Fragment-Based Drug Discovery Targeting Inhibitor of Apoptosis Proteins: Discovery of a Non-Alanine Lead Series with Dual Activity Against cIAP1 and XIAP

Article abstract of DOI:10.1021/acs.jmedchem.5b00706

(Figure Presented) Inhibitor of apoptosis proteins (IAPs) are important regulators of apoptosis and pro-survival signaling pathways whose deregulation is often associated with tumor genesis and tumor growth. IAPs have been proposed as targets for anticanc

Full text of DOI:10.1021/acs.jmedchem.5b00706

Subscriber access provided by UNIV OF CAMBRIDGE
Article
Fragment-Based Drug Discovery Targeting Inhibitor
of Apoptosis Proteins: Discovery of a Non-Alanine
Lead Series with Dual Activity Against cIAP1 and XIAP
Gianni Chessari, Ildiko M Buck, James Edward Harvey Day, Philip John Day, Aman Iqbal,
Christopher Norbert Johnson, Edward Jonathan Lewis, Vanessa Martins, Darcey Lynn Miller, Michael
Reader, David Charles Rees, Sharna Jayne Rich, Emiliano Tamanini, Marc Vitorino, George A
Ward, Pamela Ann Williams, Glyn Williams, Nicola Elizabeth Wilsher, and Alison J.-A. Woolford
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.5b00706 • Publication Date (Web): 28 Jul 2015
Downloaded from http://pubs.acs.org on August 1, 2015
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 45
Journal of Medicinal Chemistry
1
2
3
4
Fragment-Based Drug Discovery Targeting Inhibitor of Apoptosis Proteins: Discovery
of a Non-Alanine Lead Series with Dual Activity Against cIAP1 and XIAP
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Gianni Chessari,* Ildiko M. Buck, James E. H. Day, Philip J. Day, Aman Iqbal, Christopher
N. Johnson, Edward J. Lewis, Vanessa Martins, Darcey Miller, Michael Reader, David C.
Rees, Sharna J. Rich, Emiliano Tamanini, Marc Vitorino, George A. Ward, Pamela A.
Williams, Glyn Williams, Nicola E. Wilsher, Alison J.ꢀA. Woolford
Astex Pharmaceuticals, 436 Cambridge Science Park, Milton Road, Cambridge CB4 0QA,
U.K.
Page 1 of 45
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 45
1
2
3
4
ABSTRACT
5
6
7
8
9
Inhibitor of apoptosis proteins (IAPs) are important regulators of apoptosis and proꢀsurvival
signaling pathways, whose deregulation is often associated with tumor genesis and tumor
growth. IAPs have been proposed as targets for anticancer therapy and a number of
peptidomimetic IAP antagonists have entered clinical trials. Using our fragmentꢀbased
screening approach, we identified nonꢀpeptidic fragments binding with millimolar affinities
to both cellular inhibitor of apoptosis protein 1 (cIAP1) and Xꢀlinked inhibitor of apoptosis
protein (XIAP). Structureꢀbased hit optimization together with an analysis of proteinꢀligand
electrostatic potential complementarity allowed us to significantly increase binding affinity of
the starting hits. Subsequent optimization gave a potent nonꢀalanine IAP antagonist
structurally distinct from all IAP antagonists previously reported. The lead compound had
activity in cellꢀbased assays and in a mouse xenograft efficacy model and represents a highly
promising start point for further optimization.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Page 2 of 45
ACS Paragon Plus Environment

Page 3 of 45
Journal of Medicinal Chemistry
1
2
3
4
INTRODUCTION
Evasion of apoptosis is one of the hallmarks of cancer1 and can be achieved by over
expression of antiꢀapoptotic proteins. Inhibitor of apoptosis proteins (IAP), such as cellular
IAP1 (cIAP1) and Xꢀlinked IAP (XIAP), are key regulators of antiꢀapoptotic and proꢀsurvival
signaling pathways.2 Their deregulation occurs in various cancers and is associated with
tumor growth, poor prognosis and resistance to treatment, making them attractive targets for
antiꢀcancer therapy.3
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
IAPs are characterized by baculovirus IAP repeat (BIR) domains which mediate proteinꢀ
protein interactions; some members of the family such as cIAP1 and XIAP also possess
RING (Really Interesting New Gene) zinc finger domains with E3 ubiquitin ligase activity.4
XIAP exerts antiꢀapoptotic activity by binding and inactivation of caspases 3, 7 and 9 via its
BIR domains; caspases 3 and 7 bind to the linker between BIR domains 1 and 2, whilst the
Nꢀterminal ATPF sequence of caspase 9 is bound in the peptide binding groove of the BIR3
domain.5 As well as inhibiting apoptosis through the inactivation of caspases, other IAPs (in
particular cIAP1 and 2) affect the balance between proꢀapoptotic and proꢀsurvival signaling,
promoting the formation of complex I6 in response to TNFα, which activates survival
signaling through the canonical NFꢀκB pathway.7 Second Mitochondria derived Activator of
Caspases (SMAC) is an endogenous IAP antagonist, which is released from mitochondria on
induction of apoptosis. SMAC binds to the BIR domains of IAPs and can disrupt interactions
between XIAP and caspases.8 SMAC binding to cIAP1 induces a conformational change
which activates its E3 ligase function, leading to rapid autoubiquitinylation and proteasomal
degradation.9 Following removal of cIAP1, RIPK1 recruits FADD and caspaseꢀ8 to form
complexꢀIIB, leading to a switch in TNFα signaling from proꢀsurvival to proꢀapoptotic.10
This loss of cIAP1, combined with antagonism of XIAPꢀmediated caspase inhibition, leads to
a sustained proꢀapoptotic effect in the presence of TNFα, suggesting that tumors with
Page 3 of 45
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 45
1
2
3
4
5
6
7
8
sufficient levels of TNFα in their environment may be particularly susceptible to IAP
antagonism. It has been suggested that potent and dual cIAP1/XIAP antagonists are more
efficient in activating apoptosis than unbalanced cIAP1 antagonists.11
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The endogenous ligand SMAC binds to the BIR3 domains of cIAP1 and XIAP via its Nꢀ
terminal peptide sequence AVPI.8b This tetrapeptide motif has provided the starting point for
various peptidomimetic approaches12 which have resulted in the discovery of several
monovalent and bivalent clinical candidates,13 some of which are illustrated in Figure 1.
Figure 1. IAP antagonists in clinical trials showing derivation from SMAC Nꢀterminal
sequence AVPI.
AVPI itself can bind to the BIR3 domain of cIAP1 or XIAP (cIAP1ꢀBIR3, XIAPꢀBIR3)
occupying subpockets P1ꢀP4 (Figure 1), and shows 18 fold selectivity for cIAP1ꢀBIR3 over
XIAPꢀBIR3. This intrinsic selectivity is mainly due to the differential interactions formed by
the alanine residue in the P1 pocket of the BIR3 domains of XIAP and cIAP1 and is usually
maintained in alanineꢀbased SMAC mimetic inhibitors.
Page 4 of 45
ACS Paragon Plus Environment

Page 5 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Fragmentꢀbased drug discovery (FBDD) has become both a complementary and an
alternative approach to more traditional drug discovery methods (e.g. highꢀthroughput
screening, peptidomimetics) for the generation of chemical leads,14 particularly for
challenging and less tractable targets such as ProteinꢀProtein Interaction (PPI) targets.15 Astex
Pharmaceuticals has applied its FBDD technology platform (PyramidTM)16 to a range of target
classes and, in this paper, we describe the use of PyramidTM to identify new nonꢀalanine hits,
with the potential to be developed into potent and balanced XIAP and cIAP1 antagonists.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
FRAGMENT SCREENING
The Astex fragment library and a focused fragment set containing 100 fragments selected by
virtual screening were screened against XIAPꢀBIR3. The first step of the screening cascade
was based on observing protein signals in 1D NMR experiments. Ligandꢀdetected NMR
methods, such as LOGSY or STDꢀNMR that are commonly used to detect fragment binding,
were found to be relatively insensitive for XIAP due to the small size of the protein (11.8
kDa). A protein concentration of 200 ꢀM allowed the use of the protein 1HꢀNMR spectrum to
1
detect fragment binding by monitoring the chemical shifts and linewidths of XIAPꢀBIR3 H
signals with δ <0.4 ppm. NMR signals in this chemical shift range arise from strongly
1
shielded protons in aliphatic side chains. None of the fragment H NMR spectra had signals
in this region. In addition the range δ 9.8ꢀ10.4 ppm was also monitored; in XIAPꢀBIR3 this
region contains four signals, arising from the indole NH protons of tryptophan side chains,
two of which are in the binding pocket of interest. Both of these regions of the NMR
spectrum contain a number of signals which are strongly affected by the binding of SMACꢀ
derived peptides with sequences AV, AVP and AVPI and so were also expected to be
sensitive to fragment binding in the canonical AVPI pocket (data not shown). In total, 1151
fragments were screened against XIAPꢀBIR3 in cocktails of two. In order to detect very
weaklyꢀbinding hits, each fragment was present at a concentration of 10 mM. We believe that
Page 5 of 45
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 45
1
2
3
4
5
6
under these conditions it was possible to detect ligands binding to XIAPꢀBIR3 with
dissociation constant (Kd) of 30 mM or better.
7
8
9
1
Figure 2 shows representative XIAPꢀBIR3 H NMR spectra obtained during fragment
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
screening. Figure 2A shows overlaid spectra (δ 0.4ꢀ0.3 ppm) from cocktails which contain no
fragment hits. In contrast Figure 2B shows the same spectral region from samples containing
hits. A number of signals shift and gain intensity as fragments bind, presumably as a result of
changes in conformation and mobility of amino acid side chains around the AVPI pocket of
XIAPꢀBIR3.
Figure 2. Overlaid 1H NMR spectra of XIAPꢀBIR3 in the presence of fragment cocktails: (A)
containing no hits and (B) containing hits.
Cocktails which contained hits were deconvoluted and fragments responsible for the shift
changes were identified and further characterized by a 1Hꢀ15N HSQC NMR experiment using
15N labeled XIAPꢀBIR3 (data not shown).
An Xꢀray based fragment screen of XIAPꢀBIR3 was also undertaken using crystals of XIAPꢀ
BIR3 (residues 250ꢀ354) grown in the absence of compound (for details, see Experimental
Section). Fragments were soaked into crystals for 24ꢀ72 hours either as singletons (at 50ꢀ100
Page 6 of 45
ACS Paragon Plus Environment

Page 7 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
mM fragment, 5ꢀ10% DMSO) or doublets (2 x 50 mM, 10% DMSO). In addition, fragment
hits, previously identified by NMR screening and/or high concentration bioassay, were
soaked into XIAP crystals.
9
Description of fragment hits
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The majority of the fragment hits contained small primary or secondary aliphatic amines,
consistent with the known binding site preference for Nꢀterminal alanineꢀcontaining peptides.
The chemical structures for four fragment hits from the focused fragment set are reported in
Table 1, together with IC50 values or %I for the BIR3 domains of XIAP and cIAP1. The
binding modes were determined by Xꢀray crystallography.
Table 1. Chemical structures of four hits from fragment screening against XIAPꢀBIR3 with
binding affinity for the BIR3 domains of XIAP and cIAP1.
XIAPꢀBIR3 LEb
XIAPꢀBIR3a
IC50 (µM) or %I
cIAP1ꢀBIR3a
IC50 (µM) or %I
Compound
Structure
(kcal molꢀ1 per
nonꢀH atom)
38% at 5 mM
33% at 5 mM
1200
<0.21
<0.26
0.30
16% at 5 mM
1
55% at 5 mM
2
3
24
27
110
0.42
4
a
Values were determined by fluorescence polarization assay (see Experimental Section).
b
Potency data are reported as the mean of at least two runs. Values calculated according to
the Hopkins formula.17
Page 7 of 45
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 45
1
2
3
4
5
6
7
8
9
Compounds 1 and 2 are weakly binding fragments with affinities above 5 mM for both
XIAPꢀBIR3 and cIAP1ꢀBIR3. Their detection was only possible because of the high
sensitivity of Xꢀray crystallography and NMR techniques used in the fragment screening.
Ligand efficiencies (LEs)17 are low and inferior to the LE values observed for the alanine
based fragments 3 and 4. The binding mode for compound 1 in the BIR3 domain of XIAP is
shown in Figure 3A. The piperazine ring lies in the P1 pocket with the protonated nitrogen
forming a bidentate hydrogen bond interaction with the Glu314 side chain and the backbone
carbonyl of Asp309. The carbonyl of the central amide forms a hydrogen bond with the
backbone NH of Thr308 and, in addition, the piperidine ring stacks against the side chain of
Trp323 in the P3 pocket. Piperazine containing molecules have been independently identified
in a biophysical screen of a XIAPꢀBIR3 G306E mutant,18 but subsequent optimization was
not described.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The piperidine ring of compound 2 also forms a bidendate hydrogen bond interaction in the
P1 pocket of XIAPꢀBIR3 similar to that observed for compound 1 (see Figure 3B). The
bromoꢀpyrazole is in van der Waals contact with the side chain of Leu307 and forms an edgeꢀ
toꢀface aromatic interaction with Trp323, while the nonꢀsubstituted pyrazole nitrogen forms a
hydrogen bond with the backbone NH of Thr308.
Fragment 3 is an alanineꢀvaline dipeptide, which binds to the P1 and P2 pockets of XIAPꢀ
BIR3 (Figure 3C). The positively charged primary amine of the alanine forms a chargeꢀ
charge interaction with the carboxylic group in the side chain of Glu314 and an additional
hydrogen bond with the side chain of Gln319. The methyl group of the alanine inserts into
the small P1 hydrophobic pocket where it stacks on top of Trp310. The valine forms two βꢀ
sheetꢀlike hydrogen bonds with Thr308. The binding mode of 3 is consistent with Xꢀray
crystal structure of the Nꢀterminal sequence of SMAC bound to XIAPꢀBIR3.8b
Page 8 of 45
ACS Paragon Plus Environment

Page 9 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Compound 4 is another example of an alanineꢀlike fragment identified from the screen, where
the alanine recognition is similar to that observed for fragment 3. The pyridine ring stacks on
the top of Leu307 and forms an edgeꢀtoꢀface aromatic interaction with the side chain of
Trp323, whilst the pyridine nitrogen is hydrogen bonded to the backbone NH of Thr308.
Heterocycles properly functionalized with an alanine warhead similar to compound 4 have
previously been reported to bind to XIAPꢀBIR3 with similar affinity to the natural AVPI
peptide.19
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
All alanine based hits from our screen are characterized by relatively high LE for both XIAP
and cIAP1, but these fragments show an unbalanced profile with greater than 10 fold
selectivity towards cIAP1. We reasoned that their optimization could lead to compounds with
a biological profile similar to peptidomimetics currently in clinical development. Nonꢀalanine
fragment hits based on the piperazine or piperidine scaffold are weak binders of both XIAP
and cIAP1, with LEs well below the benchmark value of 0.3. However, we were interested to
explore their potential as a starting point for a more balanced dual XIAP and cIAP1
antagonist. In this paper we report the optimization of fragment 1 into a lead molecule with in
vivo antitumor activity.
Page 9 of 45
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 45
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Xꢀray crystal structures of fragments 1 ꢀ 4 bound to XIAPꢀBIR3 (A to D
respectively). The purple mesh represents the initial electron density map (Fo – Fc maps
contoured at 1σ and clipped around the ligand) and hydrogen bonds are shown as dashed red
lines.
PIPERAZINE FRAGMENT OPTIMIZATION
Comparison of the binding mode of fragment hit 1 with that of alanine fragment hits 3 and 4
provided a good hypothesis for improving affinity. The crystal structure of XIAPꢀBIR3 with
1 shows that the piperazine does not fill the small lipophilic pocket in P1, where the alanine
methyl side chain binds forming a van der Waals contact with the Trp310 side chain. An
overlay between fragment 1 and AlaꢀVal 3 led to the design and synthesis of compound 5,
where the carbon adjacent to the piperazine nitrogen has been substituted by a methyl group,
Page 10 of 45
ACS Paragon Plus Environment

Page 11 of 45
Journal of Medicinal Chemistry
1
2
3
4
in order to increase the contacts with the P1 pocket. Compound 5 showed improved potency
5
6
7
8
and LE and its superior filling of the P1 pocket was confirmed by crystallography (see Figure
4).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. Compound 5 affinity data and binding mode in XIAPꢀBIR3 as determined by Xꢀ
ray crystallography showing good shape complementarity in the P1 pocket between the
Connolly surfaces of protein (solid gray) and the Connolly surface of the ligand (orange
mesh).
The next design iteration investigated a selection of tertiary amides to replace the pyrrolidine
in compound 5. A small virtual library (Heavy Atom Count < 24 and ClogP < 3) was
designed, in order to identify side chains which could form good interactions with the P3
pocket and also could provide good vectors to allow entry into the P4 pocket. All the
members of the virtual library were docked against XIAPꢀBIR3. Binding modes were
visually inspected and 30 compounds were selected for synthesis. Compounds were prepared
by coupling 2ꢀ[(3R)ꢀ4ꢀ[(tertꢀbutoxy)carbonyl]ꢀ3ꢀmethylpiperazinꢀ1ꢀyl]acetic acid (6, R2 = H,
see Figure 5) with a range of aliphatic and aromatic amines, and subsequently removing the
Boc protecting group.
Page 11 of 45
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 45
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5. Retrosynthetic analysis for nonꢀalanine IAP antagonists. Protected piperazineꢀ
carboxylic acids (see Scheme 1) and bicyclic amines (see Scheme 3) are coupled together
before final Boc deprotection.
The resulting compounds were initially evaluated using fluorescence polarization (FP) assays
and Xꢀray crystal structures were attempted by soaking to confirm binding to XIAPꢀBIR3.
Unfortunately, all members of the compound set showed potencies which were comparable
or inferior to compound 5, and improvements in binding affinity and LE were not achieved.
However, it was possible to obtain Xꢀray crystal structures of some members of the set bound
to XIAPꢀBIR3. In particular the structure of indoline 7 (Table 2 and Figure 6) is in good
agreement with the design hypothesis; the piperazine warhead forms the same interactions as
the optimized fragment 5 and the indoline side chain binds to the P3 region as expected from
the modeling.
Page 12 of 45
ACS Paragon Plus Environment

Page 13 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 6. (A) Orthogonal views of compound 7 (green) bound to XIAPꢀBIR3 as determined
by Xꢀray crystallography. The Connolly surface of the protein is colored by electrostatic
potential (red = negative, blue = positive) and ligand Cresset field points as colored spheres
(red = positive, blue = negative, orange = lipophilic). (B) Plots of Hammett σp coefficients
versus XIAPꢀBIR3 affinities for Cꢀ6 substituted indolines 7 - 16. (C) XIAPꢀBIR3 affinity
data for compounds 8 and 17 with comparison of molecular electrostatic potential surfaces.
We investigated the electrostatic environment of the XIAPꢀBIR3 binding pocket, in order to
explain the lack of improved affinity for compound 7 (XIAPꢀBIR3 52% at 495 µM, cIAP1ꢀ
BIR3 42% at 495 µM). The electrostatic potential map obtained for the XIAPꢀBIR3 binding
Page 13 of 45
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 45
1
2
3
4
5
6
7
8
9
site (see Figure 6A) clearly indicated an electronegative patch in the P3 subꢀpocket which is
due to the presence of the backbone carbonyl of Gly306 and the phenolic oxygen in the side
chain of Tyr324. The Xꢀray crystal structure of compound 7 showed that the aromatic portion
of the indoline ring stacks on the top of the negative patch. In order to understand this
interaction, the molecular electrostatic potential of 7 was calculated with the XED forceꢀ
field,20 which uses offꢀatom charges for better representation of electron anisotropy around
the ligand. Subsequently, negative and positive electrostatic field points were derived using
Cresset technology.21 Figure 6A shows that the πꢀcloud of the indoline ring (represented by
negative field points above and below the aromatic ring) stacks on the top of the negative
patch on the protein surface, generating a potential electrostatic repulsion between ligand and
protein, which could explain the low affinity of compound 7.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In order to test this hypothesis, the electronegativity of the indoline πꢀclouds was modulated
by functionalizing the indoline Cꢀ6 with a range of electronꢀwithdrawing and electronꢀ
donating substituents (Figure 5; R1 = R2 = H, X = CH). Hammett σp values were used to
guide the selection of the substituents.22 The resulting SAR is summarized in Table 2 and
Figure 6B.
Page 14 of 45
ACS Paragon Plus Environment

Page 15 of 45
Journal of Medicinal Chemistry
1
2
3
4
Table 2. XIAPꢀBIR3 affinity of substituted indolines 7 - 16
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
XIAPꢀBIR3 LEb
(kcal molꢀ1 per
nonꢀH atom)
XIAPꢀBIR3a IC50
Compound
R
Hammett σp
(µM) or %I
ꢀH
ꢀNH2
ꢀOMe
ꢀMe
0.00
ꢀ0.66
ꢀ0.27
ꢀ0.17
ꢀ0.15
0.06
0.23
0.23
0.54
0.72
~0.24
~0.20
~0.25
0.30
0.26
0.29
0.33
0.34
0.31
0.32
7 (23050)
8 (23307)
9 (23305)
10 (23306)
11 (23352)
12 (23330)
13 (23304)
14 (23132)
15 (23488)
16 (23561)
52% @ 495 ꢀM
56% @ 1000 ꢀM
49% @ 155 ꢀM
46
59
ꢀiPr
ꢀF
51
ꢀCl
13
ꢀBr
9.8
5.9
4.1
ꢀCF3
ꢀSO2Me
a See footnote to Table 1
XIAPꢀBIR3 binding affinities are highly sensitive to the electronic properties of the indoline
πꢀsystem, with electronꢀpoor indolines (e.g. 15 and 16) being two orders of magnitude more
potent than electronꢀrich indolines (e.g. 8 and 9). This dependency is also shown in Figure 6B
where binding affinities correlate well with relative Hammett substituent constants σp.
Interestingly, the unsubstituted indoline 7 appears to be an outlier. This discrepancy could be
explained by the van der Waals contribution to binding of the different substituents which is
absent for the unsubstituted indoline 7.
Page 15 of 45
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 45
1
2
3
4
5
6
7
8
9
Electrostatic potential surfaces generated with ab initio methods also suggested that it was
possible to further discharge the πꢀcloud of the indoline by introducing a nitrogen into the
aromatic ring (see Figure 6C). Based on this observation compound 17 was synthesized
bearing a chloro substituent suitable for further elaboration to access the P4 pocket. Whilst
the binding modes of 17 and 7 are very similar (Figure 7B), careful optimization of the
electrostatic interaction between ligand and protein has enabled a >50 fold jump in potency
both for XIAPꢀBIR3 (IC50 = 7.7 ꢀM, LE = 0.35) and cIAP1ꢀBIR3 (IC50 = 20 ꢀM, LE 0.32)
without dramatically changing the size of the ligand. The azaꢀindoline scaffold also has a
positive impact on the lipophilicity of ligand 17, which has a ClogP of 1.1 and a LLEAstex
value23 of 0.35 (based on XIAPꢀBIR3 IC50). Importantly, 17 had a balanced XIAP/cIAP1
antagonist profile, vindicating the original decision to pursue a nonꢀalanine fragment hit. The
balanced profile of the piperazines can be rationalized by the fact the piperazine warhead
forms two similar polar interactions in the P1 pockets of both XIAPꢀBIR3 and cIAP1ꢀBIR3: a
salt bridge to Glu314 for XIAPꢀBIR3 or Asp314 for cIAP1ꢀBIR3 and an Hꢀbond interaction
with the backbone carbonyl of Asp309 for XIAPꢀBIR3 or Cys309 for cIAP1ꢀBIR3 (see
Figure 7). In contrast, the recognition of the alanine warhead (as shown for fragment 3 in the
P1 pocket of cIAP1ꢀBIR3) is driven by two chargeꢀcharge interactions to Asp309 and
Glu319, while in the P1 pocket of XIAPꢀBIR3 only a single chargeꢀcharge interaction is
possible, namely to Glu319.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Page 16 of 45
ACS Paragon Plus Environment

Page 17 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 7. Comparison of binding modes of alanine 3 and piperazine 17 in XIAPꢀBIR3 (A
and B respectively) as determined by Xꢀray crystallography. Key residue changes for cIAP1
vs. XIAP are shown in orange and hydrogen bonds are shown as purple dashed lines.
The interaction between the indoline side chain and the P3 pockets of XIAPꢀBIR3 and
cIAP1ꢀBIR3 was further optimized by introducing two methyl groups at Cꢀ3 of the bicycle,
with the aim of increasing the van der Waals contact with Trp323. This change was also
introduced to discharge the risk of metabolic oxidation of the indoline to the indole. Indoline
18 showed modestly improved potency compared to the nonꢀmethylated analog 13 (see Table
3) and its binding mode in XIAPꢀBIR3 was consistent with the other indolines (Figure 8Aꢀ
B). The overlay between the XIAPꢀBIR3 structure of compound 18 and the tetrapeptide
AVPF (see Figure 8A) shows that both P2 and P4 pockets could be targeted by
functionalizing Cꢀ2 of the piperazine and Cꢀ6 of the indoline scaffold respectively. These
design ideas were tested by preparing compounds 19, 20 and 21 (see Table 3).
Page 17 of 45
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 45
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 8. (A) Overlay of Xꢀray crystal structures of compound 18 (orange) and AVPF (cyan)
bound to XIAPꢀBIR3. (B) and (C) Comparison of Xꢀray crystal structures of compounds 20
and 21 respectively, showing the effect of Lys297 side chain orientation on the size of the P4
pocket.
Page 18 of 45
ACS Paragon Plus Environment

Page 19 of 45
Journal of Medicinal Chemistry
1
2
3
4
Table 3. XIAPꢀBIR3 and cIAP1ꢀBIR3 binding affinity data for compounds 18-21.
5
6
7
8
9
XIAPꢀBIR3 LE
XIAPꢀBIR3a
IC50 (µM)
cIAP1ꢀBIR3a
IC50 (µM)
Compound
Structure
(kcal molꢀ1 per
nonꢀH atom)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
5.5
1.5
0.33
0.32
0.34
3.7
18
0.65
0.32
19
0.64
20
21
0.16
0.30
0.01
See Figure 1
70% at 0.05 µM
>0.24
98% at 0.01 µM
Debio-1143
a see footnote to Table 1.
In compounds 19 and 20 the methoxymethyl side chain was introduced trans at the piperazine
Cꢀ2 and a jump in potency was observed, with compound 20 being the first compound with
subꢀmicromolar potency for both IAPs. The Xꢀray crystal structure of compound 20
confirmed the expected binding mode where, interestingly, the P4 pocket is not available
because it is occupied by the side chain of Lys297 (see Figure 8B). In compound 21, Cꢀ6 was
elaborated with a large side chain in order to target the P4 pocket. The XIAPꢀBIR3 structure
of 21 shows that the side chain of Lys297 moves allowing the benzyl group to bind to the P4
pocket (see Figure 8C). The movement of the Lys297 was anticipated from the crystal
structures of SMAC mimetics. This extra lipophilic interaction further improved the potency
Page 19 of 45
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 45
1
2
3
4
5
6
7
8
9
of the series and, gratifyingly, compound 21 achieves nM activity for both IAPs. The gain in
cIAP1 potency is more pronounced as a result of adding a large lipophilic P4 group
suggesting that further optimization of the P4 side chain is required to develop potent and
balanced cIAP1 and XIAP antagonists. The contributions to XIAP binding affinity expressed
in terms of group efficiency (GE)24 of the various side chains are summarized in Figure 9.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 9. Group efficiency (GE) schematic for compound 21 based on multiple pairꢀwise
comparisons and estimated affinity data for the piperazine core scaffold vs XIAPꢀBIR3. GE
units in kcal molꢀ1 per nonꢀH atom.
Compound 21 was tested in cell proliferation assays using two sensitive human breast cancer
cell lines (EVSAꢀT and MDAꢀMBꢀ231), with the insensitive human colon cancer cell line
HCT116 included as a control for offꢀtarget cytotoxicity. The clinical compound, Debioꢀ1143
(Debiopharm Group), was included as an assay standard (see Table 4).13c EVSAꢀT and
MDAꢀMBꢀ231 cell lines undergo apoptosis with an IAP antagonist which is driven by
autocrine production of TNFα, whilst HCT116 cells are resistant to IAP antagonists due to
reduced TNFα production.
Page 20 of 45
ACS Paragon Plus Environment

Page 21 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The EVSAꢀT and MDAꢀMBꢀ231 cell lines are particularly sensitive to cIAP1 antagonism
and we desired a direct measure of XIAP antagonism in cells. Hence, we stably transfected
HEK293 cells with fullꢀlength XIAP (FLAGꢀtagged) and caspaseꢀ9 to generate an engineered
cell line in which an immunoprecipitation assay could be used to measure levels of
XIAP:Caspaseꢀ9 interaction after a 2 h incubation with IAP antagonist. In this assay, 21 was
shown to antagonize XIAP with an EC50 of 140 nM whilst the antiꢀproliferative EC50 on the
cIAP1ꢀsensitive cell line, EVSAꢀT, was 100 nM (see Table 4). In separate cellꢀbased assays a
reduction in cIAP1 levels was measured in EVSAꢀT cells after 2 h incubation with 21; and a
significant increase in cleaved caspaseꢀ3 was induced in EVSAꢀT cells after 6 h incubation
with 21 (data not shown).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 4. Compound 21 and Debioꢀ1143 cellꢀbased assay data
21
Debio-1143
Cell Line
Cellꢀbased assay
EC50 (µM) (n) EC50 (µM) (n)
EVSAꢀT
Proliferation (72 h)
Proliferation (72 h)
Proliferation (72 h)
Immunoprecipitation (2 h)
0.10 (2)
0.59 (2)
>10 (2)
0.14 (3)
0.0021 (51)
0.019 (51)
>10 (50)
MDAꢀMBꢀ231
HCT116
HEK293ꢀXIAPꢀCaspaseꢀ9a
0.034 (92)
a
HEK293ꢀXIAPꢀCaspaseꢀ9 is a stably transfected HEK293 cell line engineered to
overexpress fullꢀlength XIAP (FLAGꢀtagged) and caspaseꢀ9. For further details of all cellꢀ
based assays, see Supporting Information.
In order to demonstrate target engagement in vivo, a single 100 mg/kg dose of 21 was
administered intraperitoneally (i.p.) in Balb/c scid mice bearing MDAꢀMBꢀ231 xenografts.
At 1, 6 and 24 h post dose tumors were harvested for Western blot analysis of xenograft
lysates (see Figure 10A). A significant reduction in cIAP1 levels relative to the vehicleꢀdosed
Page 21 of 45
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 45
1
2
3
4
5
6
7
8
9
animal tumors was evident in tumors taken from mice 1 h post dose with compound 21 (this
was confirmed by densitometry of the blot and subsequent cIAP1 assay tumor lysates using
an MesoꢀScale Discovery (MSD) plateꢀbased method for detection of cIAP1 levels – data not
shown). However, it was noted that cIAP1 levels were variable at 6 h and had returned to
vehicle levels by 24 h, suggesting that plasma and tumor free concentrations were insufficient
to produce a prolonged pharmacodynamic effect. Additionally, there was no obvious effect
on XIAP levels (indicating that binding of compound 21 to XIAPꢀBIR3 does not lead to
ubiquitinꢀmediated degradation of XIAP) and only a slight increase in apoptosis markers
(cleaved PARP and cleaved caspaseꢀ3) in the compound 21 treated animals (see Figure 10A).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 10. (A) Pharmacodynamic (PD) and (B) Pharmacokinetic (PK) effect of a single 100
mg/kg dose of 21 i.p. in Balb/c scid mice bearing MDAꢀMBꢀ231 xenografts. (C) In vivo
Page 22 of 45
ACS Paragon Plus Environment

Page 23 of 45
Journal of Medicinal Chemistry
1
2
3
4
efficacy in Balb/c scid mice bearing MDAꢀMBꢀ231 xenografts dosed i.p. with 21 at 100
5
6
mg/kg b.i.d.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In order to demonstrate efficacy in a mouse xenograft model, 100 mg/kg i.p. doses of 21 were
administered b.i.d. in Balb/c scid mice bearing MDAꢀMBꢀ231 xenografts. As a positive
control in this study the clinical IAP antagonist, Debioꢀ1143, was dosed daily (q.d.) at 100
mg/kg p.o. Significant reduction in tumor size was measured with twice daily (b.i.d.) dosing
i.p. of 21 (see Figure 10C), with no significant effect on body weight.
CHEMISTRY
The majority of compounds described in this paper were assembled via a convergent
synthesis involving late stage amide coupling, as shown in Figure 5, allowing maximum
flexibility to explore the SAR in different regions of the protein (see Supporting Information
for detailed experimental procedures). The carboxylic acidꢀfunctionalized piperazines 22 and
23 were prepared according to Scheme 1 and indolines and azaindolines 24 – 31 according to
Scheme 2.
Piperazine 22 was readily prepared from commercially available (R)ꢀ2ꢀmethylꢀpiperazineꢀ1ꢀ
carboxylic acid tertꢀbutyl ester by alkylation with benzyl bromoacetate followed by
hydrogenolysis of the benzyl group as shown in Scheme 1. The two stereocenters in
piperazine 23 are derived from Dꢀalanine methyl ester and Cbz protected Lꢀserine. EDC
mediated coupling gave a dipeptide which cyclized in situ following hydrogenolysis of the
Cbz protecting group. The resulting diketopiperazine was reduced with borane and the
product treated with diꢀtertꢀbutyl dicarbonate. The initially formed bisꢀBoc derivative was
then treated with aqueous ethanolic sodium hydroxide, resulting in selective removal of the
Boc group adjacent to the hydroxymethyl substituent, the reaction proceeding via formation
Page 23 of 45
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 45
1
2
3
4
5
6
7
8
9
of the bicyclic oxazolidinone. Treatment with benzyl chloroformate then gave the
orthogonally protected 2,5ꢀdisubstituted piperazine. Alkylation of the hydroxyl group was
accomplished using trimethyloxonium tetrafluoroborate in the presence of ProtonꢀSponge®
(1,8ꢀbis(dimethylamino)naphthalene). Following removal of the Cbz group, elaboration to
carboxylic acid 23 was performed in a manner similar to that described for 22.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 1. Reagents and conditions: a) benzyl bromoacetate, K2CO3, ACN, R.T., 82%; b)
Pd/C, MeOH, H2, R.T., >94%; c) EDC, DIPEA, DCM, R.T., 74%; d) Pd/C, MeOH,
.
cyclohexene, reflux, 81%; e) BH3 THF, THF, 70 °C, quant.; f) Boc2O, TEA, MeOH 50 °C,
then NaOH, aq. EtOH, 100 °C, 75%; g) benzyl chloroformate, NaOHaq., THF, R.T., 76%; h)
Protonꢀsponge®, trimethyloxonium tetrafluoroborate, DCM, 5ꢀ17 °C, 65%.
The required 6ꢀsubstituted indolines were either commercially available or could be prepared
by reduction of the commercially available indole or azaꢀindole precursors, using either
Page 24 of 45
ACS Paragon Plus Environment

Page 25 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
borane or NaBH3CN (Scheme 2, compounds 24 – 29). The 3,3ꢀdimethyl substituted
chloroindoline 30 was synthesized by a modified Fischer indole procedure starting from 3ꢀ
chlorophenylhydrazine and isobutyraldehyde, in which the desired 6ꢀchloro regioisomer was
separated from the 4ꢀchloro by chromatography. The synthesis of the more elaborated 3,3ꢀ
dimethylꢀ5ꢀazaꢀ6ꢀchloroindoline 31 is accomplished using a 6 step procedure, starting from
commercially available 6ꢀchloroꢀ5ꢀazaindole, introduction of the gemꢀdimethyl groups being
effected by deprotonation of an azaꢀoxindole intermediate and subsequent treatment with
iodomethane.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 2. Reagents and conditions: a) BH3, THF then TFA, 0 °C, 64ꢀ86%; b) NaBH3CN,
AcOH, R.T., 35ꢀ66%; c) BH3.SMe2, THF then MeOH, 68 °C, 41%; d) AcOH, 60 °C then
NaBH(OAc)3, DCE, R.T., 4%; e) NaH, (CH3)3SiCH2CH2OCH2Cl, DMF, 0 ºC, 82%; f)
C5H5N·HBr·Br2, dioxane, R.T., 58%; g) Zn, NH4Cl (aq), THF, R.T., 75%; h) LiHMDS,
CH3I, THF, ꢀ78 ºC, 51%; i) TFA, DCM, R.T., 67%; j) BH3ꢀMe2S, THF, 68 ºC, 95%.
Page 25 of 45
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 45
1
2
3
4
5
6
7
8
Following the schematic in Figure 5, the appropriate amine and acid precursors were coupled
using PyBroP and the products deprotected with concentrated HCl in dioxane as illustrated
for compounds 19 and 20 (Scheme 3), to give target compounds 7, 9, 10 and 12 ꢀ 20.
Optimized fragment 5 was also prepared following similar methods using pyrrolidine in the
amide coupling step. Compound 21 was prepared by late stage Negishi coupling of the
protected chloroꢀazaindoline with benzylzinc bromide followed by Boc deprotection.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 3. Reagents and conditions: a) PyBroP, TEA, DCM, R.T., 74%; b) conc. HCl,
dioxane, R.T., 78ꢀ85%; c) Benzylzinc bromide, PEPPSITMꢀIPr, LiBr, NMP, THF, R.T., 78%.
Analogs 8 and 11 were also prepared by late stage elaboration of Cꢀ6 on the bicycle (Scheme
4). Compound 8 was prepared via reduction of a 6ꢀnitro indoline derivative, while for 11 the
isopropyl group was introduced by Suzuki coupling of 2ꢀisopropenylꢀ4,4,5,5ꢀtetramethylꢀ
1,3,2ꢀdioxaborolane with a 6ꢀbromo indoline derivative followed by hydrogenation.
Page 26 of 45
ACS Paragon Plus Environment

Page 27 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 4. Reagents and conditions: a) 22, PyBroP, TEA, DCM, R.T., 50%; b) 10% Pd/C,
H2, MeOH, R.T., quant.; c) conc. HCl, dioxane, R.T., 82%; d) 2ꢀisopropenylꢀ4,4,5,5ꢀ
tetramethylꢀ1,3,2ꢀdioxaborolane, Pd(PPh3)4, Cs2CO3, 85 °C, 86% over 2 steps.
CONCLUSION
We have applied Astex PyramidTM to XIAPꢀBIR3 in order to identify nonꢀalanine starting
points for antagonists with balanced dual cIAP1/XIAP profiles. A combination of protein
1
observed HꢀNMR screening and Xꢀray crystallography screening based on long soaking
experiments using high fragment concentrations allowed us to identify very weak nonꢀalanine
fragment hits. Structureꢀbased drug design coupled to an electrostatic analysis of the XIAPꢀ
BIR3 binding site enabled the efficient optimization of the piperazineꢀbased initial hit 1 into a
scaffold 17 with low micromolar activity and good LE for both XIAP and cIAP1 BIR3
domains. Subsequent elaboration gave compound 21, which antagonizes XIAP, induces
cIAP1 degradation and inhibits cell growth in cancer cell lines. Biomarker modulation
associated with reduction in tumor size was observed in a mouse xenograft model bearing the
breast cancer cell line MDAꢀMBꢀ231 xenografts when treated with 100 mg/kg i.p. doses of
21. Compound 21 is the first reported example of a nanomolar affinity non peptidic IAP
antagonist which does not possess an alanine warhead. Hence 21 represents a novel
chemotype for an important class of anticancer targets and has proved a promising starting
Page 27 of 45
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 45
1
2
3
4
5
6
7
8
point for further medicinal chemistry. Further optimization of 21 into an orally bioavailable
and efficacious dual XIAP/cIAP1 antagonist with low nanomolar affinity for both targets will
be reported in a subsequent publication.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
EXPERIMENTAL SECTION
General Chemistry
All solvents employed were commercially available anhydrous grade, and reagents were used
as received unless otherwise noted. Hereafter, petrol denotes the petroleum ether fraction
boiling at 40 – 60 oC. Flash column chromatography was performed on a Biotage SP1 system
(32−63 µm particle size, KPꢀSil, 60 Å pore size). NMR spectra were recorded on a Bruker
AV400 (Avance 400 MHz) spectrometer. Analytical LC−MS was conducted using an Agilent
1200 series with Mass Spec Detector coupled with an Agilent 6140 single quadrupole mass
detector and an Agilent 1200 MWD SLUV detector. LC retention times, molecular ion (m/z)
and LC purity (by UV) were based on the method below. Purity of compounds (as measured
by peak area ratio) was >95%, as determined by the LC method described below.
LC Method (BASIC)
Eluent A:
Eluent B:
Gradient:
Flow:
95:5 10 mM NH4HCO3+NH4OH:CH3CN (pH = 9.2)
CH3CN
5ꢀ95% eluent B over 1.1 minutes
0.9 ml/min
Column:
Column T:
Waters Acquity UPLC BEH C18; 1.7ꢀ; 2.1x50 mm
50°C
Page 28 of 45
ACS Paragon Plus Environment

Page 29 of 45
Journal of Medicinal Chemistry
1
2
3
4
Preparation of compound 21 (Scheme 3): 1-(6-Benzyl-3,3-dimethyl-2,3-dihydro-
5
6
7
8
pyrrolo[3,2-c]pyridin-1-yl)-2-((2
ethanone hydrochloride salt
R,5R)-2-methoxymethyl-5-methyl-piperazin-1-yl)-
9
Step 1: (2R,5R)ꢀ4ꢀ[2ꢀ(6ꢀChloroꢀ3,3ꢀdimethylꢀ2,3ꢀdihydroꢀpyrrolo[3,2ꢀc]pyridinꢀ1ꢀyl)ꢀ2ꢀoxoꢀ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ethyl]ꢀ5ꢀmethoxymethylꢀ2ꢀmethylꢀpiperazineꢀ1ꢀcarboxylic acid tertꢀbutyl ester
To
a
solution of (2R,5R)ꢀ4ꢀcarboxymethylꢀ5ꢀmethoxymethylꢀ2ꢀmethylꢀpiperazineꢀ1ꢀ
carboxylic acid tertꢀbutyl ester 23 (4.1 g, 13.7 mmol, see Supporting Information) and 6ꢀ
chloroꢀ3,3ꢀdimethylꢀ2,3ꢀdihydroꢀ1Hꢀpyrrolo[3,2ꢀc]pyridine 31 (1.2 g, 6.9 mmol, see
Supporting Information) in dichloromethane (34 mL) were added PyBroP (7.7 g, 16.5 mmol)
and triethylamine (4.9 mL, 35.7 mmol). The reaction mixture was stirred at ambient
temperature for 18 hours. The solvent was removed in vacuo and the residue was purified by
chromatography on silica gel (gradient elution with 0 – 100% ethyl acetate in petrol) to give
the title compound (2.4 g, 73%) as a yellow foam. 1H NMR (400 MHz, CDCl3): 8.09 (1H, s),
8.07 (1H, s), 4.28ꢀ4.16 (1H, m), 4.06ꢀ3.85 (3H, m), 3.66 (2H, d), 3.56ꢀ3.42 (1H, m), 3.42ꢀ
3.34 (4H, m), 3.31 (1H, dd), 3.03 (1H, s), 2.86 (1H, dd), 2.67 (1H, dd), 1.58 (3H, s), 1.49
(9H, s), 1.42 (6H, s); LCMS: [M+H]+ = 467.
Step 2: (2R,5R)ꢀ4ꢀ[2ꢀ(6ꢀBenzylꢀ3,3ꢀdimethylꢀ2,3ꢀdihydroꢀpyrrolo[3,2ꢀc]pyridinꢀ1ꢀyl)ꢀ2ꢀoxoꢀ
ethyl]ꢀ5ꢀmethoxymethylꢀ2ꢀmethylꢀpiperazineꢀ1ꢀcarboxylic acid tertꢀbutyl ester
(2R,5R)ꢀ4ꢀ[2ꢀ(6ꢀChloroꢀ3,3ꢀdimethylꢀ2,3ꢀdihydroꢀpyrrolo[3,2ꢀc]pyridinꢀ1ꢀyl)ꢀ2ꢀoxoꢀethyl]ꢀ5ꢀ
methoxymethylꢀ2ꢀmethylꢀpiperazineꢀ1ꢀcarboxylic acid tertꢀbutyl ester (1.0 g, 2.14 mmol)
was added to a solution of PePPSiꢀiPr (40 mg, 0.06 mmol) and LiBr (390 mg, 4.54 mmol) in
NMP (12 mL) and THF (12 mL) under inert atmosphere. The solution was degassed for 5
min before adding benzylzinc bromide (9.1 mL, 1.6 mmol). The reaction was stirred at room
temperature for 30 minutes and then quenched with water. The product was extracted with
Page 29 of 45
ACS Paragon Plus Environment