chemistry / AstraZeneca Mölnal
Tudor I. Oprea
Is There A Leadlike Space?
• There is a general consensus that lead discovery is
an essential goal that precludes drug discovery
• For the time being, the only way to analyse the
nature of the “leadlike” space is to examine the
structures that, historically, were leads.
• The problem is: can these structures provide an
objective link between lead-space and drug-space?
• Can we define how these two spaces overlap?
Medicinal Chemistry / AstraZeneca Mölnal
Tudor I. Oprea
Is There A Leadlike Space? (2)
• Initial goal of the retrospective leadlike analysis: gather
as much information as possible about leads, i.e., which
drug has been developed from which lead
• Only a few authors describe the chemical structure of
the lead compound that was used to derive a given drug.
• Lead structures are often disclosed in a series (SAR),
making it difficult to pinpoint at a given compound.
• Furthermore, a drug can have 1 or more leads
• a lead can be a drug
• a lead can lead to several drugs.
Medicinal Chemistry / AstraZeneca Mölnal
Tudor I. Oprea
Mifepristone (RU486):
Drug with multiple leads:
N
O
• Mifepristone originates from
progesterone and RU2323
H
H
O
O
RU486
O
H
H
H
H
O
H
Progesterone
Medicinal Chemistry / AstraZeneca Mölnal
Tudor I. Oprea
RU2323
O
Cocaine:
N
Drugs that are leads:
O
H
O
• Cocaine (local anesthetic) was
the lead for procaine (local
anesthetic) which was, in turn,
the lead for procainamide
(antiarrhythmic)
H
O
O
Cocaine
N
N
N
O
Procainamide
Medicinal Chemistry / AstraZeneca Mölnal
Tudor I. Oprea
O
N
N
O
Procaine
Diltiazem:
Classes of leads that
have generate drugs with
diverse medical
applications:
benzodiazepines are a well
described class of leadlike
structures that resulted in
several drugs, ranging from
CNS agents (hypnotics,
anxiolytics) to calcium
channel blockers and ACE
inhibitors
S
O
N
H
N
Thiazesim
N
O
O
O
N
Cl
N
O
N
Oxazepam
Tudor I. Oprea
H
S
26 launched BZDs in MDDR
Medicinal Chemistry / AstraZeneca Mölnal
O
O
Diltiazem
Av. 302
Md. 315
16
14
12
10
8
6
4
2
0
Av. 391
Md. 384
20
15
Frequency
Frequency
M.W. & CMR
10
5
0
129 182 234 286 339 391 444 496 548
MW
156
208
260
312
364
416
468
572
520
624
14
Av. 8.3
Md. 8.4
12
10
Frequency
Frequency
10
Av. 10.6
Md. 10.2
12
8
6
4
8
6
4
2
2
0
3
4
5
6
7
8
9
10 11 12 13 14 15 16
62 “pure” leads
Medicinal Chemistry / AstraZeneca Mölnal
Tudor I. Oprea
CMR
0
3
4
5
6
7
8
9
10
11
12
13
75 “pure” drugs
14
15
16
RNG & RTB
Av. 2.37
Md. 2
25
18
16
Frequency
Frequency
20
15
10
Av. 3.36
Md. 3
20
14
12
10
8
6
5
4
2
0
0
1
2
3
4
5
RNG
6
0
0
10
7
6
5
4
3
2
3
4
5
6
7
Av. 7.51
Md. 6
10
Frequency
Frequency
8
2
12
Av. 5.48
Md. 4
9
1
8
6
4
2
RTB
1
0
0
1
2
3
4
5
6
7
62 “pure” leads
Medicinal Chemistry / AstraZeneca Mölnal
Tudor I. Oprea
8
9
10
More
0
0
1
2
3
4
5
6
7
8
9
10
75 “pure” drugs
11
12
More
CLogP & LogD7.4
Frequency
6
5
CLogP
12
10
Frequency
7
Av. 1.57
Md. 2.11
n=58
4
3
2
8
6
4
2
1
0
0
12
-2.3 -1.8 -1.3 -0.8 -0.3 0.3 0.8 1.3 1.8 2.3 2.8 3.3 3.8 4.3 4.8 5.3 5.8
Av. 0.72
Md. 0.72
n=57
LogD7.4
14
12
8
Frequency
Frequency
10
Av. 2.73
Md. 2.54
n=70
6
4
2
10
Av. 1.69
Md. 1.69
n=71
8
6
4
2
0
0
- 7.5 - 6.5 - 5.5 - 4.5 - 3.5 - 2.5 - 1.5 - 0.5 0.5
1.5
62 “pure” leads
Medicinal Chemistry / AstraZeneca Mölnal
Tudor I. Oprea
2.5
3.5
4.5
5.5
- 7.5 - 6.5 - 5.5 - 4.5 - 3.5 - 2.5 - 1.5 - 0.5 0.5 1.5
75 “pure” drugs
2.5 3.5 4.5
5.5
Av. 2.3
Md. 2
16
14
12
10
8
6
4
2
0
Av. 2.1
Md. 2
25
20
15
10
5
0
0
1
2
3
4
5
6
7
8
9
10
Av. 4.8
Md. 4
14
12
10
11
0
HAC
1
2
3
4
5
6
7
8
9
Av. 5.8
Md. 5
14
12
10
8
6
Frequency
Frequency
HDO
Frequency
Frequency
Donors & Acceptors
4
2
8
6
4
2
0
0
0
1
2
3
4
5
6
7
8
9 10 11 12 More
62 “pure” leads
Medicinal Chemistry / AstraZeneca Mölnal
Tudor I. Oprea
0
1
2
3
4
5
6
7
8
75 “pure” drugs
9
10
11
12
More
Druglike Score
Av. 0.66
Md. 0.66
20
Frequency
Frequency
15
10
5
15
10
5
0
0
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
DFPS
0.4
Av. 0.45
Md. 0.54
14
12
10
8
6
4
0.5
0.6
0.7
0.8
0.9
1.0
Av. 0.54
Md. 0.66
20
15
Frequency
Frequency
Av. 0.81
Md. 0.81
20
10
5
2
0
0.4
0.5
0.6
0.7
0.8
0.9
62 “pure” leads
Medicinal Chemistry / AstraZeneca Mölnal
Tudor I. Oprea
1.0
PPFS
0
0.3
0.4
0.5
0.6
0.7
0.8
75 “recent” drugs
0.9
1.0
Is There A Difference?
Index
Mean
Std Err.
Median
STDEV
Min.
Max.
RNG
0.55
0.13
0
1.06
-2
5
RTB
1.90
0.42
2
3.41
-12
12
RGB MW
3.34 78.97
0.67 10.13
3
69.88
5.45
82.95
-11
-120.1
24
386.3
HDO HAC
-0.18 0.45
0.2
0.23
0
1
1.62 1.88
-5
-5
3
5
Data from 67 drug-lead unique pairs.
For more than 1:1 correspondence, only larger structures were considered.
Medicinal Chemistry / AstraZeneca Mölnal
Tudor I. Oprea
CLP
1.25
0.31
0.67
2.58
-5.95
9.7
Is There A Real Difference?!
Index
Mean
Std Err.
Median
STDEV
Min.
Max.
RNG
1.0
0.14
1
0.78
0
3
RTB
2.48
0.61
2
3.3
-3.0
12
RGB
5.86
0.73
6
3.93
-1
13
MW HDO HAC
111
-0.21 0.41
16.6
0.38 0.41
96.2 0
1
89.2
2.0
2.2
-19
-5.0 -5.0
386.3 3
4
Data from 29 recent drug-lead unique pairs.
For more than 1:1 correspondence, only larger structures were considered.
Medicinal Chemistry / AstraZeneca Mölnal
Tudor I. Oprea
CLP
1.90
0.47
0.21
2.54
-3.74
9.7
What Have We Learned?
• There is a difference between “leads” and “drugs”
• However, this difference is apparent mostly in
pairwise comparisons, as the “leadspace” and
“drugspace” appear to overlap
• Average difference: 1 ring, 2 rotatable bonds, 100
daltons, 1 acceptor and 0.5-1 LogP unit
• Scoring leads yields 0.1-0.2 less units than drugs…
• The optimal way to use this information is to provide
“leadlike” profiles for combichem libraries, as well as
guidelines for HTS “hit” analysis in medicinal
chemistry efforts
Medicinal Chemistry / AstraZeneca Mölnal
Tudor I. Oprea
How to Reach a Leadlike Profile?
• There is no universal answer for a leadlike library
• However, depending on the final goal (e.g., CNS
vs. urinary antiinflammatory drugs), one can
generate criteria to focus combichem or HTS libraries
(go into “cherry-picking” mode right from the start)
• For example: 1-5 rings, 2-15 rotatable bonds, up to
400 daltons, 0-2 donors, 1-8 acceptors and 0-3
LogP units would cover MDDR-like space (75%)
• Additional criteria provided by druglike scoring
schemes (e.g., DFPS > 0.4, PPFS > 0.4)
• PSA < 60 Å2 for CNS; PSA < 140 Å2 for oral activity...
Medicinal Chemistry / AstraZeneca Mölnal
Tudor I. Oprea
Lipinski’s Rule of 5: Just How Good Is It?!
80%
ACD
70%
MDDR
PDR
Towards a computational
model for oral absorption
HB-acceptors
2
3
4
20%
10%
0%
PASS
FAIL
SKIPPED
More reagents than drugs pass the “Rule of 5” test
Medicinal Chemistry / AstraZeneca Mölnal
Tudor I. Oprea
5
6
7
8
9
There Is A Leadlike Space
• The bottom line is that we can direct discovery
towards leadlike profiling
• For the time being, we are limited by the number of
lead-drug pairs in our historycal analysis
• An ISIS database is available for those interested.
• We are currently working on expanding this for
including validated HTS hits (is there a “HTS-hits”
space? Is it different from the leadlike space?)
Medicinal Chemistry / AstraZeneca Mölnal
Tudor I. Oprea
Thanks to...
• Andy Davis, Simon Teague & Paul Leeson
• Bertil Samuelsson
• Mark Divers & Lennart Svensson
• Johan Gottfries, Ismael Zamora
• Thomas Kühler & Bob Carter
Medicinal Chemistry / AstraZeneca Mölnal
Tudor I. Oprea