Synthesis of potent and highly selective nonguanidine azetidinone inhibitors of human tryptase

Article abstract of DOI:10.1016/j.bmcl.2004.02.011

Azetidinones such as BMS-363131 (2) and BMS-363130 (3), which contain a guanidine group in the C-3 side chain were previously shown to be very potent inhibitors of human tryptase with high selectivity versus other serine proteases, including trypsin. In this letter, we describe the discovery of a number of potent azetidinone tryptase inhibitors in which the guanidine moiety at the ring C-3 position is replaced with primary or secondary amine or aminopyridine functionality. In particular, BMS-354326 (4) is a highly potent tryptase inhibitor (IC₅₀=1.8nM), which has excellent selectivity against trypsin and most other related serine proteases.

Full text of DOI:10.1016/j.bmcl.2004.02.011

Bioorganic & Medicinal Chemistry Letters 14 (2004) 2227–2231
Synthesis of potent and highly selective nonguanidine azetidinone
inhibitors of human tryptase
Gregory S. Bisacchi,* William A. Slusarchyk, Scott A. Bolton, Karen S. Hartl,
Glenn Jacobs, Arvind Mathur, Wei Meng, Martin L. Ogletree, Zulan Pi, James C. Sutton,
Uwe Treuner, Robert Zahler, Guohua Zhao and Steven M. Seiler
The Bristol-Myers Squibb Pharmaceutical Research Institute, PO Box 5400, Princeton, NJ08543-5400, USA
Received 17 December 2003; revised 1 February 2004; accepted 4 February 2004
Dedicated to the memory of Steven M. Seiler, Ph.D. (deceased March 31, 2003). SteveÕs untimely passing will not diminish the
continuing impact of his drug discovery research or the admiration of his fellow co-workers
Abstract—Azetidinones such as BMS-363131 (2) and BMS-363130 (3), which contain a guanidine group in the C-3 side chain were
previously shown to be very potent inhibitors of human tryptase with high selectivity versus other serine proteases, including trypsin.
In this letter, we describe the discovery of a number of potent azetidinone tryptase inhibitors in which the guanidine moiety at the
ring C-3 position is replaced with primary or secondary amine or aminopyridine functionality. In particular, BMS-354326 (4) is a
highly potent tryptase inhibitor (IC50 ¼ 1.8 nM), which has excellent selectivity against trypsin and most other related serine pro-
teases.
Ó 2004 Elsevier Ltd. All rights reserved.
The serine protease tryptase, which is the major protein
component of mast cells, is released along with hist-
amine, chymase, and other mediators of inflammatory
and allergic responses when mast cells degranulate upon
stimulation. Tryptase has been implicated in inflamma-
tory and allergic diseases and is believed to play a sig-
nificant role in asthma.1 Consequently, inhibition of
tryptase has become an important therapeutic target for
asthma and other inflammatory and allergic conditions.2
We previously reported the straight-chained terminal
guanidine compound BMS-262084 (1), which displayed
potent inhibition of tryptase (IC50 ¼ 4 nM) and moder-
ate to good selectivity against related serine proteases
with the exception of trypsin. We also demonstrated the
efficacy of BMS-262084 upon intratracheal dosing in
guinea pig models of bronchoconstriction and lung
inflammation.3 In a separate report, we described the
conformationally constrained guanidine epimers BMS-
363131 (2) and BMS-363130 (3), both of which dis-
played potent inhibition of tryptase (IC50<1.7 nM) with
excellent selectivity against other serine proteases
including trypsin. BMS-363131, upon intratracheal
dosing, was also efficacious in a guinea pig model of
lung inflammation.4
HN
CO2H
O
H2N
N
H
N
H
N
N
O
N
O
BMS-262084 (1)
R or S
H
CO2H
O
Ph
N
N
O
N
N
O
HN
NH2
R-isomer BMS-363131 (2)
S-isomer BMS-363130 (3)
CO2H
N
O
HN
H
Ph
N
N
O
O
BMS-354326 (4)
Because a significant portion of an inhaled drug is
swallowed, and considering the potential for preclinical
* Corresponding author. Tel.: +1-609-818-5229; fax: +1-609-818-3450;
e-mail: gregory.bisacchi@bms.com
0960-894X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.02.011

2228
G. S. Bisacchi et al. / Bioorg. Med. Chem. Lett. 14 (2004) 2227–2231
toxicity associated with trypsin inhibition,5 selectivity
against the digestive protease trypsin (>1000-fold) was
considered an essential requirement for an inhaled drug
candidate. We were also interested in exploring whether
this guanidine-containing chemotype might be modified
for oral delivery through replacement of the guanidine
moiety with less basic functionality.
Table 2. Tryptase inhibition for guanidine and nonguanidine com-
pounds containing the N-1 group derived from 12
R1
O
CO2H
iPr
iPr
N
N
N
O
O
O
R1
Tryptase IC50 (nM)
NH
In this letter, we report on our efforts to search for
nonguanidine compounds similar to BMS-262084 and
BMS-363131. These efforts led to the discovery of a
number of potent tryptase inhibitors in which a primary/
secondary amine or aminopyridine functionality replace
the guanidine functionality. In particular, we identified
BMS-354326 (4), which is a highly potent tryptase
inhibitor (IC50 ¼ 1.8 nM) having excellent selectivity
against trypsin and most other related serine proteases.
19
4
H2N
H2N
NH
NH
20
31
21
22
23
>33,000
NH
39
7
H2N
H2N
H2N
The syntheses of the compounds shown in Tables 1–3
are described in Schemes 1 and 2, with the exception of
reference compounds 193 and 326, which are described
elsewhere. As indicated in Scheme 1, (S)-azetidinone 5
was alkylated with Boc-protected halides 6a–k in the
presence of LDA to afford intermediates 7. N-Desilyl-
ation, followed by benzylation of the carboxylate
afforded intermediates 8. The preparation of interme-
diates 8l and 8m were described previously.4 The prep-
aration of several of the starting halides 6 is also shown
in Scheme 1. Iodides 6c–f were synthesized by reacting
the corresponding chloro iodo-n-alkanes with potassium
di-t-butyl iminodicarbonate followed by treatment with
sodium iodide in acetone. The aminopyridine bromides
6g and 6h were prepared by bis-Boc protection of
2-amino-5-picoline and 2-amino-4-picoline, respectively,
followed by mono bromination of the methyl groups.
Reaction of a,a0-dibromo-para- and meta-xylenes with
potassium di-t-butyl iminodicarbonate gave 6i and 6j,
respectively. Iodides 6a, 6b, and 6k were prepared by
literature procedures.7
34
32
H2N
24
17,899
N
H
25
26
2605
1214
HN
HN
27
1053
H2N
N
N
28
29
61
H2N
211
H2N
H2N
30
19
Table 1. Tryptase inhibition for compounds containing a 3-(4-pipe-
ridinylmethyl) group
R1
CO2H
N
R2
O
Table 3. Tryptase inhibition for nonguanidine compounds containing
the N-1 group derived from 33
O
R1
R2
Tryptase
IC50 (nM)
R1
O
CO2H
Ph
N
N
N
NH
H
N
O
14
515
O
H2N
N
R1
Tryptase IC50 (nM)
H
N
15
16
147
30
HN
HN
HN
32
34
1.0
O
O
<1.7
N
N
N
N
N
H
tBu
H2N
HN
iPr
iPr
17
29
HN
4 (BMS-354326)
1.8
O
H

G. S. Bisacchi et al. / Bioorg. Med. Chem. Lett. 14 (2004) 2227–2231
2229
CO2H
NTBS
CO2Bn
NH
CO2H
NTBS
R
O
R
b,c
a
halide 6
O
O
7
5
8
halide 6
R for 7 and 8
n
I
n
N
N
Boc
Boc
6a n = 1
6b n = 2
7a, 8a n = 1
7b, 8b n = 2
d, e
n
n
n
Cl
I
Boc2N
I
Boc2N
6c n = 1
6d n = 2
6e n = 3
6f n = 4
7c, 8c n = 1
7d, 8d n = 2
7e, 8e n = 3
7f, 8f n = 4
CH3
CH3
Br
f
f
N
N
N
H2N
Boc2N
Boc2N
6g
7g, 8g
Br
N
N
N
NH2
NBoc2
NBoc2
6h
7h, 8h
Br
Br
Br
d
d
Br
Boc2N
Boc2N
7i, 8i
6i
6j
Br
Br
Boc2N
Boc2N
7j, 8j
N
N
Boc
I
Boc
6k
7k, 8k
n
n
I
N
N
Boc
6l n = 1
6m n = 2
Boc
8l n = 1
8m n = 2
Scheme 1. (a) LDA, THF; (b) TBAF, THF; (c) BnBr, NaHCO3; (d) Boc2NK, DMF; (e) NaI, acetone; (f) (i) Boc2O, CH2Cl2, DMAP, iPr2EtN; (ii)
NBS, benzoyl peroxide, CCl4.
As shown in Scheme 2, compound 8a was N-acylated
with phenylisocyanate in the presence of LDA to afford
9. Removal of the Boc group of 9 with TFA, followed
by conversion of the piperidine amine to a bis-CBZ
protected guanidine function using N,N0-bis-CBZ-1-
guanylpyrazole, and finally reductive removal of the
benzyl and CBZ groups afforded 14. The corresponding
nonguanidine piperidine analog 15 was prepared by
sequential deprotection of the Boc and benzyl groups of
9. Intermediate 8a was also converted to compounds 16
and 17 by reaction with the chlorocarbonyl reagents 10
and 12 in the presence of TEA/DMAP, followed by
removal of the Boc and benzyl groups. Intermediates
8b–j, 8l, and 8m were acylated with 12, and afforded
compounds 20–30 following deprotection. Compound
31, the guanidine analog of 21, was prepared from 18d
by sequential removal of the Boc group, reaction of the
primary amine with N,N0-bis-CBZ-1-guanylpyrazole,
and reductive removal of the CBZ and benzyl groups.
Intermediates 8d and 8k were converted to compounds
34 and 4 by N-acylation with the chlorocarbonyl reagent
33, followed by deprotection of the Boc and benzyl
groups.
During the course of our SAR studies of guanidine-
containing azetidinone tryptase inhibitors we prepared
compound 14, a conformationally constrained guani-
dine analog of one of our early lead compounds. While

2230
G. S. Bisacchi et al. / Bioorg. Med. Chem. Lett. 14 (2004) 2227–2231
e
d,
,
c
14
15
CO2Bn
c,e
N
N
NH
O
Boc
O
9
a
CO2Bn
O
10
c,e
8a
tBu
b
N
N
16
17
N
O
N
N
H
Boc
O
12
b
11
CO2Bn
c,e
c,e
O
iPr
N
N
N
O
N
O
iPr
Boc
O
13
CO2Bn
12
b
R
O
iPr
iPr
8b-8j,
8l & 8m
20-30
N
N
O
N
O
O
18b-j, 18l, 18m
(R groups correspond
to Scheme 1)
CO2Bn
O
Boc2N
iPr
c,d,e
31
N
N
N
O
iPr
O
O
18d
c,e
CO2Bn
R
O
O
33
Ph
5
8d and 8k
34 and 4
N
N
O
N
b
19d and 19k
iPr
iPr
O
O
O
O
N
O
O
Ph
tBu
Cl
12
N
N
O
Cl
N
Cl
10
N
N
N
H
5
33
Scheme 2. (a) LDA, THF, phenylisocyanate; (b) TEA, DMAP, DMF; (c) TFA, CH2Cl2; (d) TEA, N,N0-bis-CBZ-1-guanylpyrazole; (e) Pd/C, H2,
dioxane.
compound 14 had only modest potency (IC50 ¼ 515 nM),
we were intrigued when we discovered that the corre-
sponding nonguanidine piperidine analog 15 showed
improved potency (IC50 ¼ 147 nM). Keeping the C-3
4-(piperidinylmethyl) substituent constant, we next
prepared analog 16, which contained the t-butyl-
aminocarbonylpiperizinylcarbonyl N-1 group found in
BMS-262084 (1), as well as analog 17, containing the
diisopropylmethoxypiperizinylcarbonyl N-1 group.
These two N-1 groups had previously been recognized
to be superior to the phenylaminocarbonyl N-1 group in
the C-3 guanidinylpropyl series, and in the C-3 pipe-
ridine series afforded more potent tryptase inhibitors: 16
displayed an IC50 of 30 nM and 17 displayed an IC50 of
29 nM. The selectivity of 17 was profiled in a panel of
related serine proteases (Table 4). While 17 is highly
selective versus thrombin, factor Xa, uPA, and tPA and
is modestly selective for plasmin, it is poorly selective for
trypsin (32-fold).
Table 4. Tryptase inhibition and selectivities for 17 and 4
17
4 (BMS-354326)
IC50 (nM) Selectivity
IC50 (nM)
Selectivity
Tryptase
Trypsin
Thrombin
FXa
29
935
––
32X
1.8
––
5583X
10,005
>33,000
>33,000
309
>33,000
>33,000
6220
>1138X
>1138X
214X
>18,333X
>18,333X
169X
Plasmin
uPA
tPA
6999
>33,000
241X
>1138X
>33,000
>33,000
>18,333X
>18,333X

G. S. Bisacchi et al. / Bioorg. Med. Chem. Lett. 14 (2004) 2227–2231
2231
Keeping the diisopropylmethoxypiperizinylcarbonyl
N-1 group constant, we compared two additional gua-
nidine-containing beta-lactams with their corresponding
nonguanidine counterparts, namely 19 and 20, having
guanidinylpropyl and aminopropyl C-3 substituents,
respectively, and 31 and 21, having guanidinylbutyl and
aminobutyl C-3 substituents, respectively (Table 2).
Whereas guanidine 19 is quite potent (IC50 ¼ 4 nM), the
corresponding amino compound 20 completely lacks
potency. In contrast, the analogs in the N-butyl series
are both fairly potent, with the amino analog 21 sixfold
more potent than the corresponding guanidine 31
(IC50 ¼ 39 nM vs 7 nM, respectively). It was evident that
simple amine functionality at the azetidinone C-3 could
afford very potent tryptase inhibitors, and that the SAR
was different from the corresponding guanidine analogs
having the same alkyl linker. This latter observation is
not surprising given the assumption that both the gua-
nidine and the amine C-3 functional groups are salt-
bridging to aspartate 189 at the bottom of the tryptase
S1 pocket8, and that the proper placement of the positive
charge is a critical determinant of potency.
inhibitor (IC50 ¼ 1.8 nM), and, except for only moderate
selectivity versus plasmin (169-fold), the selectivity of
BMS-354326 versus a panel of related serine proteases is
excellent. In particular, it is over 5000-fold selective
against trypsin. By comparison the selectivity of 32
against trypsin is 310-fold.3 BMS-354326 also demon-
strated good aqueous hydrolytic stability, with a half-
life of >48 h at pH 7 and 35 h at pH 9.
In summary, we have identified a number of potent,
nonguanidine azetidinone inhibitors of human tryptase.
One of them, BMS-354326 has an IC50 for tryptase of
1.8 nM, and displays excellent selectivity against trypsin
and most other related serine proteases.
References and notes
1. (a) Abraham, W. M. Am. J. Physiol. Lung Cell Mol.
Physiol. 2002, 282, L193; (b) Krishna, M. T.; Chauhan, A.;
Little, L.; Sampson, K.; Hawksworth, R.; Mant, T.;
Kjukanovic, R.; Lee, T.; Holgate, S. J. Allergy Clin.
Immunol. 2001, 107, 1039; (c) Elrod, K. C.; Numerof, R. P.
Emerging Ther. Targets 1999, 3, 203.
2. (a) Costanzo, M. J.; Yabut, S. C.; Almond, H. R., Jr.;
Andrade-Gordon, P.; Corcoran, T. W.; de Garavilla, L.;
Kauffman, J. A.; Abraham, W. M.; Recacha, R.; Chatto-
padhyay, D.; Maryanoff, B. E. J. Med. Chem 2003, 46,
3865; (b) Newhouse, B. J. IDrugs 2002, 5, 682; (c) Burgess,
L . E.Drug News Perspect 2000, 13, 147.
3. Sutton, J. C.; Bolton, S. A.; Hartl, K. S.; Huang, M.-H.;
Jacobs, G.; Meng, W.; Ogletree, M. L.; Pi, Z.; Schumacher,
W. A.; Seiler, S. M.; Slusarchyk, W. A.; Treuner, U.;
Zahler, R.; Zhao, G.; Bisacchi, G. S. Bioorg. Med. Chem.
Lett. 2002, 12, 3229.
4. Slusarchyk, W. A.; Bolton, S. A.; Hartl, K. S.; Huang,
M.-H.; Jacobs, G.; Meng, W.; Ogletree, M. L.; Pi, Z.;
Schumacher, W. A.; Seiler, S. M.; Sutton, J. C.; Treuner,
U.; Zahler, R.; Zhao, G.; Bisacchi, G. S. Bioorg. Med.
Chem. Lett. 2002, 12, 3235.
Further survey of amino functionality at the azetidinone
C-3 position showed the aminopentyl (22, IC50 ¼ 34 nM)
and aminohexyl (23, IC50 ¼ 32 nM) analogs to be
somewhat less potent than aminobutyl 21. In contrast to
the 4-piperidinylmethyl compound 17, the 3-piperid-
inylmethyl (24), 2-pyrrolidinylmethyl (25), and 4-piper-
idinlyethyl (26) analogs were all weakly potent. Whereas
the 4-amino-3-pyridinylmethyl analog 27 was only
modestly potent (IC50 ¼ 1053 nM), the isomeric 3-amino-
4-pyridinylmethyl analog 28 had relatively good potency
(IC50 ¼ 61 nM). The pKa of 2-aminopyridine is around
7.0, and thus 28 is the least basic inhibitor of this series
having good potency. Finally, 4-aminomethylbenzyl
analog 29 had an IC50 of 211 nM while its 3-amino-
methyl analog 30 was 10-fold more potent (IC50 ¼ 19),
an interesting positional effect somewhat mirroring the
relative potencies of the aminopyridine isomeric pair, 27
and 28.
5. (a) Flavin, Dana F. Veterinary and Human Toxicology
1982, 24, 25; (b) Chan, J.; De Luman, B. O. I. J. Agric. and
Food Chem. 1982, 30, 46; (c) Ekpenyong, T. E.; Borchers,
R. L. Nutr. Rep. Int. 1981, 23, 865.
The more recent discovery of the potency- and selec-
tivity-enhancing properties of certain lipophilic N-1
groups such as the phenyl-n-pentylcarbonyl-piperizi-
nylcarbonyl group4;6 prompted us to survey select
nonguanidine analogs in that N-1 series. The C-3 4-
piperidinylmethyl analog 32 has been reported sepa-
rately6, and is shown in Table 3 for comparison. Thus,
32 is about 30-fold more potent than the corresponding
N-1 diisopropyl-methoxycarbonyl analog 17 (IC50Õs
1 nM versus 29 nM, respectively). Analog 34 containing
the C-3 aminobutyl group is also more potent than the
corresponding analog 21 (IC50Õs < 1.7 nM versus 7 nM,
respectively). Additionally we prepared an analog in this
N-1 series having a 3-piperidinylethyl group at C-3 (4,
BMS-354326), a group not surveyed in earlier N-1 ser-
ies. BMS-354326 was found to be a very potent tryptase
6. Sutton, J. C.; Bolton, S. A.; Davis, M. E.; Hartl, K. S.;
Jacobson, B.; Mathur, A.; Ogletree, M. L.; Slusarchyk, W.
A.; Zahler, R.; Seiler, S. M.; Bisacchi, G. S. Bioorg. Med.
Chem. Lett. 2004, 14, following paper in this issue.
doi:10.1016/j.bmcl.2004.02.012.
7. (a) Villalobos, A.; Blake, J. F.; Biggers, C. K.; Butler, T.
W.; Chapin, D. S.; Chen, Y. L.; Ives, J. L.; Jones, S. B.;
Liston, D. R.; Nagel, A. A.; Nason, D. M.; Nielsen, J. A.;
Shalaby, I. A.; White, W. F. J. Med. Chem 1994, 37, 2721;
(b) Lowen, G. T.; Almond, M. R.; Rideout, J. L.
J. Heterocycl. Chem 1992, 29, 1663.
8. (a) Pereira, P. J. B.; Bergner, A.; Macedo-Ribeiro, S.;
Huber, R.; Matschiner, G.; Fritz, H.; Sommerhoff, C.;
Bode, W. Nature 1998, 392, 306; (b) An X-ray crystal
structure of BMS-262084 (1) in trypsin (Ref. 6) showed a
classic salt bridge of the C-3 guanidine to the trypsin
Asp189.