Optimization of a series of 4,6-bis-anilino-1H-pyrrolo[2,3-d]pyrimidine inhibitors of IGF-1R: Elimination of an acid-mediated decomposition pathway

Article abstract of DOI:10.1016/j.bmcl.2008.11.065

Initial evaluation of a series 4,6-bis-anilino-1H-pyrrolo[2,3-d]pyrimidines revealed a C(1′) carboxamide was preferred for sub-micromolar in vitro potency against IGF-1R. Subsequent solution stability studies with 1 revealed a susceptibility toward acid-i

Full text of DOI:10.1016/j.bmcl.2008.11.065

Bioorganic & Medicinal Chemistry Letters 19 (2009) 373–377
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Optimization of a series of 4,6-bis-anilino-1H-pyrrolo[2,3-d]pyrimidine
inhibitors of IGF-1R: Elimination of an acid-mediated decomposition pathway
Stanley D. Chamberlain a, Anikó M. Redman a, Samarjit Patnaik a, Keith Brickhouse d, Yen-Chiat Chew d,
Felix Deanda b, Roseanne Gerding a, Huangshu Lei a, Ganesh Moorthy c, Mark Patrick d, Kirk L. Stevens a,
Joseph W. Wilson a, J. Brad Shotwell a,
*
a GlaxoSmithKline, Oncology R&D, 5 Moore Drive, PO Box 13398, Research Triangle Park, NC 27709-3398, USA
b GlaxoSmithKline, Computational and Structural Chemistry, 5 Moore Drive, PO Box 13398, Research Triangle Park, NC 27709-3398, USA
c GlaxoSmithKline, Oncology R&D, 1250 S. Collegeville Road, Collegeville, PA 19426, USA
d GlaxoSmithKline, Physical Properties and Developability, 5 Moore Drive, PO Box 13398, Research Triangle Park, NC 27709-3398, USA
a r t i c l e i n f o
a b s t r a c t
Initial evaluation of a series 4,6-bis-anilino-1H-pyrrolo[2,3-d]pyrimidines revealed a C(10) carboxamide
was preferred for sub-micromolar in vitro potency against IGF-1R. Subsequent solution stability studies
with 1 revealed a susceptibility toward acid-induced intramolecular cyclization with the C(10) carboxam-
ide. Herein, we describe several successful approaches toward generating both potent and acid-stable
inhibitors of IGF-1R within the 4,6-bis-anilino-1H-pyrrolo[2,3-d]pyrimidine template.
Ó 2008 Elsevier Ltd. All rights reserved.
Article history:
Received 20 October 2008
Revised 18 November 2008
Accepted 19 November 2008
Available online 24 November 2008
Keywords:
IGF-1R
Pyrrolopyrimidine
IGF-IR
Kinase
The activation of IGF-1R in many human cancers coupled with
the observation that inhibition of IGF-1R signaling results in a de-
creased proliferation of tumor cells in vitro and in vivo has made it
an attractive oncology target.1 As such, the optimization of inhibi-
tors of IGF-1R has been reported by multiple groups in diverse
chemical space.2 We have previously disclosed the synthesis and
optimization of a series of 4,6-bis-anilino-1H-pyrrolo[2,3-d]pyrim-
idines wherein a C(10) carboxamide, which interacts with Asp1150
through a water-mediated hydrogen bond, was believed a prere-
quisite for potent inhibition of IGF-1R.3,4
Subsequent to our initial biological characterization of 1, solu-
tion stability studies revealed a tendency toward decomposition
in acidic media. Further investigation identified tetracycle 2 and
acid 3 as the principle degradants (Scheme 1). Tetracycle 2 ap-
peared to account for the unusual sensitivity of the carboxamide
to mildly acidic media. In enzymatic estimates of IGF-1R potency,
2 and 3 showed reductions in potency relative to 1 of 20- and 4-
fold, respectively. With the solution instability of 1 well-character-
ized, the design of potent analogs not suffering from this liability
became a key medicinal chemistry aim.
istered tool for ongoing drug-discovery efforts relating to IGF-1R.
Isolation of authentic standards of 2 and 3 allowed their in vivo
pharmacokinetics to be investigated. Tetracycle
2 exhibited
slightly better exposure than parent 1 when dosed as a suspension
in rat (Table 1), and accounted for about 10% of the total AUC when
1 was dosed as an oral suspension (Table 1). Carboxylic acid 3
showed no oral exposure, was rapidly cleared when dosed as an
iv solution, and could not be detected in plasma following oral dos-
ing of 1 or 2. As a consequence of these data, medicinal chemistry
efforts were focused on accessing potent analogs which would be
more resistant toward in vivo hydrolysis in the stomach and espe-
cially in pharmaceutical formulations with pH <4.
Suspecting the electrophilicity of the carboxamide might corre-
late with decomposition rate (i.e., 1 ? 2 ? 3), we measured the
half-life for decomposition of parent (0.1 N HCl, 25 °C, LC–MS) as
Table 1
In vivo rat DMPK properties for 1–3 (data represents an average of three animals)
Compound
Dose
Cl (mL/min/kg)
DNAUC
%F (parent)
(mg/kg)
(ng h/mL/mg/kg)
The propensity toward hydrolysis did not prevent reasonable
oral exposure of 1, and as such allowed its use as an orally admin-
IV
PO
Parent
2
3
1
2
3
2.1
2.2
1.4
9.2
9.8
9.8
51
47
>200
156
203
0
18
—
0
0
0
0
65
81
0
* Corresponding author.
E-mail address: JBS26900@GSK.com (J. B. Shotwell).
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.11.065

374
S. D. Chamberlain et al. / Bioorg. Med. Chem. Lett. 19 (2009) 373–377
F
OH
F
NH2
1'
1'
O
H
O
H
N
N
F
N
N
1'
H2O
3
a.
4
3
4
3
O
N
N
N
N 1
H
6
N
N 1
H
6
N
N
N 1
H
6
N
N
O
O
O
1
2
3
N
N
N
N
N
N
Scheme 1. Acid-induced intramolecular cyclization of 1. Conditions: (a) 0.1 N HCl (aq), ambient temperature, t1/2 = 17 h.
Table 2
analogs, dihydro-3H-indazole-3-one 11, indoline 12, and quinoline
13 retained reasonable potency at the enzyme level, but proved
much less promising in the context of cellular estimates of IGF-
1R inhibition (Table 3). Similarly, 9, 10, and 14–16 suffered from
10- to 100-fold losses in potency relative to 1 in both the IGF-1R
enzyme and cellular assays. A suitably potent IGF-1R inhibitor
lacking a carboxamide was not identified in the context of the iso-
propylpiperazine tail.
As we considered possible alternative strategies toward
improving the stability within our series, a past synthetic problem
piqued our interest. Specifically, we had found that although thio-
phene 17 added smoothly to key dichloro intermediate 8, the sub-
sequent formation of key, activated intermediate 21 and C(6)
chloride displacement by aniline 19 does not occur to any appre-
ciable extent upon exposure to protic or Lewis acids (i.e.,
18 ? 22, Scheme 2), in direct contrast to the corresponding benz-
amides.3 Although this unanticipated stability to acid was difficult
to rationalize, we reasoned the same stability to acid that makes 18
an undesirable synthetic intermediate might ultimately impart
acid stability on analogs related to it. We developed a modified
synthetic protocol to allow access to heterocyclic 23–28, which
was routed through activated intermediate 21, accessible via a
non-acidic cyclization event (Scheme 2).
Rates of decomposition (0.1 N HCl (aq) at 23 °C) and potency for substituted
carboxamides
H
H
R2
N
N
N
R3
R4
O
H
N
4
N
N
1
N
H
6
N
H
O
Compound
R2
R3
R4
t
(h)
IGF-1R enzyme IC50 (nM)
½
1
4
5
6
7
F
H
F
H
F
H
H
F
H
F
17
2
4
17
39
2.0
0.9
0.7
1.3
0.8
H
H
F
F
H
Assay conditions used have been previously described and represent an average of
P2 independent measurements (see Ref. 4).
a function of substitution about the C(4) aniline (Table 2). Adding
an electronegative fluorine at R2 afforded half-lives 5- to 10-fold
longer than identical substitution at R3 or R4. We hypothesized a
hydrogen-bonding network including electronegative R2 might ac-
count for the increased activation barrier toward hydrolysis for 1,
6, and 7. Disubstitution (7) appeared to afford an additive impact
on elongating t1/2, but the overall impact of various substituents
on R2, R3, and R4 was subtle. All compounds (1 and 4–7)
shared an unacceptable propensity (t1/2 < 100 h) toward
decomposition.
Acid-catalyzed addition of thiophene 20 to 8 was followed by
base-mediated saponification of the methyl ester and exposure
to oxalyl chloride to give key, activated intermediate 21. Displace-
ment of the C(6) chloride of isolated 21 with aniline 19 proceeded
smoothly to afford 22. Finally, ring-opening with ammonium
hydroxide in THF/H2O in a sealed tube followed by base-mediated
hydrolysis of the tosylate protecting group afforded 23 in reason-
able overall yield. This route allowed us to both bypass acid-stable
but synthetically useless 18 and access primary carboxamides de-
rived from amino-thiophenes as final analogs.
The most straightforward improvement in cyclization half-life
would be anticipated to arise from removal of the offending car-
boxamide all-together. This proved a unique synthetic challenge,
as the syntheses of 1 and 4–7 necessitated the design of an internal
carboxamide-mediated activation strategy routed through the
intermediacy of cyclized intermediates related to 22 (vide infra).
This synthetic approach and its limitations has been described in
detail elsewhere.3 Ultimately we acknowledged that our chemis-
tries for the preparation of 1 and 4–7 had evolved to rely on the
very carboxamide functionality causing stability problems in acidic
media. However, using a Buchwald–Hartwig approach,5 we found
we could prepare milligram quantities of a variety of analogs
(Table 3) lacking a carboxamide in very modest yields. Of these
The reduced reactivity of 18 was found to correlate with desir-
able stability properties for 23–28. Gratifyingly, 23–28 showed a
tremendously reduced propensity toward acid-induced decompo-
sition in 0.1 N HCl despite possessing a primary carboxamide.
Additionally, no reduction in potency was observed via direct sub-
stitution of an amino-thiophene for the C(4) aniline. In fact 26–28
proved to be the most potent inhibitors of IGF-1R at both the cel-
lular and enzyme level prepared during the course of our efforts
(Table 4).
With the recognition that a C(4) heteroaryl amine could impart
acid-stability in the context of exceptional potency, we undertook
a more systematic investigation of related modifications at C(4).
Furans (31 and 32), regioisomeric thiophenes (30), and pyridines

S. D. Chamberlain et al. / Bioorg. Med. Chem. Lett. 19 (2009) 373–377
375
Table 3
NH2
H2N
O
NH2
IGF-1R enzyme and phospho cellular IGF-1R IC50 results for 1, 3, and 9–16
S
O
O
N
N
S
N
X
N
N
R
N
19
H2N
N
N
17
N
HCl, TFE,
N
100 ºC, 5 Days
TFA, TFE, 80 ºC
(<10% conversion)
85
N
H
N
H
(65% Yield)
Cl
O
ts
18
Compound
R
X
IGF-1R
PhosphoI GF-1R
enzyme IC50
(nM)
Cellular IC50 (nM)
1. TFA, TFE
O
S
O
NH2
F
S
H2N
N
20
O
1
2.0
7.9
117
2. LiOH/THF/H2O
3. (COCl)2, THF
O
N
HN
OH
N
N
N
(46% Yield, 3 steps)
ts
F
O
S
O
2
30,000
N
22
THF, 80 ºC,
HN
N
N
19, 2 h
O
N
N
NC
HN
Cl
N
ts
9
10
11
79
32
6.0
2745
982
21
O
1. 27% NH4OH
THF/H2O
O
NH2
S
2. 2.0 N NaOH
HN
H
N
H2O/dioxane
N
N
O
(22% Yield, 3 steps)
N
546
N
H
N
N
HN
N
O
H
23
Scheme 2. General synthesis for heterocyclic C(4) anilines containing a carbox-
12
13
10
13
1770
773
amide activating group.
N
N
N
(35 and 36) all proved more potent at both the enzyme and cellular
level than benzamide 29 and could be synthesized in analogy to
23–28. Additionally, these analogs showed a >100-fold improve-
ment in acid stability over benzamide 29 with the exception of
pyridyl analogs 35 and 36 (Table 5). Additionally, constrained lac-
tam 37 retained both remarkable potency and stability. Finally,
indolines 33 and 34 proved much more potent than initial indoline
12, wherein the improvement in potency is likely attributable to a
specific bidentate interaction between the N,N-dimethyl glycine
moiety and Asp1056.4
In summary, a key developability issue was observed for po-
tent IGF-1R inhibitor 1 wherein an acid-mediated cyclization of
the pyrimidine moiety onto the pendant carboxamide led to
facile hydrolysis in vitro and in vivo. Remarkable improvements
in both inhibitor stability and potency were realized via substi-
tution at C(4) with carboxamide-containing 5-membered het-
eroaryl amines (23–28 and 30–32), a constrained lactam (37),
or indolines (33 and 34).6 Further biological characterization
and in vivo pharmacokinetics of this subclass of exceptionally
potent and acid-stable inhibitors of IGF-1R will be reported in
due course.
N
14
15
16
250
400
25
2534
1695
654
N
N
N
H
N
N
HN
Assay conditions used have been previously described and represent the average of
P2 independent measurements (see Ref. 4).

376
S. D. Chamberlain et al. / Bioorg. Med. Chem. Lett. 19 (2009) 373–377
Table 4
Table 5
IGF-1R enzyme and phospho cellular IGF-1R IC50 results for 23–28 (data represents
P2 independent measurements)
IGF-1R enzyme and phospho cellular IGF-1R IC50 results for 29–37 (data represents
P2 independent measurements)
R
NH2
4
N
O
S
H
N
H
N
N
N
H
O
4
O
N
N
N
N
R
N
X
H
Compound
R
X
t
(h)
IGF-1R
enzyme
IC50 (nM)
PhosphoI
GF-1R Cellular
IC50 (nM)
½
Com-
pound
R
t
(h)
IGF-1R
enzyme
PhosphoI
GF-1R
½
IC50 (nM) Cellular
IC50 (nM)
N
F
N
O
O
29
CH2
2
1
2
54
25
N
23
N
N
>1000 1.6
106
N
NH
N
O
O
O
30
CH2
390
S
N
24
25
26
430 5.0
920 1.3
250 0.3
92
52
10
N
NH2
N
SO2Me
O
N
O
31
32
CH2
CH2CH2
>1000
>1000
0.8
0.3
28
9
N
N
N
N
33
34
CH2
CH2CH2
>1000
>1000
0.8
0.5
160
15
O
O
N
O
N
N
N
N
O
O
35
36
CH2
CH2CH2
15
27
0.7
0.5
38
18
N
N
O
27
830 <0.2
6
N
N
37
CH2
>1000
0.8
16
N
N
t1/2 measured at 23 °C in 0.1 N HCl.
O
O
28
720 0.2
<2
N
N
t1/2 measured at 23 °C in 0.1 N HCl.

S. D. Chamberlain et al. / Bioorg. Med. Chem. Lett. 19 (2009) 373–377
377
Hassell, A.; Lei, H.; Mook, R.; Moorthy, G.; Rowand, J.; Stevens, K.; Kumar, R.;
Shotwell, J. Bioorg. Med. Chem. Lett. 2008. doi:10.1016/j.bmcl.2008.11.046; (b)
Chamberlain, S.; Redman, A.; Wilson, J.; Deanda, F.; Shotwell, J.; Gerding, R.; Lei,
H.; Yang, B.; Stevens, K.; Hassell, A.; Shewchuk, L.; Leesnitzer, M.; Sabbatini, P.;
Atkins, C.; Groy, A.; Rowand, J.; Kumar, R.; Mook, R.; Moorthy, G.; Patnaik, S.
Bioorg. Med. Chem. Lett. 2008. doi:10.1016/j.bmcl.2008.11.077.
References and notes
1. Sachde, V. D.; Yee, D. Mol. Cancer Ther. 2007, 61, 1.
2. (a) García-Echeverría, C.; Pearson, M. A.; Marti, A.; Meyer, T.; Mestan, J.;
Zimmermann, J.; Gao, J.; Brueggen, J.; Capraro, H. G.; Cozens, R.; Evans, D. B.;
Fabbro, D.; Furet, P.; Porta, D. G.; Liebetanz, J.; Martiny-Baron, G.; Ruetz, S.;
Hofmann, F. Cancer Cell 2004, 53, 231; (b) Mulvihill, M. J.; Ji, Q. S.; Coate, H. R.;
Cooke, A.; Dong, H.; Feng, L.; Foreman, K.; Rosenfeld-Franklin, M.; Honda, A.;
Mak, G.; Mulvihill, K. M.; Nigro, A. I.; O’Connor, M.; Pirri, t C.; Steinig, A. G.; Siu,
K.; Stolz, K. M.; Sun, Y.; Tavares, P. A.; Yao, Y.; Gibson, N. W. Bioorg. Med. Chem.
2008, 163, 1359; (c) Ji, Q. S.; Mulvihil, M. J.; Rosenfeld-Franklin, M.; Cooke, A.;
Feng, L.; Mak, G.; O’Connor, M.; Yao, Y.; Pirritt, C.; Buck, E.; Eyzaguirre, A.;
Arnold, L. D.; Gibson, N. W.; Pachter, J. A. Mol. Cancer Ther. 2007, 68, 2158; (d)
Mulvihill, M. J.; Ji, Q. S.; Werner, D.; Beck, P.; Cesario, C.; Cooke, A.; Cox, M.;
Crew, A.; Dong, H.; Feng, L.; Foreman, K. W.; Mak, G.; Nigro, A.; O’Connor, M.;
Saroglou, L.; Stolz, K. M.; Sujka, I.; Volk, B.; Weng, Q.; Wilkes, R. Bioorg. Med.
Chem. Lett. 2007, 174, 1091; (e) Saulnier, M. G.; Frennesson, D. B.; Wittman, M.
D.; Zimmermann, K.; Velaparthi, U.; Langley, D. R.; Struzynski, C.; Sang, X.;
Carboni, J.; Li, A.; Greer, A.; Yang, Z.; Balimane, P.; Gottardis, M.; Attar, R.; Vyas,
D. Bioorg. Med. Chem. Lett. 2008, 185, 1702; (f) Velaparthi, U.; Liu, P.;
Balasubramanian, B.; Carboni, J.; Attar, R.; Gottardis, M.; Li, A.; Greer, A.;
Zoeckler, M.; Wittman, M. D.; Vyas, D. Bioorg. Med. Chem. Lett. 2007, 1711, 3072;
(g) Velaparthi, U.; Wittman, M.; Liu, P.; Stoffan, K.; Zimmermann, K.; Sang, X.;
Carboni, J.; Li, A.; Attar, R.; Gottardis, M.; Greer, A.; Chang, C. Y.; Jacobsen, B. L.;
Sack, J. S.; Sun, Y.; Langley, D. R.; Balasubramanian, B.; Vyas, D. Bioorg. Med.
Chem. Lett. 2007, 178, 2317; (h) Bell, I. M.; Stirdivant, S. M.; Ahern, J.; Culberson,
J. C.; Darke, P. L.; Dinsmore, C. J.; Drakas, R. A.; Gallicchio, S. N.; Graham, S. L.;
Heimbrook, D. C.; Hall, D. L.; Hua, J.; Kett, N. R.; Kim, A. S.; Kornienko, M.; Kuo, L.
C.; Munshi, S. K.; Quigley, A. G.; Reid, J. C.; Trotter, B. W.; Waxman, L. H.;
Williams, T. M.; Zartman, C. B. Biochemistry 2005, 4427, 9430.
5. Analogs lacking a carboxamide could not be prepared as previously described; 9
was accessed by direct dehydration of the corresponding carboxmide with TFAA
followed by neutralization with K2CO3/MeOH; 10–17 were accessed by initial
base-mediated displacement with R1NH2 at C(4) followed by Buchwald–
Hartwig amination using R2NH2 at C(6) and base-mediated detosylation.
1. R1NH2, DIPEA,
iPrOH
R1
2. Pd2(dba)3, BuOH, xantphos,
t
Cl
N
N
Na2CO3, uW, 150 ºC, R2NH2
3. NaOH, THF/H2O, uW, 100 ºC
N
N
R2
N
Cl
N
(5-20% Yield, 3 steps)
N
H
N
H
ts
8
9-16
6. The overall kinase selectivity for 1, 4–7, 9–16, and 23–37 is relatively unchanged
from the key molecules highlighted in Refs. 4,5. For example: (a) thiophene
carboxamide 26 has a measured IC50 P1
IGF-1R, IR, and ALK giving IC50s of 0.3, 0.4, and 0.4 nM respectively; (b) furan
carboxamide 31 has a measured IC50 P 1 M for 34 of 36 kinases tested, with
IGF-1R and IR giving IC50s of 0.8 and 0.8 nM, respectively (ALK not tested); (c)
constrained lactam 37 has a measured IC50 of P1 M for 35 of 38 kinases tested
with IGF-1R, IR, and ALK giving IC50s of 0.8, 0.8, and 2.5 nM, respectively. For 26,
31, and 37 measured IC50s for PI3K, GSK3, and AKT were all >10 M.
lM for 41 of 44 kinases tested, with
l
3. Chamberlain, S.; Lei, H.; Moorthy, G.; Patnaik, S.; Gerding, R.; Redman, A.;
Stevens, K.; Wilson, J.; Yang, B.; Shotwell, J. J. Org. Chem. 2008. doi:10.1021/
jo8020693.
4. (a) Chamberlain, S.; Wilson, J.; Deanda, F.; Patnaik, S.; Redman, A.; Yang, B.;
Shewchuk, L.; Sabbatini, P.; Leesnitzer, M.; Groy, A.; Atkins, C.; Gerding, R.;
l
l