Pyridone-conjugated monobactam antibiotics with gram-negative activity

Article abstract of DOI:10.1021/jm400560z

Herein we describe the structure-aided design and synthesis of a series of pyridone-conjugated monobactam analogues with in vitro antibacterial activity against clinically relevant Gram-negative species including Pseudomonas aeruginosa, Klebsiella pneumoniae, and Escherichia coli. Rat pharmacokinetic studies with compound 17 demonstrate low clearance and low plasma protein binding. In addition, evidence is provided for a number of analogues suggesting that the siderophore receptors PiuA and PirA play a role in drug uptake in P. aeruginosa strain PAO1.

Full text of DOI:10.1021/jm400560z

Subscriber access provided by UNIV OF ARIZONA
Article
Pyridone-Conjugated Monobactam Antibiotics with Gram-Negative Activity
Matthew F. Brown, Mark J. Mitton-Fry, Joel T Arcari, Rose Barham, Jeffrey Casavant, Brian S
Gerstenberger, Seungil Han, Joel R Hardink, Thomas M Harris, Thuy Hoang, Michael D Huband,
Manjinder S. Lall, M Megan Lemmon, Chao Li, Jian Lin, Sandra P McCurdy, Eric McElroy,
Craig McPherson, Eric S. Marr, John P. Mueller, Lisa Mulllins, Antonia A. Nikitenko, Mark C
Noe, Joseph Penzien, Mark S Plummer, Brandon P. Schuff, Veerabahu Shanmugasundaram,
Jeremy T. Starr, Jianmin Sun, Andrew P. Tomaras, Jennifer A. Young, and Richard P. Zaniewski
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/jm400560z • Publication Date (Web): 11 Jun 2013
Downloaded from http://pubs.acs.org on June 11, 2013
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 36
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
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 36
1
2
3
4
Pyridone-Conjugated Monobactam Antibiotics with Gram-Negative Activity
5
6
7
8
9
Matthew F. Brown,*a Mark J. MittonꢀFry,a Joel T. Arcari,a Rose Barham,a Jeffrey Casavant,a
Brian S. Gerstenberger,a Seungil Han,e Joel R. Hardink,d Thomas M. Harris,a Thuy Hoang,a
Michael D. Huband,c Manjinder S. Lall,a M. Megan Lemmon,c Chao Li,a Jian Lin,d Sandra P.
McCurdy,c Eric McElroy,a Craig McPherson,c Eric S. Marr,e John P. Mueller,c Lisa Mullins,c
Antonia A. Nikitenko,a Mark C. Noe,a Joseph Penzien,c Mark S. Plummer,a Brandon P. Schuff,a
Veerabahu Shanmugasundaram,b Jeremy T. Starr,a Jianmin Sun,a Andrew Tomaras,c Jennifer A.
Young,a Richard P. Zaniewskic
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
b
c
aWorldwide Medicinal Chemistry, Computational Chemistry, Antibacterials Research Unit,
e
dPharmacokinetics, Dynamics & Metabolism, Structural Biology, Pfizer Global Research and
Development, Eastern Point Road, Groton, CT 06340.
ABSTRACT
Herein we describe the structureꢀaided design and synthesis of a series of pyridoneꢀconjugated
monobactam analogs with in vitro antibacterial activity against clinically relevant Gramꢀnegative
species including Pseudomonas aeruginosa, Klebsiella pneumoniae and Escherichia coli. Rat
pharmacokinetic studies with compound 17 demonstrate low clearance and low plasma protein
binding. In addition, evidence is provided for a number of analogs suggesting that the
siderophore receptors PiuA and PirA play a role in drug uptake in P.aeruginosa strain PAO1.
INTRODUCTION
Infections caused by multidrugꢀresistant Gramꢀnegative bacteria, particularly in the hospital
setting, result in thousands of deaths per year and are the source of considerable concern in the
ACS Paragon Plus Environment

Page 3 of 36
Journal of Medicinal Chemistry
1
2
medical community1. One strategy to develop new antibacterial agents capable of eradicating
these pathogens involves the soꢀcalled “Trojan Horse” approach wherein the bacterial iron
acquisition system is utilized to deliver a drug to the periplasmic space, thus enabling target
engagement and cell death2. We and others have exploited this strategy to identify
monocarbams3 such as 1 (MCꢀ1)4, monosulfactams such as 2 (BAL30072)5 and monobactams6
such as isoxazole 37 (Figure 1). In compounds 1-3, the presence of a pyridone moiety is critical
to the active uptake process which likely involves drug interaction with one or more bacterial
siderophore receptors (vide infra)8. Our initial report of pyridoneꢀconjugated monobactam
analogs exemplified by 3 highlighted the potential of the series to provide molecules with
respectable in vitro and in vivo activity against Gramꢀnegative pathogens, including some
multidrugꢀresistant (MDR) strains. We describe herein a more chemically diverse set of analogs,
the design of which benefited from cocrystal structures of several ringꢀopened βꢀlactams with
Pseudomonas aeruginosa Penicillin Binding Protein 3 (PaePBP3).
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 1. Recently reported pyridoneꢀconjugated monocyclic βꢀlactams.
CHEMISTRY
Detailed experimental methods utilized in the preparation of all analogs, including the
intermediates described below, are available in the Supporting Information. A number of
analogs in Tables 1 and 4 were prepared as shown in Scheme 1. Enantiomer 4 was isolated from
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 36
1
2
the racemicꢀcis precursor9 by use of chiral resolution via supercritical fluid chromatography.
Ester reduction of 4 with NaBH4 and removal of the CBZ protecting group by hydrogenolysis
provided aminoalcohol 5. Coupling of amine 5 with the activated ester 9, which is derived from
the corresponding carboxylic acid10, provided alcohol 6. The hydroxyl group of this versatile
intermediate could be converted to the corresponding amine 7, thus providing access to a variety
of aminoꢀlinked compounds, including carbamate 8. Lactam Nꢀsulfation, typically with
pyridine•SO3, followed by global deprotection, typically with BCl3, provided final analogs such
as 18.
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
Preparation of the pyridone intermediates utilized in the synthesis of final analogs were generally
derived from kojic acid (10). For example, monoprotection of the phenolic oxygen of kojic acid
with 4ꢀmethoxybenzyl chloride followed by heating with hydroxylamine and subsequent reaction
with 4ꢀmethoxybenzyl chloride provided intermediate 11 which was utilized in the preparation of
18. The pyridoneꢀ2ꢀcarboxylic acids utilized in the preparation of a number of analogs in Table
4 involved protection of the kojic acid phenol with a benzyl group followed by oxidation of the
2ꢀhydroxymethyl moiety to a carboxylic acid with NaClO2/TEMPO(cat)/bleach(cat)11. Reaction
of this intermediate with the requisite amine then provided the pyridoneꢀ2ꢀcarboxylic acids 12.
Standard amide coupling methodology of 12 with amine 7 provided access to Table 4 analogs.
A number of the analogs with modified oxime substituents described in Tables 2 and 3 were
prepared as shown in Scheme 2. Condensation of Oꢀalkylhydroxylamines, such as 15, with
ketone 13 provided oxime acid 14. Standard amide coupling with amine 16 (see Supplementary
Information for preparation) followed by sulfation and global deprotection as described above
provided final targets, such as 27.
ACS Paragon Plus Environment

Page 5 of 36
Journal of Medicinal Chemistry
1
2
3
4
Scheme 1. Preparation of 18.
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
Reagents and conditions: (a) NaBH4, MeOH, 20 °C; (b) 20% Pd(OH)2/C, H2 (50 psi), glacial
HOAc, EtOH; (c) 9, NEt3, MeCN, RT; (d) PPh3, imidazole, I2, CH2Cl2, RT; (e) NBu4N3, NEt3, 2ꢀ
MeꢀTHF; (f) 10% Pd/C, H2 (30 psi), EtOH; (g) 11, CDI, THF; (h) pyridine•SO3, DMF, RT; (i)
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 36
1
2
3
4
5
6
7
8
9
BCl3, CH2Cl2, RT; (j) NaHCO3, water; (k) 4ꢀmethoxybenzyl chloride, NMP, K2CO3, 75 °C; (l)
(i) NH2OH•HCl, K2CO3, NMP, (ii) 4ꢀmethoxybenzyl chloride; (m) BnBr, NaOH, MeOH, reflux;
(n) TEMPO, NaClO2, NaClO, Naꢀphosphate buffer, CH3CN, 35 °C; (o) RNH2, MeOH, RT.
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. Method for preparing oxime modified analogs.
Reagents and conditions: (a) 15, MeOH, RT; (b) 16, HATU, NaHCO3, DMF, RT; (c)
pyridine•SO3, DMF, RT; (d) BCl3, CH2Cl2, RT.
RESULTS AND DISCUSSION
All analogs were screened for minimum inhibitory concentrations (MICs) against a primary
panel of Gramꢀnegative pathogens. The panel included strains susceptible to clinically relevant
drugs as well as some with a MDR phenotype (Tables 1ꢀ4). For example, P. aeruginosa strain
1091ꢀ05 and A. baumannii strain ABꢀ3167 are susceptible to antibacterial drugs from multiple
classes while K. pneumoniae strain 1000ꢀ02 is aztreonam resistant and meropenem sensitive. P.
aeruginosa strain 1042ꢀ06 and K. pneumoniae strain 1487ꢀ07 (KPCꢀ2, TEMꢀ1 and SHVꢀ12)
display moderate to high level resistance to all of the comparator agents.
ACS Paragon Plus Environment

Page 7 of 36
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The antibacterial mechanism of action of βꢀlactams involves acylation of an active site serine
residue from one or more penicillin binding proteins (PBPs) leading to enzyme inactivation,
which produces cell wall irregularities and ultimately, cell death. In general, monobactams, such
as aztreonam and the compounds described below, are potent inhibitors of Gramꢀnegative PBP3,
display moderate PBP1a and 1b inhibition and no relevant PBP2 inhibitory activity (Table 8)5a,
12. Given the potent PBP3 inhibitory activity of this class, our initial structure based drug design
efforts focused on understanding drug interactions with this enzyme. Indeed, a number of recent
publications from our group describe cocrystal structures of ringꢀopened βꢀlactams bound to
PaePBP3 and this work contributed significantly to the design of novel analogs described below.
For example, the xꢀray structures of monocarbam 1 and monobactam 3 demonstrated that the
linker motifs connecting the βꢀlactam cores to the pyridone moieties of both analogs occupy a
fairly sizable tunnel defined, in part, by Val333 and Phe5337, 13. A similar binding mode is
observed with the novel monobactam analog 17 bound to PaePBP3, with the urea linker
occupying this tunnel (Figure 2). Overall, this structural data suggests that a wide range of linker
groups could be accommodated in this tunnel region, and indeed, a number of new compounds
with potent antiꢀpseudomonal activity were discovered in this vein. For example, urea 17,
carbamate 18, triazoles 20 and 21 and 2ꢀaminopyridyl analog 22 displayed improved activity
relative to the comparator agents against the drug resistant P. aeruginosa strain 1042ꢀ06. In
addition, most analogs proved to be superior to the classꢀcomparator aztreonam against the K.
pneumoniae strain 1000ꢀ02 with a number of compounds matching the potency of meropenem
(17, 20, 21). In addition, urea 17 was effective against the MDR KPCꢀharboring K. pneumoniae
strain, 1487ꢀ07, and all analogs (and comparator drugs) in Table 1 provided moderately low
MICs against the susceptible A. baumannii strain ABꢀ3167. However, not all linkers provided
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
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 36
1
2
3
4
5
6
7
8
9
the desired level of potency and/or spectrum. For example, while urea 17 exhibited a promising
profile, the closely related sulfamide 19 showed diminished potency in the primary panel relative
to 17, especially vs. the drug resistant strains 1042ꢀ06 and 1487ꢀ07 and the pyridylꢀlinked analog
22 displayed inferior activity vs. the K. pneumoniae strain 1487ꢀ07.
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
While the PaePBP3 cocrystal structures catalyzed the design of the various linkers described
above, the enhanced cellular potency of the analogs relative to aztreonam is not likely due to
enhanced inhibition of PBP3. For example, compound 3 and aztreonam have similar enzyme
IC50s (Table 8), while urea 17 is approximately 10ꢀfold less potent. Both 3 and 17 display
improved PBP1a and 1b IC50s relative to aztreonam, which, when combined with the
siderophore uptake mechanism, likely explains the enhanced cellular potency of these analogs
relative to aztreonam.
Figure 2. Cocrystal structure of 17 with PaePBP3.
ACS Paragon Plus Environment

Page 9 of 36
Journal of Medicinal Chemistry
1
2
3
4
Table 1. C4 Linker modification.
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
P. aeruginosa
strains (ꢀg/mL)
K. pneumoniae
A. baumannii
Cmpd
Linker
strains (ꢀg/mL)
strain (ꢀg/mL)
1091-05 1042-06 1000-02 1487-07
AB-3167
aztreonam
4
64
32
64
64
8
>64
NT a
0.25
NT
>64
>64
64
4
2
cefepime
meropenem
ciprofloxacin
amikacin
2
2
0.25
0.06
1
0.125
2
>64
64
NT
3b
0.25
0.25
1
NT
0.125
0.25
0.25
1
2
2
0.25
4
2
0.25
0.125
1
17
18b
19
NT
32
16
4
0.125
0.125
0.25
NT
0.125
20
0.25
1
1
0.25
0.5
NT
16
0.25
21
22
0.125
0.125
a NT = not tested
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 36
1
2
b bisꢀNaꢀsalt
3
4
5
6
7
8
The βꢀlactam C3 carboxamide moiety in the analogs described above is found in a variety of βꢀ
lactam drugs (Figure 3). The recently published cocrystal structures of monocarbam 1,
aztreonam and ceftazidime with PaePBP3 illustrate multiple productive drugꢀprotein interactions
with this sidechain and learnings from these structures laid the groundwork for C3 monobactam
optimization efforts13. For example, in all cases, the cis-oxime linked gemꢀdimethyl group was
found to interact with a hydrophobic pocket formed by Tyr503, Tyr532, Phe533 and Val333, and
the carboxylic acid participates in a salt bridge interaction with Arg489. Analogous interactions
are observed in the cocrystal structure of PaePBP3 with urea 17 (Figure 2) and molecular
modeling studies (see experimental section for model details) suggest that increasing the size of
the gemꢀdialkyl group could potentially provide a better fit with the hydrophobic pocket
described above. Dimethyl substitution (17) was roughly similar to the desmethyl analog 23
with regard to P. aeruginosa MICs, and somewhat improved relative to the monoꢀalkyl analogs
24 and 25 (Table 2). Constraining the gem dialkyl substituents via ring formation provided
interesting SAR with the 4ꢀmembered ring analog 26 demonstrating roughly equivalent potency
and spectrum as compared to the dimethyl analog 17. Ring expansion to a cyclopentyl group
(27) provided no additional benefit while incorporation of a cyclohexyl group (28) led to a 4 to
8ꢀfold loss of potency across the primary strain set relative to the cyclobutyl analog 26, thus
providing some guidance regarding the preferred substituent size for this hydrophobic pocket.
The similar P. aeruginosa MICs observed for 26 and 17 translated to PBP enzyme inhibitory
activity as the two compounds display similar potencies against PBP3 as well as PBP1a, 1b and
3 (Table 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. Selection of βꢀlactam drugs with similar carboxamide moieties.
ACS Paragon Plus Environment

Page 11 of 36
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
Table 2. Oxime group modifications of 17.
P. aeruginosa
K. pneumoniae
A. baumannii
strains (ꢀg/mL)
strains (ꢀg/mL)
strain (ꢀg/mL)
Cmpd
R1
R2
1091-05 1042-06
1000-02
>64
1487-07
AB-3167
aztreonam
cefepime
4
2
64
32
64
64
>64
>64
64
4
NTa
2
meropenem
ciprofloxacin
2
0.25
NT
0.25
0.06
0.125
>64
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 36
1
2
3
4
5
6
7
8
amikacin
2
8
2
4
NT
0.25
1
64
2
1
Me
H
Me
H
0.25
0.125
0.25
0.25
17
23
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
H
H
1
16
0.5
1
1
24
Bn
1
0.125
0.25
1
8
1
1
8
1
1
1
25
26
27
28
ꢀ(CH2)3ꢀ
0.125
0.25
0.5
0.5
0.5
4
0.125
0.125
1
ꢀ(CH2)4ꢀ
ꢀ(CH2)5ꢀ
a NT = not tested
Attempts to improve potency by modifying the cis-oxime substituent of compound 17 were
generally unsuccessful (Table 3). For example, homologation (29) or isosteric replacement of
the carboxylate with a tetrazole (31) led to a moderate loss of potency relative to 17 against the
MDR P. aeruginosa strain 1042ꢀ06. Replacement of the carboxylate with a hydroxamic acid
(30) led to a more pronounced loss of activity vs. 1042ꢀ06, and this is also reflected in the
diminished PBP1a, 1b and 3 inhibitory activity of 30 relative to compound 17 (Table 8).
Analogs with lipophilic moieties lacking the carboxylic acid, such as the phenylꢀ (32),
cyclopentylꢀ (33) and thiazolylꢀcontaining (34) compounds, were also inferior vs. 1042ꢀ06 as
was replacement of the acid with either an alcohol (35) or amine (36). Comparison of
cyclopentyl analog 33 with the analogous carboxylateꢀcontaining compound 27 demonstrates the
contribution of the carboxylateꢀArg489 saltꢀbridge interaction to potency, and the reduced
activity of amine 36 is not surprising given the potential for charge repulsion with nearby
Arg489. All analogs in this set were significantly less potent than 17 vs. both K. pneumoniae
strains. Respectable activity is observed against the A. baumannii strain ABꢀ3167, with MICs
falling in the same range as that observed for the comparator drugs.
ACS Paragon Plus Environment

Page 13 of 36
Journal of Medicinal Chemistry
1
2
3
4
Table 3. Additional oxime modifications of 17.
5
6
7
8
9
10
11
12
13
14
15
16
17
P. aeruginosa
K. pneumoniae
A. baumannii
strains (ꢀg/mL)
strains (ꢀg/mL)
strain (ꢀg/mL)
Cmpd
R
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
1091-05 1042-06 1000-02 1487-07
AB-3167
aztreonam
4
64
32
64
64
8
>64
NTa
0.25
NT
>64
>64
64
4
2
cefepime
meropenem
ciprofloxacin
amikacin
2
2
0.25
0.06
1
0.125
2
>64
64
NT
17
29
0.25
0.5
2
8
0.25
32
2
0.25
0.5
>64
1
>64
16
8
8
16
32
1
30
31
0.5
0.5
Ph
0.5
1
16
16
>64
8
>64
32
0.5
1
32
33
cyclopentyl
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 36
1
2
3
4
5
6
7
8
9
1
>64
16
16
1
34
35
0.5
4
16
32
32
0.5
4
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
>64
>64
>64
36
a NT = not tested
Bacteria express a wide variety of siderophores and siderophore receptors, and as previously
reported, uptake of 1 and 2 in P. aeruginosa strain PAO1 likely involves the siderophore
receptors PiuA and PirA8. These receptors are known to be involved in recognition and uptake
of bacterial catecholꢀcontaining siderophores, such as enterobactin14. Enterobactin is a wellꢀ
characterized siderophore which is produced by E. coli and other Enterobacteriaceae and is
utilized by a number of bacteria (including P. aeruginosa) to sequester iron15. One plausible
theory regarding the cellular uptake of the monobactam analogs described here is that the
pyridone moiety mimics the catechol groups found in enterobactin and related siderophores
leading to receptor binding and delivery of drug to the bacterial periplasmic compartment. To
explore the pyridone SAR, a variety of structurally similar analogs were prepared (Table 4). The
Nꢀunsubstituted (37) and Nꢀhydroxy (38) pyridones displayed the best potency vs. both P.
aeruginosa strains. In general, our experience across a range of monobactam analogs suggests
that the Nꢀhydroxy pyridone moiety present in compound 38 typically provides the best overall
potency and spectrum. Moderate activity and spectrum was observed for a number of other
analogs, including the Nꢀmethoxy (39), Nꢀmethyl (40), NꢀCH2CO2H (41) and NꢀNHMe (43) and
significantly diminished potency and spectrum was observed for both the Nꢀphenyl pyridone 42
and pyranone 44, with 42 also demonstrating somewhat reduced PBP1b and 3 inhibitory activity
(Table 8). It is interesting to note that compounds 45 and 46 retain structural features most
ACS Paragon Plus Environment

Page 15 of 36
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
similar to the catechol moieties found in enterobactin, yet are the least active compounds in this
set. Unfortunately, PBP enzyme inhibition data is not available for either analog, so we are
unable to determine if the loss of cellular potency is due to reduced enzyme inhibition, reduced
siderophore receptorꢀmediated cellular uptake, or a combination of both factors.
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. Pyridone modification.
P. aeruginosa
K. pneumoniae
A. baumannii
strains (ꢀg/mL)
strains (ꢀg/mL)
strain (ꢀg/mL)
Cmpd
X
1091-05
1042-06 1000-02 1487-07
AB-3167
aztreonam
cefepime
4
2
64
32
64
64
8
>64
NTa
0.25
NT
>64
>64
64
4
2
meropenem
2
0.25
0.06
1
ciprofloxacin
amikacin
0.125
2
>64
64
NT
NH
0.125
0.5
0.5
NT
0.125
37
38b
39
NOH
NOMe
NMe
0.5
1
0.5
4
<0.06
1
0.25
8
0.5
1
4
4
0.25
4
4
40
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 36
1
2
3
4
5
6
7
8
9
NCH2CO2H
NPh
1
32
2
2
32
8
0.25
1
4
16
8
1
32
2
41
42
43
44
NNHMe
O
0.5
32
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
8
64
64
8
Siderophore
mimic
16
32
32
64
32
64
64
16
32
45
46
>64
a NT = not tested
b bisꢀNaꢀsalt
Since compound 17 displayed good overall potency against the initial Gramꢀnegative panel, it
was further evaluated against a broader set of clinically relevant pathogens known to express a
variety of resistance mechanisms (Table 5, SI1ꢀSI9). As in Table 1, representatives from several
drug classes are included for comparison. The majority of cystic fibrosis (CF) patients suffer
chronic P. aeruginosa infections, requiring nearly continual antibiotic therapy16. When
evaluated against a panel of 17 P. aeruginosa strains from CF patients, compound 17 provided
superior MIC50/90 values relative to the comparator drugs, and also performed well vs. 15 P.
aeruginosa strains harboring metalloꢀβꢀlactamases17 which showed high level resistance to all
comparator drugs, except aztreonam. Meropenem proved superior to 17 in K. pneumoniae
ACS Paragon Plus Environment

Page 17 of 36
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
panels expressing a variety of βꢀlactamases17, the exception being a KPCꢀpositive panel which
exhibited high level resistance to all comparator drugs, but was moderately susceptible to
compound 17. Against several E. coli panels composed of a variety of βꢀlactamase producers,
both 17 and meropenem consistently delivered low MIC50/90 values relative to the other
comparator drugs. While good potency was observed for both 17 and comparator drugs against
the susceptible A. baumannii strain ABꢀ3167 (Table 1), no compound provided a low MIC90
against a collection of 31 clinical strains of A. baumannii, with meropenem and ciprofloxacin
providing the best MIC50 values (1 and 0.5 ꢀg/mL). The relatively poor performance of
ciprofloxacin and amikacin against most panels described in Table 5 suggests the presence of
one or more additional resistance mechanisms beyond the expression of βꢀlactamases. Examples
could include upregulation of efflux pumps, downregulation of porin channels, drug target
binding site mutations and expression of aminoglycosideꢀmodifying enzymes.
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
Table 5. MIC50/90 data for compound 17 and comparator drugs.
MIC50/90 (ꢀg/mL)
Bacterial Strain # of strains
(phenotype)
tested
17
Aztreonam Cefepime Meropenem Ciprofloxacin
Amikacin
P. aeruginosa
(CF patients)
P. aeruginosa
17
0.25/4
8/64
8/>64
1/16
2/16
8/>64
15
17
0.25/1
8/16
>64/>64
8/>64
>64/>64
32/>64
>64/>64
4/64
(metalloꢀβꢀlactamase)
49
50
51
52
53
54
55
56
57
58
59
60
K. pneumoniae
0.125/1
32/>64
0.03/0.25
0.25/>64
(ESBL)
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 36
1
2
3
4
K. pneumoniae
30
22
0.5/16
4/16
64/>64
4/>64
0.03/0.25
32/>64
0.5/64
4/>64
5
6
7
8
(defined βꢀlactamase)
K. pneumoniae
(KPC)
9
>64/>64
>64/>64
64/>64
32/>64
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
E. coli
16
18
0.25/2
64/>64
4/>64
64/>64
0.25/4
0.03/0.06
0.03/0.06
>64/>64
8/64
1/16
(ESBL)
E. coli
0.125/1
0.03/0.06
(defined βꢀlactamase)
E. coli
(CTXꢀM)
11
31
0.5/1
64/>64
32/>64
64/>64
8/>64
0.03/0.06
1/32
64/>64
0.5/>64
8/32
Acinetobacter spp.
8/>64
4/>64
In order to determine which bacterial siderophore receptors may be involved in drug uptake,
monobactams 3, 17 and 38 were evaluated against a panel of siderophore receptorꢀdeficient P.
aeruginosa strains. This panel was reported previously for determining the specific siderophore
receptors utilized in the uptake of 1 and 28. No relevant MIC shift relative to the parent strain
PAO1 was observed with 3, 17 and 38 when evaluated against mutant strains lacking the
receptors involved in uptake of the major P. aeruginosa siderophores pyoverdin (FpvA) and
pyochelin (FptA) (Table 6). However, a significant (16ꢀ to 32ꢀfold) shift in MIC was observed
for the PiuA receptorꢀdeficient strain relative to the parent strain when the assay utilized standard
MuellerꢀHinton broth (MHB). The MIC shift was much less pronounced when conducted in
ACS Paragon Plus Environment

Page 19 of 36
Journal of Medicinal Chemistry
1
2
modified lowꢀiron MHB8, supporting the notion that siderophore receptor expression is regulated
3
4
5
6
7
by iron availability.18 Evaluation of 3, 17 and 38 vs. the ꢁpirA single mutant strain in either
8
9
normal or lowꢀiron media showed no MIC shift, however, evaluation vs. the ꢁpiuAꢁpirA double
mutant strain provided a >64ꢀfold MIC shift relative to the parent strain for all compounds in
both standard and lowꢀiron media. This was the only example of a second mutation on the
ꢁpiuA background which provided an additional MIC shift beyond that observed with the ꢁpiuA
single mutant. This may suggest that PiuA is the dominant uptake receptor for these compounds
in PAO1, while PirA functions as a secondary receptor whose expression is upregulated under
lowꢀiron conditions leading to enhanced drug uptake, thus explaining the reduced MIC shift for
the ꢁpiuA mutant in lowꢀiron vs. standard MHB media. These results are analogous to those
reported previously for 1 and 28 and may suggest a common uptake mechanism for pyridoneꢀ
conjugated βꢀlactams.
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 6. PAO1 isogenic siderophore receptor mutant panel.
MIC (ꢀg/mL)
3
17
Lowꢀiron MHB
38
Lowꢀiron
MHBa
0.25
8
Lowꢀironb
0.25
1
MHB
0.25
8
Strain
PAO1
piuA
0.125
0.5
0.5
8
0.5
0.5
0.5
0.5
0.5
1
pirA
0.25
0.125
0.25
8
0.25
0.125
0.25
0.5
0.25
0.25
0.25
8
0.25
0.06
0.25
0.25
0.5
0.5
1
fpvA
fptA
0.5
16
8
piuA fpvA
piuA fptA
8
1
8
0.5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 36
1
2
3
4
piuA pirA
32
16
32
32
64
64
5
6
a Mueller Hinton Broth.
7
8
9
b Ironꢀchelated version of MHB8.
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
Plasma protein binding and rat pharmacokinetic studies were conducted with a subset of analogs
to enable compound selection for advanced studies, and Table 7 provides representative data for
urea 17, carbamate 18 and triazole 21. All three compounds exhibited relatively low binding to
human, rat and mouse plasma proteins. Compounds 17 and 18 showed low plasma clearance in
rat following iv dosing, while triazole 21 was cleared more rapidly, and this was especially
apparent when comparing clearance values adjusted for protein binding and blood flow (Cl,int).
A previous report from our group demonstrated a positive correlation between Cl,int and
observed efficacy for related monobactams evaluated in a murine infection model,7 supporting
the selection of 17 for in vivo efficacy studies which will be reported in a future manuscript.
Table 7. Pharmacokinetic and protein binding data.
Plasma Protein Binding fu
Rat Plasma
Clearance
(mL/min/kg)
8.4 ± 0.4
Rat Cl,int
Cmpd
(ml/min/kg)a
Human
Rat
Mouse
1.0 ± 0.29
0.51 ±0.06
0.50 ± 0.06
0.91 ± 0.07
0.61 ± 0.04
0.60 ± 0.07
0.91 ± 0.07
0.72 ± 0.03
0.80 ± 0.03
10 ± 0.5
29 ± 15
17
18
21
14.2 ± 5.6
31.3 ± 0.4
94 ± 2.8
a Total plasma clearance corrected for protein binding and rat blood flow.
Table 8. P. aeruginosa Penicillin binding protein IC50s.
ACS Paragon Plus Environment

Page 21 of 36
Journal of Medicinal Chemistry
1
2
3
4
5
IC50 (ꢀM)
Cmpd
6
7
8
9
PBP1a
3.34 ± 0.57
0.11 ± 0.14
0.62 ± 1.0
0.34 ± 0.096
3.02
PBP1b
2.87 ± 0.64
0.15 ± 0.13
1.67 ± 3.11
2.46 ± 0
PBP2
> 300
> 300
> 300
> 300
> 300
> 300
PBP3
0.008 ± 0.004
0.006 ± 0.004
0.054 ± 0.044
0.066 ± 0.034
0.27
aztreonam
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
3
17
26
30a
42
11.1
0.58 ± 0.29
7.40 ± 0
0.14 ± 0
an = 1 data.
In summary, we have described a structurally diverse collection of pyridoneꢀconjugated
monobactams with in vitro activity against a variety of clinically relevant Gramꢀnegative
pathogens. MIC50/90 profiling of compound 17 demonstrated balanced antibacterial activity
against a panel of P. aeruginosa, K. pneumoniae and E. coli strains harboring resistance
mechanisms for a number of drug classes, including βꢀlactams, fluoroquinolones and
aminoglycosides. Cocrystal structures of βꢀlactam ringꢀopened compounds, such as 17, bound to
P. aeruginosa PBP3 enzyme provided insight into drugꢀprotein interactions and aided in the
design of analogs described here. In addition, utilization of siderophore receptor knockout
strains provided evidence suggesting involvement of the siderophore receptors PiuA and PirA in
drug uptake in the P. aeruginosa strain, PAO1. Finally, rat PK studies conducted with urea 17
demonstrated low unbound clearance, supporting advancement of this analog to in vivo efficacy
studies, the results of which will appear in due course.
EXPERIMENTAL SECTION
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 36
1
2
3
4
5
6
7
8
9
The experimental methods utilized to generate the P. aeruginosa siderophore receptor KO data
have been described previously8. The minimum inhibitory concentration (MIC) values were
determined using the broth microdilution protocol according to the methods of the Clinical and
Laboratory Standards Institute (CLSI)19. Plasma protein binding values were determined using a
modification of a reported method20. The assay was conducted at ambient temperature rather
than 37 °C. Positive controls were utilized for the protein binding studies; phenytoin was used
for compound 17 and sertraline was utilized for compounds 18 and 21 and in all cases,
acceptable positive control results were obtained relative to historical data.
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
P. aeruginosa PBP IC50 assay. Membrane preparation (ATCC27853) was conducted according
to a published method21. Membrane samples (75 ꢀg) were mixed with 1.5 ꢀL 100 ꢀM EDTA,
and the total volume of the resulting mixture was brought to 15 ꢀL with PBS. The βꢀlactams
tested were diluted in a 1 to 3 scheme in 1.5% DMSO so that 5 ꢁL would give a final
concentration of 300 ꢁM to 0.005 ꢁM (11 wells serially diluted 1 to 3 including a no compound
control) in a 25 ꢁL final assay volume. Samples were incubated at 35 °C for 20 minutes,
followed by the addition of 5 ꢁL of the fluorescent penicillin Bocillin FL (Molecular Probes,
Inc.) suspended in PBS to yield a final assay concentration of 0.65 ꢁM. Reaction mixtures were
incubated at 35 °C for 20 min, then terminated by the addition of 25 ꢁL 2× Laemmli buffer.
After boiling in a water bath for 4 minutes, samples were centrifuged at 14,000 rpm for 2
minutes, and 10 ꢁL was loaded onto a 10% Tris/Bis SDS gel (NuPage) run with the MES buffer
system (Invitrogen) at 150 V for 75 min. The gel was washed briefly in deionized H2O and
scanned by a Storm 860 (Molecular Dynamics) at an excitation wavelength of 450 nm and an
emission wavelength of 520 nm. An IC50 was determined for each βꢀlactam assayed
ACS Paragon Plus Environment

Page 23 of 36
Journal of Medicinal Chemistry
1
2
3
4
(concentration where the βꢀlactam causes a 50% reduction of the Bocillin FL binding to each
5
6
PBP in the P. aeruginosa total membrane preparation).
7
8
9
Rat pharmacokinetic studies. All animal care and in vivo procedures conducted were in
accordance with guidelines of the Pfizer Animal Care and Use Committee. Male Wistar Han rats
(~250g) were housed one per cage in an American Animal Association Laboratory Animal Care
accredited facility. This investigation conformed to the Guide for the Care and Use of
Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85ꢀ
23, revised 1996). Animals were allowed ad libitum access to water and food. Compounds were
administered intravenously via the jugular vein cannula (n = 2), and were administered at 1.0
mg/kg i.v.. Compounds were formulated as a solution in 20 mM phosphate buffer (pH = 6).
After dosing, serial plasma samples were collected at appropriate times and kept frozen at −20
°C until LCꢀMS/MS analysis.
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
LC-MS/MS method. A nonꢀvalidated liquid chromatographyꢀtandem mass spectrometry (LCꢀ
MS/MS) was used to determine compound concentrations in the plasma protein binding and rat
pharmacokinetic assays. Samples were injected (10 mL) onto a Phenomenex Monolithic C18
column (5 mm, 50 x 2.0 mm). The column was equilibrated with mobile phase (A: 0.1% formic
acid in water / B: 0.1% formic acid in acetonitrile) at a flow rate of 0.8 mL/min. The gradient
started at 0% B and was ramped to 95% B from 0.0 to 0.6 min, then returned to starting
conditions by 0.9 min and held for an additional 0.6 min for a total run time of 1.5 min. The
effluent was analyzed by a mass spectrometer detector (AB Sciex APIꢀ4000), fitted with a turbo
ion spray interface and operated in positive ion mode. Aztreonam was used as the internal
standard for both protein binding and rat PK studies.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 36
1
2
3
4
5
6
7
8
9
Molecular Modeling. βꢀlactam ringꢀopened models of pyridoneꢀconjugated monobactam
analogs bound to PaePBP3 were generated using the following procedure. The 2D structures of
the closed form of the compounds were converted into single low energy 3D conformations
using Corina22. Protons were titrated appropriately at pH = 7.4 using an internal software CVT23.
Multiple low energy conformations (maximum number of conformations = 100) of the
compounds were generated using Omega24. A shapeꢀbased matching protocol to select the
conformation most similar to the closest crystal structure was then performed using ROCS25.
This was taken as the initial pose in the binding site, which was then refined by an “inꢀsitu” bond
clipping of the βꢀlactam in the enzyme active site followed by minimizing the compound in the
context of the protein using Macromodel26 and the OPLS2005 force field27. Exhaustive
conformational searching using the MonteꢀCarlo Multiple Minimization (MCMM) algorithm
was implemented in MacroModel and the 10 lowest minimum energy conformations were
visually inspected to select a reasonable, representative binding conformation.
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
Chemistry. A purity of > 95% as determined by either HPLC or LCMS was achieved for all
final analogs, with the exception of compounds 19 (88%, HPLC), 22 (88%, HPLC), 23 (92%,
HPLC), 32 (88%, HPLC), 35 (92%, HPLC), 36 (90%, HPLC), 37 (>90%, HPLC) and 34 (94%,
HPLC). The synthesis of compound 18 is provided below. Detailed synthetic procedures and
spectral characterization for all compounds is provided in the Supporting Information.
Sodium
2-((Z)-1-(2-aminothiazol-4-yl)-2-((2R,3S)-2-((((1,5-dihydroxy-4-oxo-1,4-
dihydropyridin-2-yl)methoxy)carbonylamino)methyl)-4-oxo-1-sulfonatoazetidin-3-
ylamino)-2-oxoethylideneaminooxy)-2-methylpropanoate (18).
ACS Paragon Plus Environment

Page 25 of 36
Journal of Medicinal Chemistry
1
2
3
4
(2S,3S)-Methyl
3-(benzyloxycarbonylamino)-4-oxoazetidine-2-carboxylate
(4).
5
6
7
8
Chiral resolution of racemicꢀcisꢀ49 was achieved by supercritical fluid chromatography
(Chiralcel OJꢀH: CO2 / propanol) to afford 4 as a white solid. For literature characterization of 4,
see Y. Takahashi, et al.28 99.7% ee. LCMS m/z 279.2 (M+1). [α]20 +81.93o (c 0.035, CHCl3).
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
1H NMR (400 MHz, DMSOꢀd6) δ 8.60 (br s, 1H), 8.14 (d, J=8.8 Hz, 1H), 7.34ꢀ7.29 (m, 5H),
5.05ꢀ5.01 (m, 3H), 4.35 (d, J=5.5 Hz, 1H), 3.53 (s, 3H).
(3S,4S)-3-Amino-4-(hydroxymethyl)azetidin-2-one (HOAc salt) (5). A solution of 4
o
(72.5 g, 260 mmol) in methanol (725 mL) at 20 C was treated with a suspension of sodium
borohydride (18.7 g, 495 mmol) in isopropyl alcohol (145 mL) added portionwise over 30
minutes, the temperature was maintained between 26ꢀ33 oC. The reaction mixture was stirred for
20 minutes. The methanol was removed in vacuo and the mixture was treated with brine
solution (200 mL) and water (200 mL). The white slurry was extracted with ethyl acetate (700
mL) and washed with brine solution (3 x 200 mL). The aqueous layer was back extracted with
ethyl acetate / isopropyl alcohol (10:1, 2 x 220 mL) and the combined organic layers were dried
over magnesium sulfate. The suspension was filtered under vacuum and the filtrate concentrated
in vacuo to give crude material (81.2 g) as a solid. The crude material was treated with ethyl
acetate (400 mL) followed by Darco KB (2 g) and Celite (5 g) and the mixture was stirred at
room temperature for 30 minutes. The mixture was filtered and the solids washed with ethyl
acetate (100 mL). The filtrate was treated with heptane (750 mL) over 30 minutes. The white
slurry was filtered and the solid washed with ethyl acetate / heptane (2 : 3, 150 mL) to afford
benzyl (2S,3S)ꢀ2ꢀ(hydroxymethyl)ꢀ4ꢀoxoazetidinꢀ3ꢀylcarbamate as a white solid. Yield: 59.3 g,
1
237 mmol, 91%. LCMS m/z 251.6 (M+1). [α]20 +9.03o (c 0.064, CHCl3). mp 125ꢀ127 oC. H
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 36
1
2
3
4
5
6
NMR (400 MHz, CDCl3) δ 7.35ꢀ7.28 (m, 5H), 5.67 (br s, 1H), 6.18 (d, J=9.9 Hz, 1H), 5.15 (dd,
J=9.8, 4.8 Hz, 1H), 5.08 (s, 2H), 3.85ꢀ3.79 (m, 2H), 3.65 (m, 1H), 3.36 (br s, 1H).
7
8
9
To a solution of benzyl (2S,3S)ꢀ2ꢀ(hydroxymethyl)ꢀ4ꢀoxoazetidinꢀ3ꢀylcarbamate (29.4 g,
117.6 mol) in ethanol (589 mL) was added 9.0 g of Darco (Norit KB). The resulting slurry was
stirred for 1 hour and the Darco was removed by vacuum filtration, then the Darco cake was
rinsed with ethanol (294 mL). The ethanolic filtrate was treated with glacial acetic acid (13.5
mL, 236 mmol) and the resulting mixture was charged with 5.9 g of 20% palladium hydroxide.
The mixture was purged with nitrogen followed by hydrogen and pressurized to 50 psi hydrogen.
The reaction mixture was agitated at 25 °C for 16 hours. The mixture was filtered and the filter
cake was rinsed with ethanol (150 mL). The ethanolic filtrate was concentrated to afford 5 as an
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
1
oil. Yield: 26.82 g, 152.2 mmol, >100%. H NMR (400 MHz, CD3OD) δ 4.29 (d, J=4.8 Hz,
1H), 3.89 (dd, J=13.0, 4.8 Hz, 1H), 3.76ꢀ3.81 (m, 2H), 1.98 (s, 3H).
tert-Butyl
2-((Z)-1-(2-(tert-butoxycarbonylamino)thiazol-4-yl)-2-((2S,3S)-2-
(hydroxymethyl)-4-oxoazetidin-3-ylamino)-2-oxoethylideneaminooxy)-2-methylpropanoate
(6). A flask charged with 5 (26.82 g) was treated with a solution of (Z)ꢀtertꢀbutyl 2ꢀ(1ꢀ(2ꢀ(tertꢀ
butoxycarbonylamino)thiazolꢀ4ꢀyl)ꢀ2ꢀ(2,5ꢀdioxopyrrolidinꢀ1ꢀyloxy)ꢀ2ꢀoxoethylideneaminooxy)ꢀ
2ꢀmethylpropanoate (9) (41.5 g, 78.81 mmol) in acetonitrile (200 mL), followed by triethylamine
(55.0 mL, 394.6 mmol) added over 5 minutes. The reaction mixture was stirred for 16 hours.
The solution was concentrated in vacuo to yield a yellow glass. The glass was dissolved in
methyl tertꢀbutyl ether (500 mL) and washed with water (1 x 250 mL), saturated aqueous sodium
bicarbonate solution (1 x 250 mL), saturated brine solution (1 x 250 mL), 1% aqueous potassium
carbonate solution (1 x 500 mL) and saturated brine solution (1 x 500 mL). The methyl tertꢀ
butyl ether organic layer was concentrated in vacuo to give crude material (38.0 g) as an offꢀ
ACS Paragon Plus Environment

Page 27 of 36
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
white solid. The crude material (38.0 g) was treated with acetone (95 mL) and heptane (285
mL). The mixture was heated to 45 °C and was held at this temperature for 30 minutes. The
thin slurry was cooled to 20 °C and stirred for 16 hours. The offꢀwhite solids were collected by
vacuum filtration and the isolated filter cake was washed with 25% acetone ꢀ heptane (100 mL)
to afford a white solid. The solid was dried in a vacuum oven at 30 °C to afford 6. Yield: 23.13
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
1
g, 43.8 mmol, 56%. LCMS m/z 528.1 (M+1). H NMR (400 MHz, CDCl3) δ 8.80 (br s, 1H),
8.02 (d, J=7.7 Hz, 1H), 7.31 (s, 1H), 6.48 (br s, 1H), 5.44 (br dd, J=7.7, 4.8 Hz, 1H), 4.30ꢀ4.36
(m, 1H), 4.02ꢀ4.06 (m, 1H), 3.84ꢀ3.89 (m, 2H), 1.57 (s, 3H), 1.55 (s, 3H), 1.53 (s, 9H), 1.45 (s,
9H).
tert-Butyl
2-((Z)-2-((2R,3S)-2-(aminomethyl)-4-oxoazetidin-3-ylamino)-1-(2-(tert-
butoxycarbonylamino)thiazol-4-yl)-2-oxoethylideneaminooxy)-2-methylpropanoate (7). To
o
a solution of 6 (10.0 g, 18.9 mmol) in anhydrous dichloromethane (100 mL) at 20 C under
nitrogen was added triphenylphosphine (10.0 g, 38 mmol), followed by imidazole (2.58 g, 38
mmol). The reaction mixture was treated with iodine (9.62 g, 38 mmol) in portions over 10
o
minutes. The solution was allowed to warm to 20 C and stirring was continued for 15 hours.
The reaction mixture was washed with saturated aqueous sodium thiosulfate (100 mL) and the
aqueous layer was extracted with dichloromethane (2 x 100 mL). The combined organic layers
were washed with brine solution (100 mL), dried over sodium sulfate, filtered under vacuum and
the filtrate was concentrated in vacuo to give crude material (24 g) as an orange foam. The
orange foam was purified by chromatography on silica gel (heptane /ethyl acetate 30 to 80%) to
afford tertꢀbutyl 2ꢀ((Z)ꢀ1ꢀ(2ꢀ(tertꢀbutoxycarbonylamino)thiazolꢀ4ꢀyl)ꢀ2ꢀ((2S,3S)ꢀ2ꢀ(iodomethyl)ꢀ
4ꢀoxoazetidinꢀ3ꢀylamino)ꢀ2ꢀoxoethylideneaminooxy)ꢀ2ꢀmethylpropanoate as a white foam.
1
Yield: 9.41 g, 14.8 mmol, 78%. LCMS m/z 638.0 (M+1). H NMR (400 MHz, DMSOꢀd6) δ
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 36
1
2
3
4
5
6
7
8
9
11.83 (br s, 1H), 9.24 (d, J=9.0 Hz, 1H), 8.71 (br s, 1H), 7.29 (d, J=0.8 Hz, 1H), 5.17 (ddd,
J=9.0, 4.9, 1.7 Hz, 1H), 4.11 (ddd, J=10.6, 4.8, 3.7 Hz, 1H), 3.28 (dd, J=10.4, 3.6 Hz, 1H), 3.17
(dd, J=10.5, 10.4 Hz, 1H), 1.46 (s, 9H), 1.42 (s, 3H), 1.40 (s, 12H).
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
A solution of tertꢀbutyl 2ꢀ((Z)ꢀ1ꢀ(2ꢀ(tertꢀbutoxycarbonylamino)thiazolꢀ4ꢀyl)ꢀ2ꢀ((2S,3S)ꢀ
2ꢀ(iodomethyl)ꢀ4ꢀoxoazetidinꢀ3ꢀylamino)ꢀ2ꢀoxoethylideneaminooxy)ꢀ2ꢀmethylpropanoate (6.3
o
g, 9.88 mmol) in 2ꢀmethyltetrahydrofuran (60 mL) under nitrogen at 20 C was treated with
triethylamine (2.75 mL, 19.76 mmol), followed by dropwise addition of a 15% solution
of tetrabutylammonium azide in tetrahydrofuran (22.49 g, 11.86 mmol). The reaction mixture
o
was stirred at 20 oC for 2 hours, and then heated at 35 C and stirred for 15 hours. The solution
was cooled to room temperature, filtered under vacuum, the white solid was washed with methyl
tertꢀbutyl ether (2 x 100 mL), the filtrate was collected and the solvent was removed in vacuo to
give a foam. The foam was dissolved in methyl tertꢀbutyl ether (200 mL), washed with water (2
x 100 mL), brine solution (100 mL), dried over sodium sulfate, and filtered under vacuum. The
filtrate was collected and concentrated in vacuo to afford a foam. The foam was dissolved in
acetonitrile and concentrated (3 x 15 mL) and the residue was held under high vacuum to give
tertꢀbutyl
2ꢀ((Z)ꢀ2ꢀ((2R,3S)ꢀ2ꢀ(azidomethyl)ꢀ4ꢀoxoazetidinꢀ3ꢀylamino)ꢀ1ꢀ(2ꢀ(tertꢀ
butoxycarbonylamino)thiazolꢀ4ꢀyl)ꢀ2ꢀoxoethylideneaminooxy)ꢀ2ꢀmethylpropanoate as a yellow
1
foam. Yield: 5.24 g, 9.48 mmol, 96%. LCMS m/z 553.1 (M+1). H NMR (400 MHz, DMSOꢀ
d6) δ 11.82 (br s, 1H), 9.19 (d, J=8.9 Hz, 1H), 8.67 (br s, 1H), 7.28 (s, 1H), 5.24 (ddd, J=8.7, 5.1,
1.4 Hz, 1H), 3.89ꢀ3.95 (m, 1H), 3.64 (dd, J=12.9, 3.9 Hz, 1H), 3.39 (dd, J=12.9, 8.9 Hz, 1H),
1.46 (s, 9H), 1.44 (s, 3H), 1.42 (s, 3H), 1.40 (s, 9H).
A Parr shaker vessel was charged with tertꢀbutyl 2ꢀ((Z)ꢀ2ꢀ((2R,3S)ꢀ2ꢀ(azidomethyl)ꢀ4ꢀ
oxoazetidinꢀ3ꢀylamino)ꢀ1ꢀ(2ꢀ(tertꢀbutoxycarbonylamino)thiazolꢀ4ꢀyl)ꢀ2ꢀ
ACS Paragon Plus Environment

Page 29 of 36
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
oxoethylideneaminooxy)ꢀ2ꢀmethylpropanoate (14.37 g, 26.0 mmol) and ethanol (140 mL). The
mixture was purged with nitrogen and then treated with 10% palladium on carbon (5.7 g) and
pressurized to 30 psi hydrogen. The reaction mixture was agitated at room temperature for 4
hours. The solution was filtered through a micro filter and the solvent was removed in vacuo to
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
1
afford 7 as a brown solid. Yield: 13.22 g, 25.1 mmol, 97%. LCMS m/z 527.1 (M+1). H NMR
(400 MHz, DMSOꢀd6) δ 9.1ꢀ9.3 (v br s, 1H), 8.25 (br s, 1H), 7.26 (s, 1H), 5.17 (d, J=4.9 Hz,
1H), 3.65 (ddd, J=6, 6, 5 Hz, 1H), 2.78 (dd, J=13.4, 5.8 Hz, 1H), 2.62 (dd, J=13.4, 6.3 Hz, 1H),
1.46 (s, 9H), 1.43 (s, 3H), 1.41 (s, 3H), 1.39 (s, 9H).
tert-Butyl
2-((Z)-2-((2R,3S)-2-((((1,5-bis(4-methoxybenzyloxy)-4-oxo-1,4-
dihydropyridin-2-yl)methoxy)carbonylamino)methyl)-4-oxoazetidin-3-ylamino)-1-(2-(tert-
butoxycarbonylamino)thiazol-4-yl)-2-oxoethylideneaminooxy)-2-methylpropanoate (8). A
solution of 2ꢀ(hydroxymethyl)ꢀ1,5ꢀbis(4ꢀmethoxybenzyloxy)pyridinꢀ4(1H)ꢀone (11) (see
Supplementary Information for preparation) (300 mg, 0.755 mmol) in tetrahydrofuran (10 mL)
was treated with 1,1’ꢀcarbonyldiimidazole (379 mg, 2.26 mmol) at room temperature and stirred
for 20 hours. The yellow reaction mixture was treated with a solution of 7 (286 mg, 0.543
mmol) in tetrahydrofuran (25 mL). The mixture was stirred for 6 hours at room temperature,
then treated with water (20 mL) and extracted with ethyl acetate (3 x 25 mL). The combined
organic layers were dried over sodium sulfate, filtered and concentrated in vacuo. The crude
material was purified via chromatography on silica gel (heptane / ethyl acetate / 2ꢀpropanol) to
1
afford 8 as a light yellow solid. Yield: 362 mg, 0.381 mmol, 62%. LCMS m/z 950.4 (M+1). H
NMR (400 MHz, DMSOꢀd6), characteristic peaks: δ 9.31 (d, J=8.4 Hz, 1H), 8.38 (s, 1H), 8.00
(s, 1H), 7.41 (br d, J=8.2 Hz, 2H), 7.36 (br d, J=8.8 Hz, 2H), 7.26 (s, 1H), 6.10 (s, 1H), 5.20 (s,
2H), 4.92 (br s, 4H), 3.77 (s, 3H), 3.76 (s, 3H), 1.45 (s, 9H), 1.38 (s, 9H).
ACS Paragon Plus Environment