A γ-Lactam Siderophore Antibiotic Effective against Multidrug-Resistant Gram-Negative Bacilli

Article abstract of DOI:10.1021/acs.jmedchem.0c00255

Treatment of multidrug-resistant Gram-negative bacterial pathogens represents a critical clinical need. Here, we report a novel γ-lactam pyrazolidinone that targets penicillin-binding proteins (PBPs) and incorporates a siderophore moiety to facilitate uptake into the periplasm. The MIC values of γ-lactam YU253434, 1, are reported along with the finding that 1 is resistant to hydrolysis by all four classes of β-lactamases. The druglike characteristics and mouse PK data are described along with the X-ray crystal structure of 1 binding to its target PBP3.

Full text of DOI:10.1021/acs.jmedchem.0c00255

Subscriber access provided by Uppsala universitetsbibliotek
Article
A #-Lactam Siderophore Antibiotic Effective
Against Multidrug-Resistant Gram-Negative Bacilli
Joel Goldberg, Ha Nguyen, Vijay Kumar, Elizabeth Spencer, Denton Hoyer, Emma Marshall, Anna
Cmolik, Margaret O'Shea, Steven Marshall, Andrea Hujer, Kristine Hujer, Susan D. Rudin, T. Nicholas
Domitrovic, Christopher Bethel, Krisztina M. Papp-Wallace, Latania Logan, Federico Perez, Michael
Jacobs, David van Duin, Barry Kreiswirth, Robert A. Bonomo, Mark Plummer, and Focco van den Akker
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.0c00255 • Publication Date (Web): 18 May 2020
Downloaded from pubs.acs.org on May 18, 2020
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 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 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.
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 44
Journal of Medicinal Chemistry
1
2
3
4
A -Lactam Siderophore Antibiotic Effective Against
5
6
Multidrug-Resistant Gram-Negative Bacilli
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
Joel A. Goldberg,1 Ha Nguyen,2 Vijay Kumar,2 Elizabeth J. Spencer,3 Denton Hoyer,3 Emma K.
Marshall,1 Anna Cmolik,1 Margaret O’Shea,1 Steven H. Marshall,1 Andrea M. Hujer,1,4 Kristine
M. Hujer,1,4 Susan D. Rudin,1,4 T. Nicholas Domitrovic,1,4 Christopher R. Bethel,1 Krisztina M.
Papp-Wallace,1,2,4 Latania K. Logan,5,6, Federico Perez,1,4,7 Michael R. Jacobs,4,8 David van
Duin,9 Barry M. Kreiswirth,10 Robert A. Bonomo,*1,2,4,7,11,12 Mark S. Plummer,*3 and Focco van
den Akker*2
1Research Service, Louis Stokes Cleveland Department of Veterans Affairs Medical Center,
Cleveland, OH 44106, USA
2Department of Biochemistry, Case Western Reserve University, Cleveland, OH 44106, USA
3Yale Center for Molecular Discovery, West Haven, CT 06516, USA
4Department of Medicine, Case Western Reserve University, Cleveland, OH 44106, USA
5 Department of Pediatrics, Rush University Medical Center, Rush Medical College, Chicago, IL
60612, USA
6Cook County Health and Hospital Systems, Chicago, IL 60612, USA
7Geriatric Research, Education and Clinical Center, Louis Stokes Cleveland Department of
Veterans Affairs Medical Center, Cleveland, OH 44106, USA
8Department of Pathology, University Hospitals Cleveland Medical Center, Division of Clinical
Microbiology, Cleveland, OH 44106, USA
9University of North Carolina School of Medicine, Chapel Hill, NC 27514, USA
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 44
1
2
3
4
5
6
7
8
9
10Center for Discovery and Innovation, Hackensack Meridian Health, Nutley, NJ 07601, USA
11Departments of Pharmacology, Molecular Biology & Microbiology, and Proteomics &
Bioinformatics, Case Western Reserve University, Cleveland, OH 44106, USA
12CWRU-Cleveland VAMC Center for Antimicrobial Resistance and Epidemiology (Case VA
CARES), Cleveland, OH 44106, USA
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
2
ACS Paragon Plus Environment

Page 3 of 44
Journal of Medicinal Chemistry
1
2
3
4
5
ABSTRACT
6
7
8
9
Treatment of multidrug resistant Gram-negative bacterial pathogens represent a critical clinical
need. Here we report on a novel -lactam pyrazolidinone that targets penicillin binding protein
(PBPs) and incorporates a siderophore moiety to facilitate uptake into the periplasm. The MIC
values of -lactam YU253434, 1, are reported along with the finding that 1 is resistant to hydrolysis
by all four classes of -lactamases. The drug-like characteristics and mouse PK data are described
along with the X-ray crystal structure of 1 binding to its target PBP3.
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
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 44
1
2
3
4
INTRODUCTION
5
6
7
8
9
At the present time, antimicrobial resistance (AMR) is creating significant economic hardships,
impacting public health, and is leading to untreatable infections that are associated with significant
morbidity and mortality. In the United States alone, more than two million infections each year
are caused by antibiotic resistant bacteria resulting in approximately 35,000 deaths and more than
$20 billion in healthcare costs.1 The global impact is staggering with an estimated cumulative
economic cost of $100 trillion by 2050 and 10 million deaths resulting from multidrug resistant
(MDR) infections.2 In particular, MDR Gram-negative bacterial pathogens are associated with
mortality rates upwards of 30–70% and are classified by the WHO as Priority 1 pathogens.3–7
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
The “hard-to-treat” nature of these infections is often due to extensive antibiotic-resistant
phenotypes that allow pathogens to overcome many standard-of-care antibiotic therapies.8 Despite
the increasing public health threat, few truly novel agents are in development to treat such
infections, as large pharmaceutical companies have withdrawn from antibacterial research.3,4,9,10
Government agencies are aware of the ever-growing issue of AMR and are responding with several
initiatives in an effort to incentivize renewed efforts in antibiotic development.11 These initiatives
include public-private partnerships highlighting the urgency and need for novel treatment
solutions.
Resistance in MDR Gram-negative organisms is multi-causal, due not only to reduced
permeability observed with all Gram-negative pathogens, but also to efficient and diverse efflux
pumps.12,13 Of particular concern are WHO Priority 1 pathogens, Acinetobacter baumannii,
Pseudomonas aeruginosa, and Enterobacterales5 that possess plasmid-mediated carbapenemases
(β-lactamases (blas): NDM-1, KPC and OXA-type class D carbapenemases), making them nearly
untreatable with β-lactam antibiotics.1 Nevertheless, -lactam antibiotics remain a mainstay
4
ACS Paragon Plus Environment

Page 5 of 44
Journal of Medicinal Chemistry
1
2
3
4
treatment for infections and are often used in combination with -lactamase inhibitors. Clearly,
5
6
new antibacterial agents are needed to effectively treat infections caused by MDR pathogens.
7
8
9
The search for new antibiotic targets is crucial, but not a prerequisite for success. In this
report, we exploit a well-known target, the penicillin binding proteins (PBPs), using a new non--
lactam pyrazolidinone chemical scaffold. These pyrazolidinones, coupled with a siderophore for
penetration, irreversibly inhibit PBPs by acylating the catalytic serine. Importantly
pyrazolidinones are poorly hydrolyzed by all four classes of -lactamases, thus avoiding the
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
Achilles’ heel of -lactam antibiotics, including those used as last resort treatments. Herein, we
report on the attributes of one such pyrazolidinone, YU253434, 1.
More than thirty years ago, pyrazolidinones were discovered that demonstrated potent
Gram-positive bacterial activity, low oral bioavailability, and suboptimal Gram-negative
activity.14-17 These agents did not progress to the clinic as the focus of antibacterial research at that
time was on developing broad spectrum oral agents17 with potent Gram-positive activity. Early
compounds displayed some Gram-negative activity when aminothiazoleoxime-type side chains
were employed especially when combined with very strongly electron withdrawing C(3)
substituents like a cyano group (Figure 1).16,17 The effect of C(3) substituents on activity and
stability has been established.16,17 Broadly, MIC correlates with the electron withdrawing nature
of the σp value of the C(3) substituent and is tunable for reactivity and stability.14,18 Here, we use
only a mildly electron withdrawing amide group at C(3) which imparts low reactivity and
remarkable stability, while simultaneously using the amide to append a siderophore moiety (Figure
1).
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 44
1
2
3
4
5
6
7
8
9
To improve cell uptake, we strategically exploit the notion that bacteria have highly
developed methods for acquiring iron which is crucial to their survival in their native environment
as well as in hosts.19 Generally, the siderophore is synthesized and released by the bacterium to
acquire iron and effectively compete with other siderophores, from other bacterial species or with
endogenous iron binding proteins of the host. Once the siderophore has complexed iron, it is
actively imported through a variety of specialized receptors.20,21 Often the small molecule natural
siderophores are structurally complex and species selective. To utilize the import receptors, the
siderophores fortunately can be much simpler siderophore mimics. Previously, such mimics were
used to import antibacterial agents into the Gram-negative periplasmic space where PBP targets
reside, colloquially known as a Trojan Horse strategy. Only one such agent, cefiderocol (Figure
1), has received recent approval.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
The design, synthesis, and in vitro characterization 1, a novel pyrazolidinone antibiotic
containing a dihydroxyphthalimide siderophore mimetic is reported herein.23 Antimicrobial
activity is described in comparison to current -lactam antibiotics against select clinical isolates
of P. aeruginosa, K. pneumoniae and Escherichia coli, as well as, the compound’s PBP binding
characteristics and 2.4 Å resolution crystal structure in complex with P. aeruginosa PBP3. The
chemical structures of 1 and comparator agents are shown in Figure 1.
6
ACS Paragon Plus Environment

Page 7 of 44
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
S
H2N
OH
O
3
N
N
N
O
HN
O
OH
OH
OH
HN
CO2H
O
N
N
O
O
O
N
O
O
YU253434, 1
O
N
HO
H2N
H
N
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
S
O
N
HN
O
O
3
H
O-
N
HN
O
O-
R
N
N
O
HN
N
O
O
CO2H
Substructure of the
product resulting from
hydrolysis of YU253434
O
2
HO
S
H2N
O
N
N
HN
CO2H
H2N
O
O
O N
O
O
O
S
N
N
HO
O
O
3
OH
OH
H
S
O
O
N
N+
HO
N
N
a siderophore conjugated lactivicin analog
O
4
O
OH
H2N
S
O
Ceftazidime
N
H
N
O
H
N
OH
O
N
O
O
S
N
O
N
HO
OH
O
O
S
N
5
O
6
Aztreonam
OH
O
OH
Meropenem
HO
Cl
O
H2N
S
O
HN
H2N
N
N
O
O
O
H
S
OH
O
N
N
N+
N
S
HO
O
O
N
O
7
8
O
O-
O
Avibactam
a beta lactamase inhibitor
Cefiderocol
Appproved 11/19
Figure 1. Comparator agents and select siderophore containing PBP inhibitors, as well as, the -
lactamase inhibitor avibactam. Highlighted in red is the site that covalently modifies the PBP or
-lactamase catalytic serine. When a siderophore mimic is present it is highlighted in blue.
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 44
1
2
3
4
5
6
RESULTS AND DISCUSSION
Syntheses
7
8
9
The synthesis of 1 (Scheme 1) follows, in part, the previous synthesis of pyrazolidinones
and begins with conversion of Boc-L-serine to the known cyclic hydrazide 9.14,17 Reaction of this
hydrazide with the known acryl phosphonate results in addition of the nucleophilic amine to
cleanly give the substituted phosphonohydrazide 10. The phosphonohydrazide, 10, could be used
in crude form, but was often purified by silica gel chromatography. Reaction of 10 with tert-butyl
2-chloro-2-oxoacetate in dichloromethane with excess Hunig’s base resulted in acylation followed
by slow ring closure to the pyrazolidinone nucleus 11. Selective deprotection of the allyl ester
provides the intermediate acid, 12, in six steps with one chromatographic purification. Intermediate
12 is suitably differentiated to allow incorporation of a diverse set of amide-linked siderophores
and PBP recognition elements (see below). To this end, the acid 12 was converted to its acid
chloride with oxalyl chloride and reacted in situ with the known siderophore23 13 which was best
solubilized and transiently protected using N-methyl-N-(trimethylsilyl) trifluoroacetamide
(MSTFA) to give 14. Deprotection of 14 with trifluoroacetic acid in dichloromethane using
triethylsilane as a scavenger gave the amino acid intermediate 15. The acid chloride of the suitably
protected aminothiazole 16 was formed using oxalyl chloride and reacted with the amine 15 that
had been protected and solubilized using MSTFA to give the acylated product 17. Crude 17 was
deprotected with trifluoroacetic acid in dichloromethane using triethylsilane as a scavenger to give
1. The chiral purity of 1 was not assessed after the synthesis was completed though Boc-L-serine
was employed as the starting material.
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
ACS Paragon Plus Environment

Page 9 of 44
Journal of Medicinal Chemistry
1
2
3
4
5
6
CO2Allyl
b
OH
OCH3
a
3 steps
NH
NH
N
PO3Et2
BocHN
BocHN
NH
CO2Allyl
PO3Et2
BocHN
O
O
9
O
10
7
8
9
O
N
OH
OH
O
N
N
CO2H
CO2tBu
N
d
N
N
CO2Allyl
CO2tBu
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
c
N
BocHN
BocHN
BocHN
OH
H
N
O
O
CO2tBu
O
O
O
12
11
H2N
N
O
14
OH
13
CO2tBu
O
N
O
OH
O
f
O
O
O
H
N
N
e
N
N
OH
OH
N
N
OH
N
H
H2N
.TFA
N
N
CO2tBu
S
H
O
O
O
N
CO2H
CO2H
N
O
15
O
S
TrtHN
17
OH
N
O
TrtHN
16
O
CO2H
O
N
O
H
N
N
N
OH
OH
g
N
N
S
H
O
N
O
CO2H
O
H2N
YU253434, 1
Scheme 1 Synthesis of YU253434, 1: (a) allyl 2-(diethoxyphosphoryl)acrylate, DCM, RT, 1h,
98%; (b) tert-butyl 2-chloro-2-oxoacetate, DCM, 0 ˚C to RT, Hunig’s base, 18 h 100% crude; (c)
Acetonitrile, triethylsilane, Pd(PPh3)4, 0 ˚C to RT, 65%; (d) i) oxalyl chloride, 12, ii) MSTFA,
Hunig’s base, 13, 38%; (e) 2:1 DCM/TFA, triethylsilane, 0 ˚C to RT, 3.5 h, 100% crude; (f) i)
oxalyl chloride, catalytic DMF, 16, ii) MSTFA, Hunig’s base, 15, 3.5 h, 100% crude; (g) ) 2:1
DCM/TFA, triethylsilane, 0 ˚C to RT, 1.5 h, toluene chase, reverse phase MPLC C18, 0 to 60%
acetonitrile 0.1% formic acid / water with 0.1% formic acid, 38%.
Antimicrobial activity
Pyrazolidione 1 was tested against large panels of P. aeruginosa (197 previously described
clinical isolates),24 K. pneumoniae (100 previously described clinical isolates),25 and E. coli (100
previously described clinical isolates).26,27 A summary of the minimum inhibitory concentrations
(MICs) are presented in Figure 2 and Table 1.
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 44
1
2
3
4
5
6
7
8
9
The MICs compare favorably to representative -lactam agents currently used in the clinic
to treat MDR Gram-negative infections including aztreonam (5, monobactam), ceftazidime (4,
anti-pseudomonal cephalosporin), and meropenem (6, carbapenem). Additional MICs for 15 of
the P. aeruginosa isolates can be found in Supplemental Table 10.1. MICs were also performed
on additional clinical K. pneumoniae (16 isolates) and E.coli (14 isolates) and can be found in
Supplemental Tables 10.2 and 10.3.
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 2. Minimum Inhibitory Concentrations (MICs) of 1 against P. aeruginosa (197 clinical
isolates), K. pneumoniae (100 clinical isolates) and E. coli (100 clinical isolates).
Table 1. MIC50 and MIC90 values (g/mL) for 1 and comparator antibiotics for P. aeruginosa
(197 clinical isolates), K. pneumoniae (100 clinical isolates) and E. coli (100 clinical isolates).a
P. aeruginosa
K. pneumoniae
E. colib
Compound
MIC50 MIC90 MIC50 MIC90 MIC50 MIC90
10
ACS Paragon Plus Environment

Page 11 of 44
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
8
4
2
1
4
8
1
4
1
4
Aztreonam (5)
>16
>16
>8
>32
>64
16
>32
>64
>64
>64
2
<=2
2
>16
>16
8
Ceftazidime (4)
Meropenem (6)
Ceftolozane/Tazobactam
Ceftazidime/Avibactam (8)
<=1
NA
NA
>32
64
NA
>32
0.5
NA
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
aAll strains used are from previously described collections. The reported comparator antibiotic
b
MICs are from the original studies. Data for comparator antibiotics only available for 90
(aztreonam), 98 (ceftazidime), or 99 (meropenem) of the 100 samples in the test panel; NA: No
data available.
The effects of iron concentration on the MICs of 1 were also evaluated against
representative strains of P. aeruginosa (laboratory strain PAO1 and a clinical isolate). Previously
the concentration of free iron in the culture medium was shown to inversely correlate with
antimicrobial activity of siderophore-containing -lactam antibiotics including BAL3007228 and
cefiderocol.29 As shown in Table 2, the addition of soluble iron to the culture medium caused a
similar concentration-dependent decrease in 1’s activity in both strains. This effect was observed
for cefiderocol 7 which also contains a catechol moiety which has been shown to drive its cellular
uptake and antimicrobial activity.29 As a control, ceftazidime 4, which does not have iron-chelating
ability showed MICs independent of media iron concentration. These results are consistent with
the Trojan horse mechanism of siderophore antibiotics and for this reason iron-depleted media was
used for all MIC determinations. The low-iron media follows CLSI protocols for testing
siderophore-containing antibacterial agents, as the modified broth is thought to replicate the low
iron environment found in vivo. To further inform that 1 is transported via a siderophore uptake
mechanism, the isosteric phenethyl amide analog 2, expected to have no siderophore transport
qualities, was synthesized in a manner parallel to 1. Testing in an abbreviated MIC panel of Gram-
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 44
1
2
3
4
5
6
negative isolates showed that the MIC values of 2 were all >64 g/mL, the upper concentration
tested, while 1 displayed significant activity (data not shown).
7
8
9
Table 2. The effect of iron concentration on MIC values for 1 and comparator antibiotics versus
P. aeruginosa PAO1 and the clinical isolate.
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
MIC (g/mL)a
Added iron
g/mL
PAO1
Ceftazidime 4
clinical isolate
Ceftazidime 4
1
0.5
2
16
>16
Cefiderocol 7
1
Cefiderocol 7
0
0.5
4
>16
1
1
1
1
0.13
0.13
0.25
1
1
1
1
1
0.06
0.25
0.5
2
0.1
1.0
10
>16
aMICs were determined using iron-depleted cation-adjusted Mueller-Hinton broth that was
supplemented with iron (as ferric chloride) as indicated. The initial iron-depleted media was
prepared by the standard treatment with cation-exchange resin which has been reported to reduce
iron concentrations to 0.02 g/mL.29 Therefore, the total media iron concentrations as tested are
expected to be 0.02, 0.12, 1.02, and 10.02 g/mL.
-lactamase stability
Appending the phthalimide siderophore mimic from the C(3) amide of the pyrazolidinone was
deliberate and designed to occupy space allowed by the enzyme target PBP3. Prospectively, when
1 was modeled in class A serine -lactamases in a transition state presumed for hydrolysis, we
observed the pyrazolidine ring with its additional methylene group projects the aminothiazole side
chain in a different direction when compared to the aminothiazole group in -lactams like
ceftazidime. The consequence is the -lactam in 1 positions the aminothiazole in a manner where
it is predicted to sterically clash with the omega loop of the class A -lactamases. (Supplemental
Figure 1) This could be one of the reasons why 1 is not hydrolyzed by class A -lactamases and
perhaps the related class C and class D serine -lactamases as well. In fact, testing against isolates
that express class A, C, and D -lactamases, 1 shows significant resistance to hydrolysis (enzyme
stability). This important observation is further supported by testing against representative -
lactamases: KPC-2 (class A carbapenemase), OXA-23 and OXA-24 (class D carbapenemases);
12
ACS Paragon Plus Environment

Page 13 of 44
Journal of Medicinal Chemistry
1
2
3
4
5
6
PER-2 (class A extended-spectrum -lactamase); and PDC-3 (P. aeruginosa class C -lactamase).
In all cases, hydrolysis was not detected. The observed -lactamase stability correlates with the
7
8
9
MIC data (Table 1, Figure 2), as most tested isolates had their -lactamase genes identified by
whole-genome sequencing or rapid molecular diagnostics, and there was no correlation of the
presence of any class of enzyme and the susceptibility of the organism, see Supplemental
Information Tables 9.1–9.3. The clinical implications from this analysis suggest a clinical niche
for this agent.
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
For the class B -lactamases (metallo-enzymes) resistance to hydrolysis was not
anticipated. Interestingly, 1 is a poor substrate for NDM-1, VIM-2, and IMP-1 as compared with
imipenem (Table 3). This central observation is supported by the MIC results where 1 has low
MIC values against isolates expressing class B -lactamases. Though we didn’t anticipate the poor
hydrolysis of 1 in advance, we do recognize that the product of the hydrolysis (Figure 1) has a
unique hydrazide and carboxylate functionality (bold bonds) which are suitable for zinc chelation
and as such could form inhibitory interactions in class B -lactamase active sites.
Table 3. Susceptibility of 1 to -lactamase hydrolysis. Steady state reactions of 1 were followed
against purified enzymes: KPC-2, OXA-23, OXA-24, PER-2, PDC-3, NDM-1, VIM-2, and IMP-
1. Kinetic parameters are provided below. Imipenem was used as a control. No reaction was
detected for KPC-2, OXA-23, OXA-24, or PER-2 using 200 nM enzyme concentration and 100
µM 1.
1
kcat (s–1)
Km (µM)
kcat/Km (µM–1s–1)
Metallo--
lactamase
NDM-1
7 ± 1
292 ± 25
0.020 ± 0.002
0.020 ± 0.002
VIM-2
20 ± 2
840 ± 180
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 44
1
2
3
4
5
6
7
8
9
IMP-1
5 ± 1
76 ± 8
0.070 ± 0.010
Imipenem
65 ± 7
NDM-1
VIM-2
IMP-1
173 ± 18
57 ± 6
2.7 ± 0.3
6.7 ± 0.7
3.2 ± 0.4
8.5 ± 0.9
11 ± 1
36 ± 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
Drug-like attributes
Pyrazolidinone 1 has many characteristics that are favorable for potential administration either by
IV or inhaled routes as 1 has solubility >100 µM in pH 7.5 phosphate buffer as measured by
nephelometry. In this same buffer, 1 has remarkable stability with only about 4% hydrolysis to the
ring opened product after 2 weeks at room temperature. Meropenem in comparison is known to be
unstable after reconstitution in IV bags often showing >30% degradation in 24 hours after
reconstitution. The stability of 1 is a reflection of the mild electron withdrawing nature of the C(3)
amide group as mentioned earlier.18 Though highly soluble, 1 has poor Caco-2 permeability and
would not be expected to have oral bioavailability. Stability to both human and CD-1 mouse
microsomes is notable with a half-life >60 minutes in both preparations. Furthermore, 1 does not
inhibit any of the 8 CYP enzymes it was tested against at a 30 µM concentration. The
pharmacokinetics (PK) of 1 was evaluated both subcutaneously and by intravenous administration
at a dose of 50 mg/kg in CD-1 mice (Table 4). Both routes of dosing result in nearly identical PK
data.
Figure 3. The mean plasma concentration-time profile of 1 after IV (50 mg/kg) administration in
mice.
14
ACS Paragon Plus Environment

Page 15 of 44
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 4. The plasma PK parameters after IV administration of 1 at 50 mg/kg in mice.
The clearance is rapid, but similar to intravenous -lactam antibiotics and may be driven by
excretion through the urine, which was not measured. Blood levels of 1 are maintained above
1µg/mL at the 2-hour timepoint (Figure 3). Notably, 1 µg/mL is the MIC50 of 1 against P.
aeruginosa, K. pneumoniae and E. coli strains tested. Toxicity was also not found when 1 was
tested against primary human hepatocytes at the maximum concentration of 100 µM after 72 hours
of incubation. We further tested 1 against 44 enzyme receptors and channels that might be of a
safety concern at a maximum dose of 30 µM and found only 63% and 66% inhibition of COX-1
and PDE3A, respectively in this panel. Further examination of ion channels shows that 1 does not
bind to hERG at 30 µM concentrations nor to seven other tested ion channels. Collectively, these
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 44
1
2
3
4
data demonstrate promising characteristics of 1 and the pyrazolidinone scaffold as new
5
6
antibacterial agents (Supplementary Information).
7
8
9
PBP binding
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
Previous investigations proposed that pyrazolidinones targeted the crucial PBP3 enzyme. To verify
that 1 could effectively target relevant PBPs consistent with its antimicrobial activity, kinetic
studies were performed using a homogenous preparation P. aeruginosa PBP3, an essential
transpeptidase following a previously established protocol using bocillin, a fluorescent penicillin
analog.30 As shown in Figure 4, increasing concentrations of 1 inhibited bocillin’s ability to
fluorescently label the protein, with an IC50 of 2.5 M (Figure 3). This value is comparable to a
previously published value for doripenem (IC50 = 2.3 ± 0.5 µM, Acinetobacter baumanni PBP3).30
Figure 4. Determination of the IC50 for P. aeruginosa PBP3 using a competitive assay. Bocillin,
a fluorescent substrate of PBP3 was reacted with enzyme that had been pre-incubated with
increasing concentrations of 1. An IC50 was calculated as the concentration of 1 required to reduce
the fluorescence intensity of the bocillin-labeled protein by 50%.
DSF thermal shift analysis of ligand binding to PBP3
To further support that 1 binds PBP3, we next examined the stabilizing effect of 1 binding to P.
aeruginosa PBP3 using the DSF thermal shift assay. Pyrazolidinone 1 binding increased the
melting temperature of P. aeruginosa PBP3 (with 1% DMSO) from 44.1 ± 0.14 °C to 51.0 ± 1.13
°C (Figure 5), about a 7 °C increase. Ceftazidime (4) yielded a larger stabilizing effect; the melting
16
ACS Paragon Plus Environment

Page 17 of 44
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
temperature of its P. aeruginosa PBP3 complex was 56.5 ± 0.14 °C (Figure 5). Despite the
differences in increases in melting temperature between the two ligands, we could not discern
whether there was a correlation of the magnitude of melting temperature increases with the affinity
of ligands for proteins due to other effects such as entropy loss upon binding.
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. DSF assay of 1 and ceftazidime binding to P. aeruginosa PBP3. The derivative of the
change in fluorescence is plotted versus temperature.
PBP3 crystal structure complexed with 1
To probe the mechanism of PBP3 inhibition by the pyrazolidinone γ-lactam, including how an
additional carbon atom in the lactam ring can be accommodated, we determined the 2.4 Å
resolution crystal structure of 1 bound to P. aeruginosa PBP3. The PDBid code is: 6VOT.
The crystallographic refinement of the P. aeruginosa PBP3 structure revealed strong
difference density for the 1 ligand indicating that the catalytic S294 was acylated by the ligand
(Figure 6, 6VOT). Clear electron density is present for most of 1 including the aminothiazole
moiety, the amide moiety adjacent to the carbonyl bond, and the dihydropyrazole-carboxyl moiety
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 44
1
2
3
4
5
6
7
8
9
(Figure 6). Weaker density is present for the other moieties, in particular, the
dihydroxyphthalimide siderophore mimic. This latter groups main function is to promote uptake
via the iron uptake/catechol pathway. Density for the 2-methylpropanoic acid of 1 is also not
particularly strong (Figure 6). The active site protein residues are generally well-ordered around
the ligand 1 (Figure 6C).
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. Electron density of 1 in the active site of P. aeruginosa PBP3. (A) Unbiased electron
difference density (|Fo|-|Fc|) showing density for a covalently bound 1 in the active site. Density
is contoured at 3 σ level. Prior to map calculation, 10 cycles of Refmac refinement were carried
out with 1 removed from the structure. The S349 side chain was observed in two conformations
(labeled 1 and 2, refined with occupancies of 0.6 and 0.4, respectively). 1 is shown with cyan-
colored carbon atoms. (B) Same as in (A) but view is rotated about 100° around the x-axis. (C)
Stereo diagram of 2|Fo|-|Fc| density map of the active site region; density is contoured at 1 σ level.
PDBid 6VOT.
18
ACS Paragon Plus Environment

Page 19 of 44
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The pyrazolidinone 1, when covalently bound in the active site of PBP3, makes several
important interactions (Figure 7). A covalent bond is formed between the carbonyl carbon of 1 and
the side chain of the catalytic S294. Several hydrogen bonds are made and include hydrogen bonds
of the aminothiazole with residues E291 and backbone nitrogen and oxygen atoms of R489. The
aminothiazole ring also makes hydrophobic interactions with A488, Y409, and G293. The amide
moiety is forming hydrogen bonds across the width of the active site with N351 and backbone
oxygen of T487. The carbonyl oxygen is situated in the oxyanion hole making hydrogen bonds
with the backbone nitrogen atoms of S294 and T487. The carboxyl moiety attached to the
dihydropyrazole ring is making hydrogen bonds with S485, T487, and S349 (conformation 2 of
the S349 side chain, see Figure 6A): this pocket is the conserved carboxyl binding pocket. The
amide moiety attached to the dihydropyrazole ring is less well defined, but the structure does
suggest a water-mediated hydrogen bond with water W#1 (Figure 7). The dihydroxyphthalimide
siderophore mimic is not making any direct hydrogen bonding interactions and is pointing to the
solvent in this structure.
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. Interactions of 1 in the active site of P. aeruginosa. PBP3. Hydrogen bonds are depicted as black
dashed lines. Water molecules are shown as round red spheres. Key active site residues are labeled.
Structural comparison of the 1 and ceftazidime complexes of P. aeruginosa PBP3
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 44
1
2
3
4
5
6
7
8
9
The entire left-hand side of 1 is chemically like ceftazidime (4, Figure 1). Previously, the structure
of ceftazidime bound to P. aeruginosa PBP3 was determined.32 We, therefore, compared 1 and
ceftazidime (4) covalently bound PBP3 structures via superposition (Figure 8). This analysis
showed that the aminothiazole moieties in both bound 1 and ceftazidime are in similar although
slightly shifted position (shifted by ~0.5 Å) making almost identical interactions in the PBP3 active
site (Figure 8). The minor shift of the aminothiazole away from the center of the active site is also
apparent for the adjacent amide moiety in 1. These shifts in those moieties of 1 are likely due to
accommodating the extra methylene group in 1 since it contains a γ-lactam, not a β-lactam found
in ceftazidime (4) or other approved PBP-targeting antibiotics. The superposition showed that the
carboxyl moiety attached to the dihydropyrazole ring is in a similar position as the equivalent
carboxyl group in ceftazidime; the oxygens are making hydrogen bonds with both S485 and T487
in both structures (Figure 8). Despite making similar hydrogen bonds, one of the oxygens of this
carboxyl group in 1 is slightly shifted allowing it to also make a hydrogen bond with the side chain
of S349 in conformation 2 (Figure 8). The 5-membered and 6-membered rings attached to these
carboxyl moieties in 1 and ceftazidime, respectively, are in similar positions although of different
sizes. The prime side of the ligands are chemically very different. Compound 1 has a 5,6-
dihydroxyphthalimide siderophore mimic connected via an amide containing linker whereas
ceftazidime contains a pyridine ring at that position (Figure 1). This pyridine ring was not included
in the model of the ceftazidime PBP3 complex (PDBid 3PBO)32 perhaps resulting from a lack of
density due to conformational heterogeneity or due to possible pyridine elimination which can
occur for ceftazidime.38
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
An additional difference between the 1- and ceftazidime-bound PBP3 structures is that the 2-
methylpropanoic acid of 1 does not make a salt bridge interaction with the side chain of R489. The
20
ACS Paragon Plus Environment

Page 21 of 44
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
R489 side chain is more distant compared to that in the ceftazidime bound structure (Figure 8).
This could be because this region of the protein and the ligand are in general more flexible or
perhaps due to that ~0.5 Å shift relative to ceftazidime in that section of the 1 compound. Data
collection and refinement statistics for the structure of P. aeruginosa PBP3 complexed to 1 can be
found in Table 5.
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. Comparison of the 1 and ceftazidime (4)-bound structures of P. aeruginosa PBP3. (A)
Superposition of the PBP3–1 complex (gray protein carbon atoms) onto the PBP3–ceftazidime
structure (gold colored carbon atoms). The ligands are depicted in ball-and-stick and 1 is shown
with cyan carbon atoms. (B) Same as in (A) but a close-up view of the active site region near the
covalent bond with S294 to highlight differences in the β-lactam (ceftazidime, 4) and γ-lactam (1)
based ligands. Hydrogen bonds of 1 and ceftazidime with their respective PBP3 coordinates are
shown as black and grey dashed lines, respectively.
Table 5. Data collection and refinement statistics for crystal structure of P. aeruginosa PBP3
complexed to 1
Wavelength (Å)
Resolution range (Å)
Space group
0.97946
50.00 - 2.40 (2.44 - 2.40)
P21 21 21
Unit cell (Å, °)
68.107 83.813 89.74 90 90 90
167,804
Total reflections
Unique reflections
Multiplicity
20,549 (987)
8.2 (6.8)
Completeness (%)
Mean I/sigma (I)
99.7 (100)
16.1 (4.1)
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 44
1
2
3
4
5
6
7
8
9
Wilson B-factor
31.60
R-merge (%)
Reflections used in refinement
Reflections used for R-free
R-work
13.6 (58.8)
20,503 (1922)
1,057 (80)
0.182 (0.239)
0.236 (0.290)
3951
R-free
Number of non-hydrogen
atoms
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
Macromolecules
Solvent
Protein residues
3866
85
500
0.014
1.85
97.78
2.22
0.00
RMS(bonds, Å)
RMS(angles, °)
Ramachandran favored (%)
Ramachandran allowed (%)
Ramachandran outliers (%)
Statistics for the highest-resolution shell are shown in parentheses.
CONCLUSION
In summary, we described a novel non--lactam penicillin binding protein (PBP) inhibitor,
Compound 1, that contains a -lactam moiety whose carbonyl group covalently modifies the
catalytic serine of the target PBPs (PBP3). The pyrazolidinone 1 has excellent chemical stability
and remarkable activity against WHO priority 1 pathogens such as P. aeruginosa, and
Enterobacterales. Furthermore, 1 has been found to be stable to all 4 classes of -lactamases. X-
ray crystallography helps to define where further optimization can occur and illustrates unique
interactions with P. aeruginosa PBP3. Additionally, 1 was found to have drug like characteristics.
Further testing of 1 and the design and synthesis of additional analogs is planned.
EXPERIMENTAL SECTION
Minimum Inhibitory Concentrations, MICs
22
ACS Paragon Plus Environment

Page 23 of 44
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Bacterial strains used were from previously described collections. MICs were determined using
the general recommendations of the Clinical and Laboratory Standards Institute (CLSI). Standard
broth microdilution methods were followed but with slightly lower inoculum (6 x 104 cfu/mL)
which afforded no difference in MICs in our testing. MICs were also performed using iron-
depleted cation-adjusted Mueller-Hinton broth, except as mentioned elsewhere, using a standard
protocol used with other siderophore-containing antibiotics.
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
PBP Binding Kinetics
Method adapted from the work using purified, soluble P. aeruginosa PBP3 and bocillin, a
fluorescent β-lactam and substrate.30 Reactions were conducted in 10 mM phosphate-buffered
saline at pH 7.4 using 1.6 µM P. aeruginosa PBP3 incubated with increasing concentrations of 1.
To ensure that equilibrium between 1 and PBP3 occurred, the enzyme was preincubated with the
compound for 20 min at 37 °C before addition of 50 µM bocillin and then incubated for an
additional 20 min. The reactions were stopped by adding SDS-PAGE loading dye and boiling for
2 min. Samples were analyzed by SDS-PAGE and the gel illuminated at λ = 365 nm and imaged
with a Fotodyne gel imaging system. ImageJ analysis software was used to assign Fluorescence
intensity (FI). The IC50 was calculated as the concentration of 1 required to reduce the FI of the
bocillin-labeled protein by 50%.
-lactamase stability testing
Steady-state reactions were followed with purified enzymes (KPC-2, OXA-23, OXA-24, PER-2,
PDC-3, NDM-1, VIM-2, and IMP-1) using an Agilent model 8453 diode array spectrophotometer
as previously described.31 Assays were performed at 25 °C (room temperature) using either 10
mM PBS, pH 7.4 (KPC-2, PER-2, and PDC-3), 50 mM sodium phosphate buffer supplemented
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 44
1
2
3
4
5
6
7
8
9
with 20 mM sodium bicarbonate (OXA-23, OXA-24) or 10 mM HEPES, pH 7.5, 0.2 M NaCl,
50µg/ml bovine serum albumin, and 50 µM Zn (NDM-1, VIM-2, and IMP-1).
Compound 1 was used as a substrate at excess molar concentrations to establish pseudo
first-order kinetics and Imipenem (IMI) was used as a control. The following extinction
coefficients were used: 1, Δɛ336 = –5,500 M−1 cm−1 ; IMI, Δɛ300 = –9,000 M−1 cm−1. For velocity
determinations, a 1 cm path length quartz cuvette was employed. A nonlinear least square fit of
the data (Henri Michaelis–Menten equation) using Origin 8.1 (OriginLab, Northampton, MA) was
employed to obtain the steady-state kinetic parameters Vmax, kcat, and Km according to Eqs. 1 and
2:
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
v = (Vmax*[S])/Km + [S])
Eq. 1
kcat = Vmax/[E]
Eq. 2
Summary of Rodent PK Screening
Species: CD-1 mouse
Protocol Summary for PK in rodents (In-life and bioanalysis):
Plasma sample collection from mice (parallel sampling)
Compound 1 was administered to groups of 24 animals. Blood aliquots (300-400 μL) are
collected via cardiac puncture from individual mice anesthetized with isoflurane vapor. For
untreated control animals, blood was collected by anesthetized animals by cardiac puncture.
Blood was drawn into tubes coated with lithium heparin or K2EDTA, mixed gently, then kept
on ice and centrifuged at 2,500 ×g for 15 minutes at 4°C, within 1 hour of collection. The
plasma was then harvested and kept frozen at -70°C until further processing. Quantitative
24
ACS Paragon Plus Environment

Page 25 of 44
Journal of Medicinal Chemistry
1
2
3
4
5
6
Bioanalysis (Plasma): The plasma samples are processed using acetonitrile precipitation and
analyzed by HPLC-MS/MS.
7
8
9
A plasma calibration curve with an internal standard was generated. Aliquots of drug-free plasma
were spiked with the test compound at the specified concentration levels. The spiked plasma
samples were processed together with the unknown plasma samples using the same procedure.
The processed plasma samples are stored at -70°C until the HPLC-MS/MS analysis, at which time
peak areas were recorded, and the concentrations of the test compound in the unknown plasma
samples were determined using the respective calibration curve. The reportable linear range of the
assay is determined, along with the lower limit of quantitation (LLQ). Pharmacokinetics: Plots of
plasma concentration of the compound versus time were constructed. The fundamental
pharmacokinetic parameters of each compound after intravenous and oral dosing (AUClast,
AUCINF, T1/2, Cl, Vz, Vss, Tmax, and Cmax) were obtained from the non-compartmental
analysis (NCA) of the plasma data using WinNonlin. The oral bioavailability will be calculated.
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
Animal use complies with International Guiding Principles for Biomedical Research (CIOMS) and
has an IACUC that operates in accordance with “Regulation for Establishing the Committee of
Care and Use of Laboratory Animals”. The facility and our procedures have been fully accredited
by the Association for Assessment and Accreditation of Laboratory Animal Care International
(AAALAC International) in 2014 and 2017 (#001553). We have a current Foreign Assurance from
the Office of Laboratory Animal Welfare (#A5890-01), issued by the PHS/NIH for vertebrate
animal studies.
Structure determination
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 44
1
2
3
4
5
6
7
8
9
His-tagged P. aeruginosa PBP3 was expressed, purified, and crystallized as described
previously.32 To obtain the complex of PBP3, compound 1 was soaked into apo PBP3 crystals for
15 minutes before freezing the crystal in liquid nitrogen. In these experiments, 5 mM 1 in mother
liquor for 15 minutes was used for the soaking experiments.
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 2.4 Å resolution diffraction data set was collected at the SSRL synchrotron in Stanford
and processed using HKL-3000 (data collection statistics are listed in Table 4).33 The structure was
solved by Molecular Replacement using PHASER34 with the PBP3 protein coordinates as the
starting model.33 The structure was refined using REFMAC535 and model building was carried out
using COOT.36 After refining the protein coordinates, strong difference density was present in the
active site for the 1 compound (Fig. 2). Topology and parameter files for an acylated 1 ligand were
generated using PRODRG;37 coordinates for 1 were included in subsequent cycles of refinement
being covalently attached to S294. The final model included PBP3 residues 53-491 and 501-560,
a single 1 ligand, and 84 water molecules and was refined to an R-factor/Rfree of 0.182/0.236
(additional refinement statistics in Table 4). Figures were generated using Pymol
(www.pymol.org). The coordinates and structure factors of the PBP3:YU253434 complex were
deposited with the Protein Data Bank (PDBid = 6VOT). Authors will release the atomic
coordinates upon article publication.
Differential Scanning Fluorimetry (DSF) to probe 1 binding to PBP3
DSF experiments were carried out with the purified PBP3 at 3 µM protein concentration in 14 mM
Tris pH 8.0, 280 mM NaCl, 7% glycerol, 10X SYPRO Orange, and with or without 1 mM 1. As a
positive control, 0.2 mM ceftazidime was used. The negative control reaction without ligand also
included 1% DMSO to match the DMSO concentration in the ligand containing experiments.
26
ACS Paragon Plus Environment

Page 27 of 44
Journal of Medicinal Chemistry
1
2
3
4
5
6
Experiments were carried out in duplicate using a CFX96 Touch Real-Time PCR Detection
System (Bio-Rad) and the temperature was ramped from 25 to 95 °C.
7
8
9
Synthesis
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 reagents and solvents used for synthesis were of reagent grade quality. Dry solvents were
purchased and used as such. All compounds were individually purified by chromatography on
silica gel or by recrystallization and were of >95% purity for characterization purposes as
determined by LCMS using UV absorption at 220 or 280 nM and/or NMR integration. Each of the
synthesis steps has been replicated by more than one individual in the lab. In practice compounds
were not always purify to >95% purity prior to using in the next synthetic step and often crude
material was of sufficient purity and was carried forward.
Synthesis of tert-butyl (S)-(1-hydrazineyl-3-hydroxy-1-oxopropan-2-yl)carbamate.
To L-Boc-Ser-OMe (25 g, 114 mmol) in methanol (100 mL) was added hydrazine hydrate (25 mL,
402 mmol). The solution was stirred at 50 ˚C overnight. The following day, the solvent was
removed under reduced pressure and the resulting solid dried under hi-vac to a constant weight to
give tert-butyl (S)-(1-hydrazineyl-3-hydroxy-1-oxopropan-2-yl)carbamate (23.4 g, 93.7%). NMR
DMSO 9.0 (s, 1H), 6.5 (d, 1 H), 4.8 (s, 1H), 4.2 (s, 2H), 3.9 (m, 1 H), 3.5 (s, 2 H), 1.4 (s, 9H).
Synthesis of tert-butyl (S)-(3-hydroxy-1-oxo-1-(2-(2,2,2-trifluoroacetyl)hydrazineyl)propan-
2-yl)carbamate.
tert-Butyl (S)-(1-hydrazineyl-3-hydroxy-1-oxopropan-2-yl)carbamate (25 g, 114 mmol) and S-
ethyl trifluorothioacetate (22 mL, 171 mmol) were combined in ethanol (150 mL). The reaction
was stirred at room temperature overnight under a stream of nitrogen. The reaction was
concentrated further to a thick residue, which was crystallized from Et2O/hexanes to give tert-
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 44
1
2
3
4
5
6
7
8
9
butyl (S)-(3-hydroxy-1-oxo-1-(2-(2,2,2-trifluoroacetyl)hydrazineyl)propan-2-yl)carbamate (27.5
g, 76.4%). 1H-NMR DMSO 11.5 (s, 1 H), 10.3 (s, 1H), 6.8 (d, 1H), 4.9 (s, 1H), 4.1 (m, 1 H), 3.6
(d, 2 H), 1.4 (s, 9H).
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
Synthesis of tert-butyl (S)-(3-oxo-1-(2,2,2-trifluoroacetyl)pyrazolidin-4-yl)carbamate.
tert-Butyl (S)-(3-hydroxy-1-oxo-1-(2-(2,2,2-trifluoroacetyl)hydrazineyl)propan-2-yl)carbamate
(20.0 g, 63.4 mmol) and triphenylphosphine (20.0 g, 76.1 mmol) in THF (200 mL) were cooled in
an ice bath. Diethyl azodicarboxylate (31.8 mL, 69.8 mmol, 40% solution in toluene) was added
dropwise, keeping internal temperature below 10 ˚C. The reaction was allowed to slowly come to
room temperature and stirred overnight. The following day the solvent was removed under reduced
pressure and the residue was taken up in 300 mL EtOAc. This mixture was extracted with 3 x 100
mL saturated NaHCO3 solution. The combined aqueous layers were washed with 3 x 100 mL
EtOAc. The aqueous layer was carefully brought to pH 5.0 with 4 M HCl solution to give a white
precipitate, which was filtered and air dried to give tert-butyl (S)-(3-oxo-1-(2,2,2-
trifluoroacetyl)pyrazolidin-4-yl)carbamate (17.6 g, 93% yield). Mass spectrum M+H+(-tBu) =
242.0.
Synthesis of tert-butyl (S)-(3-oxopyrazolidin-4-yl)carbamate (9).
tert-butyl (S)-(3-oxo-1-(2,2,2-trifluoroacetyl)pyrazolidin-4-yl)carbamate (25.7 g, 86.6 mmol) was
suspended in 50 mL H2O and NaOH solution (124 mL, 1 M solution) was added. The reaction was
stirred at room temperature for 5 h, at which point all material was in solution. The reaction was
neutralized to between pH 6-7 using 4 M HCl solution. Solid began to precipitate and the solution
was cooled in an ice bath to facilitate precipitation. The solid was filtered, washed with cold water,
28
ACS Paragon Plus Environment

Page 29 of 44
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
and air dried to give tert-butyl (S)-(3-oxopyrazolidin-4-yl)carbamate (7.3 g, 41.7% yield). H
NMR (400 MHz, DMSO-d6) δ 9.21 (s, 1H), 7.01 (d, J = 8.89 Hz, 1H), 4.42 – 4.02 (m, 1H), 3.53
– 3.19 (m, 2H), 2.90 (t, J = 10.94 Hz, 1H), 1.39 (s, 9H). Mass spectrum M+H+ = 202.1.
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
Synthesis of allyl 2-(diethoxyphosphoryl)acrylate.
A solution of paraformaldehyde (2.24 g, 74.5 mmol) and piperidine (184 µL, 1.86 mmol) in
ethanol (50 mL) was heated at reflux for 1 h. To this was added allyl 2-
(diethoxyphosphoryl)acetate (11.0 g, 46.6 mmol) and the solution was refluxed for 72 hours,
adding additional piperidine (184 µL, 1.86 mmol) and paraformaldehyde (250 mg, 8.3 mmol) after
24 and 48 hours. The solvent was removed under reduced pressure and the reaction was taken up
in toluene (60 mL). To this was added p-toluenesulfonic acid monohydrate (140 mg, 0.81 mmol).
The reaction was distilled to remove ethanol and toluene. Kugelrohr distillation of the remaining
material provided allyl 2-(diethoxyphosphoryl)acrylate (7.37 g, 64% yield), which was stored as a
0.5 M solution in benzene at 0 ˚C. NMR CDCl3 7.02 (d, 1H), 6.77 (d, 1H), 5.95 (m, 1 H), 5.40 (d,
1H), 5.27 (d, 1H), 4.72 (m, 2H) 4.17 (q, 4H), 1.34 (t, 6H).
Synthesis
of
allyl
3-((S)-4-((tert-butoxycarbonyl)amino)-3-oxopyrazolidin-1-yl)-2-
(diethoxyphosphoryl)propanoate (10).
To a suspension of tert-butyl (S)-(3-oxopyrazolidin-4-yl)carbamate (5.0 g, 24.8 mmol) in DCM
(40 mL) was added allyl 2-(diethoxyphosphoryl)acrylate (7.40 g, 29.8mmol, 0.5 M solution in
benzene). The reaction was stirred for 1 h at room temperature. The reaction solution was directly
injected onto a silica gel column equilibrated with DCM. The column was run at 0-10%
MeOH/DCM to give allyl 3-((S)-4-((tert-butoxycarbonyl)amino)-3-oxopyrazolidin-1-yl)-2-
29
ACS Paragon Plus Environment