Strategic Approaches to Overcome Resistance against Gram-Negative Pathogens Using β-Lactamase Inhibitors and β-Lactam Enhancers: Activity of Three Novel Diazabicyclooctanes WCK 5153, Zidebactam (WCK 5107), and WCK 4234

Article abstract of DOI:10.1021/acs.jmedchem.8b00091

Limited treatment options exist to combat infections caused by multidrug-resistant (MDR) Gram-negative bacteria possessing broad-spectrum β-lactamases. The design of novel β-lactamase inhibitors is of paramount importance. Here, three novel diazabicyclooctanes (DBOs), WCK 5153, zidebactam (WCK 5107), and WCK 4234 (compounds 1-3, respectively), were synthesized and biochemically characterized against clinically important bacteria. Compound 3 inhibited class A, C, and D β-lactamases with unprecedented k₂/K values against OXA carbapenemases. Compounds 1 and 2 acylated class A and C β-lactamses rapidly but not the tested OXAs. Compounds 1-3 formed highly stable acyl-complexes as demonstrated by mass spectrometry. Crystallography revealed that 1-3 complexed with KPC-2 adopted a "chair conformation" with the sulfate occupying the carboxylate binding region. The cefepime-2 and meropenem-3 combinations were effective in murine peritonitis and neutropenic lung infection models caused by MDR Acinetobacter baumannii. Compounds 1-3 are novel β-lactamase inhibitors that demonstate potent cross-class inhibition, and clinical studies targeting MDR infections are warranted.

Full text of DOI:10.1021/acs.jmedchem.8b00091

Subscriber access provided by Universiteit Utrecht
Strategic approaches to overcome resistance against Gram
negative pathogens using #-lactamase inhibitors and #-lactam
enhancers: The activity of three novel diazabicyclooctanes,
WCK 5153, zidebactam (WCK 5107), and WCK 4234
Krisztina M. Papp-Wallace, Nhu Q. Nguyen, Michael R. Jacobs, Christopher R. Bethel,
Melissa D. Barnes, Vijay Kumar, Saralee Bajaksouzian, Susan D. Rudin, Philip N. Rather,
Satish Bhavsar, Tadiparthi Ravikumar, Prasad K. Deshpande, Vijay Patil, Ravindra Yeole,
Sachin S. Bhagwat, Mahesh V. Patel, Focco van den Akker, and Robert A. Bonomo
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.8b00091 • Publication Date (Web): 07 Apr 2018
Downloaded from http://pubs.acs.org on April 7, 2018
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 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Strategic approaches to overcome resistance against Gram negative pathogens
using βꢀlactamase inhibitors and βꢀlactam enhancers: The activity of three novel
diazabicyclooctanes, WCK 5153, zidebactam (WCK 5107), and WCK 4234
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
ꢀ3,*
3
2,6
1
Krisztina M. PappꢀWallace , Nhu Q. Nguyen , Michael R. Jacobs , Christopher R. Bethel ,
1,2
3
2,6
1,2
Melissa D. Barnes , Vijay Kumar , Saralee Bajaksouzian , Susan D. Rudin , Philip N.
7
,8
9
9
9
9
Rather , Satish Bhavsar , Tadiparthi Ravikumar , Prasad K. Deshpande , Vijay Patil , Ravindra
9
9
9
3,*
1ꢀ
Yeole , Sachin S. Bhagwat , Mahesh V. Patel , Focco van den Akker , and Robert A. Bonomo
5,*
1
Research Service, Louis Stokes Cleveland Department of Veterans Affairs Medical Center, 10701 East
2
3
4
Boulevard, Cleveland, OH 44106, USA; Departments of Medicine, Biochemistry, Pharmacology,
5
Molecular Biology and Microbiology, Case Western Reserve University School of Medicine, Cleveland,
6
OH 44106, USA; Department of Pathology, University Hospitals, Cleveland Medical Center, Cleveland,
7
OH 44106, USA; Department of Microbiology and Immunology, Emory University School of Medicine
8
Atlanta, GA 30322, USA; Research Service, Atlanta VA Medical Center, Decatur, GA 30033; and
9
Wockhardt Research Centre, Aurangabad, India
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Abstract
Limited treatment options exist to combat infections caused by multidrugꢀresistant (MDR) Gram
negative bacteria possessing broadꢀspectrum βꢀlactamases. The design of novel βꢀlactamase
inhibitors is of paramount importance. Here, three novel diazabicyclooctanes (DBOs), WCK
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
153, zidebactam (WCK 5107), and WCK 4234 (compounds 1ꢀ3) were synthesized and
biochemically characterized against clinically important bacteria. Compound 3 inhibited class
A, C and D βꢀlactamases with unprecedented k /K values against OXA carbapenemases.
2
Compounds 1 and 2 acylated class A and C βꢀlactamses rapidly, but not the tested OXAs.
Compounds 1ꢀ3 formed highly stable acylꢀcomplexes via mass spectrometry. Crystallography
revealed that the KPCꢀ2ꢀcompounds 1ꢀ3 structures adopted a “chair conformation” with the
sulfate occupying the carboxylate binding region. The cefepimeꢀ2 and meropenemꢀ3
combinations were efficiacous in murine peritonitis and neutropenic lung infection models using
MDR Acinetobacter baumannii. Compounds 1ꢀ3 are novel βꢀlactamase inhibitors that
demonstate potent crossꢀclass inhibition and clinical studies targeting MDR infections are
warranted.
ACS Paragon Plus Environment

Page 3 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
INTRODUCTION
In February of 2017, carbapenemꢀresistant Enterobacteriaceae (CRE), Pseudomonas aeruginosa,
and Acinetobacter baumannii were designated as the “top priority pathogens” for the
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
development of novel antibiotics by the World Health Organization . The presence of βꢀ
lactamases in these Gram negative pathogens is largely responsible for this global threat. Of the
serine βꢀlactamases, Klebsiella pneumoniae carbapenemase, KPC, Pseudomonasꢀderived
cephalosporinase, PDC, Acinetobacterꢀderived cephalosporinase, ADC, and the carbapenem
hydrolyzing class D OXA βꢀlactamases (e.g., OXAꢀ23, OXAꢀ24/40, and OXAꢀ48) represent the
most prevalent and problematic βꢀlactam inactivating hydrolases. Presently, three strategies are
used by the pharmaceutical industry to overcome βꢀlactamase mediated resistance: i) the design
of new βꢀlactams resistant to hydrolysis; ii) the synthesis of new βꢀlactamase inhibitors (BLIs);
2
and iii) the development of new antimicrobial classes . An emerging and appealing strategy to
overcome diverse mechanisms of βꢀlactamase mediated resistance deploys an unconventional
concept of a βꢀlactam enhancer (BLE). BLEs represent a new antimicrobial class and work by
providing complimentary pencillin binding protein (PBP) inhibition to a partner βꢀlactam. By
targeting two different PBPs, BLEs act synergistically to promote killing of the bacterial cell.
The remarkable aspect of the BLE mechanism is that it can operate independently of BLIs.
Currently in the United States, the BLIs available in the clinic are clavulanic acid
(approved in 1984), sulbactam (approved in 1986), tazobactam (approved in 1993), avibactam
(approved in 2015), and vaborbactam (approved in 2017). During the past 30 years, clavulanic
acid, sulbactam, and tazobactam have lost efficacy against inhibitorꢀresistant TEMs (IRTs) and
2
inhibitorꢀresistant SHVs (IRSs) . With the exception of tazobactam demonstrating fair activity
3
as an inhibitor against certain class C cephalosporinases (CMYꢀ2, K = 40 ꢁM) , these three BLIs
i
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
possess poor or no activity against the recently emerged class A carbapenemases (e.g., KPC),
class B metalloꢀβꢀlactamases (e.g., NDM and VIM), and many class D OXA βꢀlactamases.
Diazabicyclooctanes (DBOs) are a new class of BLIs (“second generation” BLIs) that
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
were developed in the early 2000s . Avibactam, the first DBO introduced into clinical practice,
inhibits class A carbapenemases, KPC, as well as class C βꢀlactamases, AmpCs, and when
2
, 4ꢀ8
combined with ceftazidime has the ability to target many drugꢀresistant Enterobacteriaceae
Avibactam, however, possesses limited activity towards class D OXA βꢀlactamases
.
4
, 9
.
1
Avibactam is slow to acylate OXAꢀtype βꢀlactamases with low k /K values in the range of 10 to
2
3
ꢀ1 ꢀ1
3
1
0 M s . In contrast, avibactam acylates class A and class C βꢀlactamases in the range of 10 ꢀ
6
ꢀ1 ꢀ1
10 M s . The recently described atomic structures of OXAꢀ24/40 and OXAꢀ48 revealed that
1
the binding pocket especially where the R side chain of avibactam resides in class D βꢀ
lactamases is more hydrophobic with fewer polar residues present, thus potentially affecting
7
, 8
binding and acylation of avibactam .
In the last decade, the DBO class of BLIs expanded to include advanced generation
DBOs with modified chemical scaffolds that enhance their activity. The first generation DBO
4
, 10
inhibitors, avibactam and relebactam inactivate class A, C, and some class D βꢀlactamases
.
The DBO inhibitors WCK 5153 (compound 1), zidebactam, formerly WCK 5107 (compound 2),
nacubactam, FPIꢀ1465, and ETX2514 are “dual action inhibitors” as they inhibit PBPs and
8
, 11ꢀ13
certain βꢀlactamases as well as possessing BLE activity
. Compound 1 and 2 were shown to
be potent inactivators of P. aeruginosa PBPꢀ2 and A. baumannii PBPꢀ2 resulting in enhanced
11
killing of bacteria when combined with a βꢀlactam partner .
Two challenges need to be overcome to design BLIs with promising therapeutic potential
against class D OXA βꢀlactamases associated with Acinetobacter spp.. The first is to readily
ACS Paragon Plus Environment

Page 5 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
access the sterically constrained active site of such βꢀlactamases found in the periplasm and
imparting inhibitory activity. The second is to effectively traverse the outer membrane and
periplasmic space embedded with diverse efflux pumps. Newer generations of BLIs such as
avibactam, relebactam and vaborbactam are devoid of activity against class D OXA βꢀ
lactamases, which illustrates the challenge in designing such compounds. Sulbactam, which is an
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
“older” or legacy BLI, effectively permeates the outer membrane and reaches the periplasmic
compartment of Acinetobacter. However, sulbactam fails to inhibit class D OXA βꢀlactamases,
additional evidence of formidable scientific intricacy of the field. This challenge led us to
explore the effects of introduction of a nitrile functional group onto the DBO core. Unlike
1
avibactam and relebactam, WCK 4234 (compound 3), possessing a nitrile side chain (R ), at the
C2 position of the DBO (Figure 1), demonstrates expanded activity against class D OXA βꢀ
lactamases, including OXAꢀ23 and OXAꢀ51 expressed in A. baumannii, while maintaining
1
4
activity against class A and C βꢀlactamases .
The structure activity relationships (SAR) that prompted the design of compounds 1ꢀ3 are
presented herein. Furthermore, we evaluate 1, 2, 3, and two DBO comparators, avibactam and
relebactam (Figure 1, chemical structures) to gain mechanistic insights into their inhibitory
activity against representative serine class A, C, and D βꢀlactamases using steadyꢀstate kinetics
and mass spectrometry. As a result of these studies, KPCꢀ2 and OXAꢀ24 were chosen for Xꢀray
crystallography analysis to define on an atomic level, the nature of the inactivation mechanism.
Finally, the abilities of cefepimeꢀ2 (WCK 5222) and meropenemꢀ3 (WCK 5999) combinations
were evaluated in susceptibility studies and mouse models of infection. Our data show that
cefepimeꢀ2 and meropenemꢀ3 offer promise against pathogens possessing βꢀlactamases from
multiple classes. Notably, the cefepimeꢀ2 combination has completed two Phase I clinical trials
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(
NCT02707107; NCT02532140) and another Phase I clinical trial is actively recruiting
participants (NCT02942810).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
RESULTS AND DISCUSSION
Synthesis of a Novel Class of DBOs. Starting with a DBO scaffold, a pathway to identify
compounds with activity against Gramꢀnegative PBPꢀ2 and βꢀlactamases led to the evolution of
1
1
this novel bicyclo hydrazide (BCH) series . The key medicinal chemistry challenge involved
1
identifying a side chain (R ) at the Cꢀ2 position of the DBO core that was able to penetrate the
periplasmic space of Gram negative bacteria and access PBP targets on the cytoplasmic
1
5
membrane. After developing SAR with linearꢀchain and cyclicꢀamines , a cyclic secondary
amine was employed on the acyl hydrazide to obtain compounds with high affinity binding for
1
1
1
the PBPꢀ2 of Enterobacteriaceae, P. aeruginosa and A. baumannii . Thus, the hydrazide as a R
side chain bearing a free amine at the appropriate length from the Cꢀ2 position of DBO core was
found to be essential for antibacterial potency. Enhancement of antibacterial activity was further
achieved by chiral separation of the isomers of the racemic secondary cyclic amine. SAR
16
17
generated the lead compounds 1 and 2 and bearing pyrrolidine and piperidine side chains,
respectively (Figure 2A). These compounds demonstrated targeted antibacterial profile
1
1
encompassing Enterobacteriaceae and nonꢀfermenter organisms .
To obtain a DBO that demonstrates expanded antimicrobial activity against A. baumannii
producing class D carbapenemases, OXAꢀ23 and OXAꢀ51, while maintaining activity in Gram
1
4
negatives producng class A and C βꢀlactamases , we next examined the properties of nitrile
1
8
group in the literature . Since nitrile bonds are polarized, they act as hydrogen acceptors and
form hydrogen bonds with amino acids and water and thus can bind to a protein backbone.
ACS Paragon Plus Environment

Page 7 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Furthermore, the strong dipolar nature of the nitrile permits it to act as a hydroxyl or carboxyl
bioisostere. The introduction of nitrile additionally reduces the clogP of the molecule thereby
reducing its lipophilicity and thus improving the chances of penetration of the porin channels in
Gramꢀnegative organisms. The nitrile group is not readily metabolized, which could be
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
advantageous. Importantly, the nitrile reduces the steric bulk of the R substituent of DBOs
such as avibactam and thus could yield a better fit into stericallyꢀconstrained active sites, such as
the class D OXA βꢀlactamase’s active site. Based on our observations, we replaced the amide
functionality of avibactam with a nitrile, generating compound 3(Figure 2B).
Compound 3 is a Highly Potent Inactivator of Class A, C, and D βꢀLactamases. Compound
3 inhibited all βꢀlactamases tested with Ki app values ranging from 0.1 to ≤ 8 ꢁM (Table 1).
Testing KPCꢀ2, 3 possessed the highest acylation rate (k /K value) by nearly 9ꢀfold compared to
2
relebactam. Additionally, 3 was the only DBO capable of acylating OXAꢀ23, OXAꢀ24/40, and
5
ꢀ1 ꢀ1
OXAꢀ48. The highest k /K value obtained was with OXAꢀ48 and 3, 6.4 ± 0.6 x 10 M s .
2
The rate (k ) at which the DBOs deacylate (recyclize) the βꢀlactamases to reꢀform active
off
DBO was also determined (Table 1) The rank order of 3 recyclization after inhibition was as
.
follows: PDCꢀ3 >>> ADCꢀ7 > OXAꢀ48 > OXAꢀ23 > OXAꢀ24/40 > KPCꢀ2. After 3 inhibition,
PDCꢀ3 recovered the ability to hydrolyze nitrocefin the fastest.
The koff values correspond to the residence time halfꢀlives as listed in Table 1.
Comparing all of the koff values, avibactam was the DBO with the lowest koff values. The βꢀ
lactamases/DBO combination with the longest residence time halfꢀlife was OXAꢀ24/40 and
1
9
avibactam at 1,823 min . The turnover number (k /k
tn) at 24 hours or the number of
cat inact =
inhibitor molecules hydrolyzed before the βꢀlactamase is inactivated was also measured (Table
1). KPCꢀ2 possessed a k /k value of 1 for all the DBOs tested. In contrast, ADCꢀ7
cat inact
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
demonstrated turnover of all of the DBOs; the t = 6 at 24 hr. Although insignificant in terms of
n
the bacterial cell cycle, the basis for this catalytic ability is being investigated. To put relative tn
values into context, the t of KPCꢀ2 for clavulanic acid is 2,500 at 15 min with KPCꢀ2 being
n
2
0
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
resistant to inhibition by clavulanic acid .
Compounds 1 and 2 are Potent Inhibitors of Representative AmpC βꢀLactamases.
Compounds 1 and 2 possessed 3ꢀ8 fold and 18ꢀ25ꢀfold lower Ki app values against ADCꢀ7 and
PDCꢀ3, respectively compared to avibactam and relebactam (Table 1). Against ADCꢀ7 and
4
ꢀ1 ꢀ1
5
ꢀ1 ꢀ1
PDCꢀ3, the highest acylation rates (3.1 ± 0.3 x 10 M s ꢀ 6.3 ± 0.6 x 10 M s ) were for 1 and
2.
1
, 2, and 3 Form Highly Stable Complexes with Class A, C, and D βꢀLactamases. Upon
formation of the acylꢀenzyme complex between avibactam and KPCꢀ2, avibactam was shown to
4
desulfate (EꢀI*) and eventually deacylate from KPCꢀ2 (Figure 3A and 3B) . To determine the
intermediates in the reaction pathway between βꢀlactamaseꢀDBO complexes in this study, we
performed timed mass spectrometry with βꢀlactamases that were incubated with the DBOs at an
equimolar ratio for 5 min and 24 hr.
Analyzing the reaction intermediates between KPCꢀ2 and either 1 (+378 ± 5 amu), 2
(+392 ± 5 amu), 3 (+247 ± 5 amu) or relebactam (+349 ±5 amu), formed adducts, did not
undergo a secondary chemical reaction; notably, desulfation was also not observed (Table 2).
However with avibactam, loss of the sulfate (ꢀ97 ± 5 amu) was demonstrated with KPCꢀ2 after a
2
4 hr incubation. For ADCꢀ7, the full adducts were detected with all DBOs, except a minor loss
of the sulfate (ꢀ80 ± 5 amu) was detected with 3 and avibactam at both 5 min and 24 hr (Table
ACS Paragon Plus Environment

Page 9 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2
). Morever, hydrolysis of 1, 2, 3, and avibactam occurred as apoꢀADCꢀ7 (40,639 ± 5 amu) was
noted; only relebactam was not hydrolyzed under these conditions. These data are consistent
with the k /k values obtained for ADCꢀ7 at 24 hr (Table 1).
cat inact
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
During protein purification, two isoforms (40,654 ± 5 amu and 40,785 ± 5 amu) of the
PDCꢀ3 βꢀlactamase were detected. We attribute this finding to a “ragged” Nꢀterminus (different
amino acids at Nꢀterminus) that occurs when proteins are overexpressed. As with the ADCꢀ7 βꢀ
lactamase, a minor loss of the sulfate side chain (ꢀ80 ± 5 amu) was observed with PDCꢀ3 βꢀ
lactamase after incubation with 3 and avibactam at both 5 min and 24 hr. When the OXAꢀ23,
OXAꢀ24/40, or OXAꢀ48 βꢀlactamases were incubated with 3, three peaks were obtained, the
major peak identified was the full adduct (+246 ± 5 amu), while two minor peaks corresponding
to apoꢀenzyme and the enzymeꢀDBO complex minus the sulfate (+166 ± 5 amu) at both 5 min
and 24 hrs. In the reaction of OXAꢀ48 βꢀlactamase with avibactam, only the loss of the sulfate
(ꢀ80 ± 5 amu) was distinguished at 5 min. At 24 hr, the full avibactam adduct (+265 ± 5 amu)
was observed with the OXAꢀ48 βꢀlactamase.
Interestingly, different mass adducts were discerned, a ꢀ80 amu with 3 and avibactam
with the class C and D βꢀlactamases and ꢀ97 amu adduct with avibactam and KPCꢀ2 (Figure
3B). 3 undergoes a different desulfation mechanism (reaction) compared to the one observed
with avibactam and KPCꢀ2 as a ꢀ80 amu peak is detected corresponding to hydrolytic loss of
SO , which is speculated to result in the hydroxylamine. The significance of this finding is being
3
further explored.
Interactions of 1, 2, and 3 with βꢀLactamases are Reversible. To confirm the reversibility of
1
, 2, and 3, an acylꢀtransfer experiment was conducted using KPCꢀ2 as the donor βꢀlactamase
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
and TEMꢀ1 as the recipient βꢀlactamase. The donor βꢀlactamase was preꢀincubated with the
DBO at equimolar ratio. Then, the recipient βꢀlactamase was added to the reaction and at 15 sec
and 5 min, the reactions were terminated and ESIꢀMS was conducted. Avibactam was
previously shown to transfer slowly from TEMꢀ1 to CTXꢀMꢀ15 with a minor CTXꢀMꢀ15ꢀ
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
avibactam population observed at 2 min . The maximum transfer of avibactam from TEMꢀ1 to
CTXꢀMꢀ15 was observed at 30 min. Here, after complete inactivation (1 min incubation time)
by a equimolar ratio of KPCꢀ2 to compounds 1, 2, or 3 was obtained, then equimolar amount of
TEMꢀ1 was added and the reactions were terminated by at the addition 1% acetonitrile and 0.1%
formic acid. The mass spectrometry results implied that 1, 2, and 3 are reversible (Figure 3C).
Within 15 sec, 3 transferred to TEMꢀ1, while for 1 and 2 , 5 min was required. 1, 2, and 3
transfer faster than avibactam, which directly correlates to the k values observed (Table 1).
off
KPCꢀ2 βꢀLactamase Crystal Structures with 1, 2, and 3. The crystal structures of the DBO
inhibitors, 1, 2, and 3 bound to KPCꢀ2 were obtained via soaking experiments and using coꢀ
crystallization (Figure 4, Figure S1, and Table S1). The inhibitors were refined with 100%
occupancy yielding low Bꢀfactors for the inhibitors (Table S1); this indicates full occupancy of
the inhibitors in agreement with the appearance of strong inhibitor electron density. With each
complex structure determined via these two methods, unexpected differences were obtained for
the 1 and 2 structures: we observed desulfation of these two DBO inhibitors when coꢀ
crystallized, but not when soaked into the KPCꢀ2 crystals (Figure S1); these differences were
not observed for 3. Desulfation of a similar DBO, avibactam, by KPCꢀ2 was previously also seen
2
1
and found to be a slow process . We subsequently probed the KPCꢀ2ꢀmediated desulfation of
these two DBOs using mass spectrometry and also observed this phenomenon at lower pH values
ACS Paragon Plus Environment

Page 11 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
similar to the pH of the crystallization condition (Figure S1). Due to the nonꢀphysiological pH
and slow rate of this reaction, our discussion will therefore focus on the structures obtained by
soaking.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The three DBO inhibitors adopt a similar conformation when bound covalently to S70 in
the active site of KPCꢀ2. The 6ꢀmembered ring of the DBO adopts a chair conformation with the
sulfate moiety occupying the carboxylate binding pocket of the active site. The sulfate moiety in
3
was observed to be in two conformations (Figure 4B), whereas this moiety in the other two
DBOs adopted a single conformation. In all three structures, the sulfate makes hydrogen bonds
with T235, T237, S130, and is positioned about 4.5 Å from R220. The carbonyl oxygen in each
of the three DBO complexes is located in the oxyanion hole formed by the backbone nitrogens of
S70 and T237. The Rꢀgroup moieties that distinguish the inhibitors from each other make
different interactions with KPCꢀ2. The Rꢀgroup nitrile moiety in 3 makes a hydrogen bond with
N132 whereas the Rꢀgroups in 1 and 2 make interactions with N132, the deacylation water
(W#1), and some additional interactions with other water molecules (Figure 4B). The Rꢀgroup
in 2 also makes a hydrogen bond with the backbone oxygen of C238; this interaction is not
22
10
observed in the avibactam:KPCꢀ2 complex nor in the relebactam:AmpC complex .
The three DBO inhibitor structures are all in a very similar position and orientation as
shown by superꢀpositioning (Figure 5). These positions and orientations are very similar to that
of avibactam bound to KPCꢀ2; the ring structures of these DBOs all adopt the chair conformation
(Figure 5). The KPCꢀ2 protein in the DBO complexes also adopts a similar conformation except
for residue W105 for which there is considerable conformational variability compounded by
some structures even having two alternative conformations for this residue (Figure 5). We show
molecule B of the 3:KPCꢀ2 complex in the superpositioning to further illustrate this point
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(
Figure 5). The superpositioning also revealed a subtle clustering of the DBO ring positions with
1
and 2 in one cluster, the 3 molecule A and B in a second cluster, and avibactam in a third
slightly different orientation (Figure 5). These differences are likely due to the different Rꢀ
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
group substituents.
OXAꢀ24/40ꢀCompound 3 Crystal Structure. We next analyzed a structure of 3 bound to the
OXAꢀ24/40 βꢀlactamase (Figure 6A and 6B). Overall, 3 binds to the OXAꢀ24/40 βꢀlactamase in
a similar fashion to when bound to KPCꢀ2 except for a few changes mostly due to differences
between the protein. 3 is covalently bound to the catalytic S81 residue. The 3’s sulfate moiety is
interacting with R261, S219, K218, and S128. The carbonyl oxygen atom is situated in the
oxyanion hole, as in KPCꢀ2. Unlike KPCꢀ2, the nitrile moiety of 3 is not directly interacting with
the protein.
In the 3ꢀOXAꢀ24/40 βꢀlactamase complex, we do not observe carboxylation of K84 as
23
was observed in OXAꢀ24/40 complexes bound to either penicillanic acid sulfones or boronic
24
acid inhibitors . Instead, this position is filled by the presence of a chloride ion observed in two
alternate positions with occupancies of 0.63 and 0.37 (labeled Cl1 and Cl2, respectively, in
Figure 6C). This chloride ion is present due to the HCl used to pH the cacodylate buffer.
Superpositioning of the 3:OXAꢀ24/40 structure with that of the avibactam:OXAꢀ24/40
25
complex indicates that avibactam and 3 have a very similar binding mode including a chair
conformation of the DBO ring. The DBO inhibitors sit on the inner side of the active site bridge
formed by residues M223 and Y112. The avibactam:OXAꢀ24/40 complex neither has a
carboxylated K84 nor does not it have a chloride ion bound; instead it has CO bound at this
2
position (Figure 6C). It is interesting to note that the class D OXA βꢀlactamases were found to
ACS Paragon Plus Environment

Page 13 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
26
be inhibited by chloride ions that compete for the carboxylation of K84 including OXAꢀ
27
2
4/40 ; the biological significance of halide binding, if any, is not yet established for OXAꢀ
24/40.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Inhibitory Activity of DBOs against Bacterial Cells. Compound 3, avibactam, and relebactam
are BLIs. 1 and 2 can also inhibit βꢀlactamases, but are additionally able to inactivate PBPs and
1
1
serve as BLEs . As a result, the DBOs either alone or in combination with either cefepime or
meropenem were tested in whole cell assays to obtain minimum inhibitory concentrations
(
MICs). The panel of isolates selected included a set of isogenic strains expressing a single βꢀ
lactamase and clinical isolates of Enterobacteriaceae producing blaKPCꢀ2 or blaOXAꢀ48
,
ceftazidimeꢀavibactamꢀresistant P. aeruginosa, and A. baumannii producing blaOXAꢀ23 and/or
blaOXAꢀ24/40. These isolates represent some of the most difficult to treat Gram negative pathogens
and they produce the most challenging βꢀlactamases to inhibit.
Isogenic E. coli DH10B with pBR322ꢀcatIꢀblaKPCꢀ2 and A. baumannii ꢂblaADC with
pWH1266 blaOXAꢀ23 or blaOXAꢀ24 demonstrated MICs of >16 ꢁg/mL for cefepime and ≥8 ꢁg/mL
for meropenem (Table 3). All MICs of cefepime or meropenem combined with all five DBOs
(
using fixed concentrations of DBOs of 4 and/or 8 ꢁg/mL) against E. coli DH10B with pBR322ꢀ
catIꢀblaKPCꢀ2 were reduced to ≤0.12 ꢁg/mL. Only cefepime combined with 3 restored
susceptibility for A. baumannii ꢂblaADC with pWH1266 blaOXAꢀ23 (MIC = 2 ꢁg/mL). A.
baumannii ꢂblaADC with pWH1266 blaOXAꢀ24/40 was resistant to all combinations tested.
A panel of clinical isolates of Klebsiella pneumoniae that carry either blaKPCꢀ2 or blaKPCꢀ3
including colisitin susceptible (ColS) and colistin resistant (ColR) strains and other
Enterobacteriaceae expressing blaKPC were also tested.
1, 2, 3, avibactam, and relebactam
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
restored susceptibility to cefepime and meropenem in all isolates (Table 3).
Thus for KPC
expressing strains, both cefepime and meropenem in combination with 3 provided comparable
activity.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The DBOs were tested against a panel of clinical isolates of Enterobacteriaceae that
expressed blaOXAꢀ48 or blaOXAꢀ48 mutants (Table 3). 1, 2, 3, and avibactam restored susceptibility
to cefepime and meropenem in all isolates tested even though 1 and 2 lack inhibiton of OXAꢀ48.
Presumably, the 1 and 2 effect is the result of PBP inhibition and their BLE property.
Combinations with relebactam were less effective. Overall, Enterobacteriaceae expressing
blaOXAꢀ48 family members were more susceptible in combinations with meropenem than with
cefepime.
Five clinical strains of P. aeruginosa from an archived collection previously found to be
2
8
resistant to ceftazidimeꢀavibactam were also tested . Combination of cefepime and 1 or 2 at 8
ꢁg/mL restored cefepime susceptibility (MICs ≤0.12ꢀ0.5 ꢁg/mL) to some of these highly drug
resistant strains (Table 3). PBPꢀ2 inhibition and BLE activity of 1 and 2 most likely provided
11
this advantage against P. aeruginosa .
The DBOs were tested against a panel of clinical isolates of A. baumannii that possess
either blaOXAꢀ23 or blaOXAꢀ24/40 or both blaOXAꢀ23 and blaOXAꢀ24/40 (Table 3). Compounds 1, 2, and
3
demonstrated some activity against these strains when combined with cefepime. However,
against dual OXAꢀcarbapenemases expressing strains, most combinations showed high MICs.
Given that 3 is a potent inhibitor of OXAꢀ23 and OXAꢀ24/40, limited penetration of cefepime
and/or 3 is a possible cause for lack of efficacy in these strains. Overall, the meropenemꢀ3
combination was more active against A. baumannii with blaOXAꢀ23 and/or blaOXAꢀ24/40, which is in
1
4
agreement with another study .
ACS Paragon Plus Environment

Page 15 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
To determine that cause of the higher MICs for the βꢀlactamꢀDBO combinations against
A. baumannii, A. baumannii ATCC 17978 was used as the parent strain and different efflux
pump components were knockedꢀout as well as blhA, (βꢀlactam hyper susceptibility), which is a
determinant of intrinsic βꢀlactam resistance and is also involved in cell division. AdeABC,
AdeFGH, AdeIJK, and AcrA systems are the major ResistanceꢀNodulationꢀDivision (RND)
efflux systems in Acinetobacter. These systems are threeꢀcomponent efflux pumps where
AdeA, AdeF, AdeI, AcrA are the membrane fusion protein (MFP), AdeB, AdeG, and AdeJ are
the multidrug transporter and AdeC, AdeI, and AdeK are the outer membrane protein (OMP).
AcrR is the transcriptional regulator of the AcrA RND efflux pump. The AdeB and AdeJ efflux
pump components may have some contribution to cefepime efflux as when the corresponding
genes were deleted, susceptibility was increased (Table S2). Overall, the strains were all
susceptible to single agents as well as the combinations.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
CefepimeꢀCompound 2 and MeropenemꢀCompound 3 was Tested against MDR A.
baumannii in Murine Models of Peritonitis and Neutropenic Lung Infection. The protective
doses (PD and PD ) or the dose of antibiotic that protects 50% or 90% of the infected mice
50
90
was determined for cefepimeꢀ2 and meropenemꢀ3 using a murine peritonitis model. Mice were
intraperitoneally infected with one of three different A. baumannii strains (NCTC 13301 carrying
blaOXAꢀ23 and blaOXAꢀ51ꢀlike, NCTC 13303 possessing blaOXAꢀ26 (blaOXAꢀ24/40ꢀlike) and blaOXAꢀ51ꢀlike
,
and SL46 with blaOXAꢀ23 and blaOXAꢀ51); MICs are presented in Table 4. The mice were treated
with either cefepime, cefepimeꢀ2, meropenem, meropenemꢀ3, tigecycline, or colistin and
survival patterns were monitored for 7 days. The PD50/90 for cefepimeꢀ2 ranged between 50ꢀ100
mg/kg for cefepime and 23.23ꢀ52.30 mg/kg for 2 (Table 4). Remarkably, the cefepimeꢀ2 PD50/90
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
29, 30
provide just 25% exposure compared to the exposures at the selected clinical dose
. For
meropenemꢀ3, the PD50/90 were meropenem: 25ꢀ50 mg/kg and 3: 16.30ꢀ66.17 mg/kg (Table 4).
Tigecycline, although employed at supraꢀtherapeutic doses (6.25 mg/kg administered as two
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
doses) failed to protect infected mice (Table 4). Considering the colistin MIC values of 0.5ꢀ1
31
ꢁ
g/mL, colistin appropriately exhibited low PD50/90 values (Table 4) .
To determine the in vivo eradication efficacy of cefepimeꢀ2 and meropenemꢀ3, a
neutropenic lung infection model was used in which mice were intranasally infected with a
clinical isolate of A. baumannii SL04 carrying blaOXAꢀ23 and blaOXAꢀ51. The MICs for this isolate
are as follows: cefepime: 256 ꢁg/mL, 2: ≥512 ꢁg/mL, cefepimeꢀ2: 32 ꢁg/mL, imipenem: 64
ꢁg/mL, meropenem: 64 ꢁg/mL, meropenemꢀ3: 8 ꢁg/mL, colistin: 2 ꢁg/mL, and tigecycline: 4
ꢁ
g/mL. Thus, the MIC of cefepime was reduced by 8ꢀfold when in combination with 2. Two
hours post infection, mice were treated by q2h dosing for 24 h by a subcutaneous route. Three
hours post last dose (i.e. 27 h post infection), the murine lungs were removed and colony
forming units (CFUs) were determined. Treatment of infected mice with cefepime (50 mg/kg),
meropenem:cilastatin (25ꢀ25 mg/kg), and imipenem:cilastatin (25ꢀ25 mg/kg) revealed a 0.67ꢀ
1
.31 log increase in bacterial burden as compared to the 2 h count (Figure 7A). Treatment
with cefepimeꢀ2 (50ꢀ8.33 mg/kg) resulted in a 3 log kill (Figure 7B). Notably, the addition of 2
32
to cefepime enhanced the bactericidal action of cefepime even at very low dose of 50 mg/kg .
Similarly, the addition of 3 at 4.68 mg/kg, to meropenem:cilastatin (25ꢀ25 mg/kg) resulted in a
2.47 log kill (Figure 7B). Conversely, relebactam at 18.75 mg/kg combined with
imipenem:cilastatin at 25 mg/kg resulted in a 0.86 log increase in CFUs (Figure 7B). It is
important to note that A. baumannii NCTC 13301, SL 46 and SL04 produce OXAꢀ23 and OXAꢀ
51, which are not inhibited by 2. Even with higher MICs, the cefepimeꢀ2 combination provided
ACS Paragon Plus Environment

Page 17 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
33
potent in vivo efficacy against these strains based on its BLE mechanism of action . The most
significant feature of the BLE mechanism is the augmentation of the pharmacodynamic action
of the partner βꢀlactam antibiotic. In addition, BLEs enhance the arsenal of agents required to
overcome βꢀlactamases. A further challenge in designing novel BLIs also stems in imparting
them with structural features which facilitate their efficient penetration into MDR Gram negative
pathogens. Many contemporary MDR pathogens express efflux pumps as well as mutations in
the genes encoding various outer membrane proteins. These mechanisms could impact the
uptake of some newer BLIs that show potent activity against inhibitor resistant enzymes. 3 by
virtue of its potent BLI activity for OXAꢀcarbapenemases and ease of penetration restores the
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
1
efficacy of meropenem against MDR A. baumannii .
CONCLUSIONS
Here, we revealed the biochemical activity of three novel DBOs, compounds 1, 2, and 3 against
some of the most challenging βꢀlactamases in Gram negative bacteria (i.e., KPCꢀ2, PDCꢀ3,
ADCꢀ7, OXAꢀ23, OXAꢀ24/40, and OXAꢀ48). We also verified in vivo efficacy of cefepimeꢀ2
and meropenemꢀ3. 3 was found to extend the kinetic inhibitory profile of DBOs to OXA
carbapenemases, while maintaining class A and C activity. 1 and 2 demonstrated increased
potency against class C βꢀlactamases compared to avibactam and relebactam. 1 and 2 are unique
DBOs in that they also target PBPs of many Gram negative pathogens and have BLE activity.
Thus, they work synergistically when paired with an appropriate βꢀlactam antibiotic and
overcome βꢀlactamase mediated resistance without the need to inhibit multiple βꢀlactamases.
The addition of compounds 1, 2, and 3 to our armamentarium would be highly beneficial as the
clinically available βꢀlactams and βꢀlactamꢀBLI combinations are not effective against these
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
refractory Gram negative pathogens. Finally, compounds 1, 2, and 3 exemplify two divergent
strategies in rejuvenating βꢀlactam antibiotics.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
EXPERIMENTAL SECTION
Compound Characterization. Nuclear magnetic resonance spectra were recorded on Mercury
4
00 MHz (Varian Inc.) or 500 MHz (Bruker). The mass spectra were recorded on TQD mass
spectrometer (Waters Corp.) using the electrospray ionization technique. Elemental analysis was
conducted on VarioꢀMicro cube elemental analyser (Elementar). Water content was determined
using Karlꢀfisher titration method. The purity (>95%) of all the compounds was established by
using high pressure liquid chromatography method with 5 µm particle size C18 columns (Bonna
Agela Technologies for 2 and YMC Technologies for remaining compounds) maintaining
solution at 10°C.
Synthesis of 1 and 2. Figure 2A represents construction of BCH backbone intermediate 1b and
2
b, as a result of coupling of corresponding chiral acid hydrazide 1a or 2a with sodium salt of
DBO nucleus represented by A, by using EDC.HCl as a coupling agent in water (~87% n = 1 and
2). Subsequent catalytic hydrogenation (10% Pd/C catalyst) furnished hydroxyl intermediate (1c,
n = 1 or 2c, n = 2), in quantitative yield which was subjected to sulfonation reaction immediately
by using sulfurtrioxide pyridine complex to afford sulfonated intermediate. Sulfonated
intermediate was isolated in the form of tetrabutylammonium salt 1d (60%, n = 1) or 2d (87% n
=
2). Finally, removal of NꢀBoc group and tetrabutyl ammonium salt was achieved in one step
using excess TFA at 0 to 5°C. Compounds were purified by crystallizing in aqueous isoꢀpropanol
to furnish 1 (1e, n = 1, 60%,), or 2 (2e, n = 2, 80%).
ACS Paragon Plus Environment

Page 19 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
34
Synthesis of 3. Figure 2B represents the synthesis of 3 , where the sodium salt of DBO core
(
A) was converted into the mixed anhydride by first treating the sodium salt with triethylamine
hydrochloride in dichloromethane and reacting the resulting compound with pivaloyl chloride in
presence of a base triethylamine at 0ꢀ5 °C, to obtain the mixed anhydride which was as such
reacted with a 25% aqueous solution of ammonia in water at ꢀ20°C, obtain the amide (3a) as offꢀ
white solid after workup and purification. The amide (3a) was dehydrated with trifluoroacetic
anhydride in the presence of triethylamine in dichloromethane at ꢀ5°C to RT, to obtain the cyano
compound (3b). The cyano compound (3b) was deꢀbenzylated with 10% Pd/C (50% wet), in a
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
: 1 (v/v) mixture of DMF: DCM (v/v) under hydrogen atmosphere (50 to 55 psi) to obtain the
hydroxyl compound (3c) which was immediately sulphated with DMF:SO complex to obtain
3
the sulphate which was isolated as its tetrabutylammonium salt (3d) by reacting with
tetrabutylammonium acetate. The tetrabutylammonium salt (3d) was converted into 3 by passing
the salt through a column filled with Indion 225 sodium resin. 3 was obtained as a white solid by
lyophyllization of the aqueous solution and reꢀcrystallization of the dry powder from a mixture
of isopropanol:water.
Synthesis
of
(2S,
5R)ꢀ6ꢀsulfooxyꢀ7ꢀoxoꢀ2ꢀ[((3R)ꢀpyrrolidineꢀ3ꢀcarbonyl)ꢀ
hydrazinocarbonyl] ꢀ1,6ꢀdiazaꢀbicyclo[3.2.1]octane, 1 (1e)ꢀ Stepꢀ1: (2S, 5R)ꢀ 6ꢀbenzyloxyꢀ7ꢀ
oxoꢀ2ꢀ[((3R)ꢀNꢀBocꢀpyrrolidineꢀ3ꢀcarbonyl)ꢀhydrazino carbonyl]ꢀ1,6ꢀdiazaꢀbicyclo[3.2.1]octane
1
(
1b): Analysis of analytical sample: White solid, 95 gm, 88%; H NMR (CDCl ): δ 8.61 (br s,
3
1
H), 8.21 (br d, 1H), 7.36ꢀ7.43 (m, 5H), 5.04 (d, J = 11.2Hz, 1H), 4.90 (d, J = 11.6 Hz, 1H), 3.99
(d, J = 6.8 Hz, 1H), 3.60ꢀ3.70 (m, 1H), 3.48ꢀ3.52 (m, 2H), 3.24ꢀ3.40 (m, 2H), 3.04ꢀ3.18 (m, 2H),
2
.98 (t, J = 11.6 Hz, 1H), 2.24ꢀ2.34 (m, 1H), 2.12 (br s, 2H), 1.91ꢀ2.04 (m, 2H), 1.58ꢀ1.66 (m,
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
H),1.43 (s, 9H); Mass: (Mꢀ1) = 486.3 for C H N O , HPLC purity: 98.89%. Stepꢀ2: (2S, 5R)ꢀ
2
4
33
5
6
6
ꢀhydroxyꢀ7ꢀoxoꢀ2ꢀ[((3R)ꢀNꢀBocꢀpyrrolidineꢀ3ꢀcarbonyl)ꢀhydrazinocarbonyl]ꢀ1,6ꢀdiazaꢀ
1
bicyclo[3.2.1] octꢀane (1c): Analysis of analytical sample: White solid, 72 gm, quantitative. H
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
NMR (400 MHz, DMSOꢀd ): δ 9.70ꢀ9.90 (m, 3H), 3.76 (d, J = 7.6 Hz, 1H), 3.59 (br s, 1H),
6
3
1
.41ꢀ3.46 (m, 1H), 3.18ꢀ3.29 (m, 2H), 3.13ꢀ3.17 (m, 2H), 2.96ꢀ2.99 (m, 2H), 1.82ꢀ2.05 (m, 4H),
.71ꢀ1.81 (m, 1H),1.53ꢀ1.71 (m, 1H), 1.37 (s, 9H); Mass: (Mꢀ1): 396.2 for C H N O . Stepꢀ3:
17 27
5
6
Tetrabutyl ammonium salt of (2S, 5R)ꢀ6ꢀsulfooxyꢀ7ꢀoxoꢀ2ꢀ[((3R)ꢀNꢀBocꢀpyrrolidineꢀ3ꢀ
carbonyl)ꢀhydrazino carbonyl]ꢀ1,6ꢀdiazaꢀbicyclo[3.2.1]octane (1d): Analysis of analytical
1
sample: White foamy solid, 106 gm, 87%; H NMR (400 MHz, CDCl ): δ 8.76 (br s, 1H), 8.60
3
(s, 1H), 4.23 (br s, 1H), 3.97 (d, J = 8.0 Hz, 1H), 3.58ꢀ3.68 (m, 1H), 3.46ꢀ3.54 (m, 2H), 3.30ꢀ
3
1
.34 (m, 1H), 3.22ꢀ3.27 (m, 9H), 3.08ꢀ3.11 (m, 1H), 12.28ꢀ2.33 (m, 1H), 2.10ꢀ2.17 (m, 5H),
.83ꢀ1.89 (m, 1H), 1.57ꢀ1.71 (m, 9H), 1.36ꢀ1.46 (m, 15H), 0.98 (t, 12H); Mass: (Mꢀ1): 476.4 as
a free sulfonic acid C H N O S.N(C H ) . Stepꢀ4: 1: (2S, 5R)ꢀ6ꢀsulfooxyꢀ7ꢀoxoꢀ2ꢀ[((3R)ꢀ
1
7
26
5
9
4
9 4
pyrrolidineꢀ3ꢀcarbonyl)ꢀhydrazinocarbonyl]ꢀ1,6ꢀdiazaꢀbicyclo[3.2.1]octane (1e): Analysis of
1
analytical sample: White crystalline solid, 33 gm, 60%; H NMR (400 MHz, DMSOꢀd D O
6
,
2
exchange) : δ 4.03 (br s, 1H), 3.85 (d, J = 7.2 Hz, 1H), 3.02ꢀ3.39 (m, 7H), 2.17ꢀ2.22 (m, 1H),
13
1
.99ꢀ2.06 (m, 2H), 1.87ꢀ1.90 (br s, 1H), 1.70ꢀ1.76 (m, 1H), 1.60ꢀ1.64 (m, 1H); C NMR (100
MHz, DMSOꢀd ): δ 171.1, 168.9, 166.1, 58.4, 57.7, 47.0, 46.9, 45.3, 40.1, 28.7, 20.6, 18.4; IR
6
ꢀ1
(cm ): 3568, 1746, 1709, 1676, 1279, 1234, 1015; Analysis calculated for 1 monohydrate
C H N O S.H O C, 36.42; H, 5.31; N, 17.70. Found: C, 36.75; H, 5.42; N, 17.69; Mass: (Mꢀ1):
12
19
5
7
2
3
76.3 for C H N O S; HPLC purity 98.46%; Column: 5 µm particle size, 25 cm length YMC
12 19 5 7
ODS AM C18 column (YMC Technologies); mobile phase: used was buffer (0.1 M ammonium
dihydrogen phosphate in water) and methanol in gradient mode; detection: 225 nm; column
ACS Paragon Plus Environment

Page 21 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
temperature: 25°C; flow rate: 0.7 mL/min.; 1 solution was maintained at 10°C. Specific rotation:
2
5
[
α] D:ꢀ47.5° (c 0.5, water).
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Synthesis of (2S, 5R)ꢀ6ꢀsulfooxyꢀ7ꢀoxoꢀ2ꢀ[((3R)ꢀpiperidineꢀ3ꢀcarbonyl)ꢀhydrazinocarbonyl]ꢀ
,6ꢀdiazaꢀbicyclo[3.2.1]octane 2 (2e)ꢀ Stepꢀ1: (2S, 5R)ꢀ 6ꢀbenzyloxyꢀ7ꢀoxoꢀ2ꢀ[((3R)ꢀNꢀBocꢀ
1
piperidineꢀ3ꢀcarbonyl)ꢀhydrazinocarbonyl]ꢀ1,6ꢀdiazaꢀbicyclo[3.2.1]octane (2b): To a clear
solution of sodium (2S, 5R)ꢀ7ꢀoxoꢀ6ꢀbenzyloxyꢀ1,6ꢀdiazaꢀbicyclo[3.2.1]octaneꢀ2ꢀcarboxylate
(A, 200 gm, 0.67 mol) in water (3.2 L) was added (R)ꢀNꢀBocꢀpiperidineꢀ3ꢀcarboxylic acid
hydrazide [2a, prepared by heating RꢀethylꢀNꢀBocꢀnipecotic acid and hydrazine, Specific
2
5
rotation: [α] = ꢀ53.5° (c 0.5, methanol), HPLC purity: 99%, 171 gm, 0.70 mol]. The resulting
D
suspension was treated with EDC.HCl (193 gm, 1.01 mol), and HOBt (90.6 gm, 0.67 mol) with
vigorous stirring at ambient temperature. The resultant precipitation was filtered after 16 h, and
the wet cake was suspended in warm water. The suspension was filtered and dried under vacuum
at 45°C to furnish coupled intermediate 2b as a white powder in 270 gm quantity in 87% yield.
NMR analyses is shown below and based on the mass, the product was identified as C H N O
6.
2
5
35
5
1
Analysis of analytical sample: H NMR (400 MHz, CDCl ): δ 8.42 (br s, 2H), 7.37ꢀ7.43 (m, 5H),
3
5
3
.06 (d, J = 12.2 Hz, 1H), 4.91 (d, J = 12.2 Hz, 1H), 4.00 (d, J = 7.6 Hz, 1H), 3.84 (br s, 1H),
.52ꢀ3.76 (m, 1H), 3.40ꢀ3.51 (m, 1H), 3.30 (br s, 1H), 3.18 (br d, J = 12.0 Hz, 1H), 3.06 (br d, J
=
12.0 Hz, 1H), 2.43 (m, 1H), 2.29ꢀ2.34 (m, 1H), 1.88ꢀ2.00 (m, 4H), 1.60ꢀ1.77 (m, 4H), 1.44 (s,
1
3
9H). C NMR (100 MHz, DMSOꢀd ) : δ 172.3, 168.6, 167.3, 154.9, 135.4, 129.2, 128.7, 128.5,
6
ꢀ1
8
0.1, 78.2, 59.1, 57.7, 47.7, 45.5, 44.5, 40.8, 28.3, 27.5, 24.0, 20.6. 17.4. IR (cm ): 2974, 1751,
1
670, 1244, 1153, 1024. Analysis calculated for C H N O : C, 59.86; H, 13.96; N, 7.03.
2
5
35
5
6
Found: C, 60.26; H, 14.20; N, 7.20. Mass: (M+1): 502.3 for C H N O . HPLC purity: 98.4%.
2
5
35
5
6
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Stepꢀ2: (2S,5R)ꢀ6ꢀhydroxyꢀ7ꢀoxoꢀ2ꢀ[((3R)ꢀNꢀBocꢀpiperidineꢀ3ꢀcarbonyl)ꢀhydrazinoꢀcarbonyl]ꢀ
1
,6ꢀdiazaꢀbicyclo[3. 2.1]octane (2c): The coupled intermediate 2b from stepꢀ1 (153 gm, 0.305
mol) was dissolved in methanol (1.53 L). To this clear solution, was added 10% Pd/C (15.3 gm,
0% wet) catalyst. The suspension was purged with hydrogen gas for 3 h at 35°C under stirring.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
The catalyst was filtered. The filtrate was evaporated under vacuum below 40°C to provide a
crude residue. The residue was triturated with hexanes. The solid was filtered to furnish hydroxyl
intermediate 2c in 125 gm quantity as a white solid in quantitative yield. The intermediate was
1
used immediately for the next reaction. Analysis of analytical sample: H NMR (500 MHz,
DMSOꢀd ) : δ 9.75 (br s, 3H), 3.90 (br d, 2H), 3.77 (d, J = 10 Hz, 1H), 3.60 (br s, 1H), 3.20 (d, J
6
= 10 Hz, 1H), 2.96ꢀ3.01 (m, 1H), 2.60ꢀ2.73 (m, 1H), 2.28ꢀ2.30 (m, 1H), 2.01ꢀ2.04 (m, 1H), 1.88ꢀ
13
1
.94 (m, 1H),1.72ꢀ1.82 (m, 3H), 1.51ꢀ1.68 (m, 4H), 1.40 (s, 9H); C NMR (125 MHz, DMSOꢀ
d6) : δ 171.87, 168.97, 166.83, 153.79, 78.78, 58.57. 57.71, 47.09, 40.40, 28.05, 27.42, 20.33,
18.29. Mass: (Mꢀ1): 410.3 for C H N O HPLC purity: 96.34%. Stepꢀ3: Tetrabutyl
1
8
29
5
6;
ammonium salt of (2S, 5R)ꢀ6ꢀsulfooxyꢀ7ꢀoxoꢀ2ꢀ[((3R)ꢀNꢀBocꢀpiperidineꢀ3ꢀcarbonyl)ꢀ
hydrazinocarbonyl]ꢀ1,6ꢀdiazaꢀbicyclo[3.2.1] octꢀane (2d): To the clear mixture of hydroxyl
intermediate 2c (113 gm, 0.274 mol), in dichloromethane (1.13 L) was added triethylamine (77
ml, 0.548 mol) followed by sulfurtrioxide pyridine complex (57 gm, 0.356 mol) at 35°C. The
reaction mixture was stirred for 3 hr and added 0.5 M aqueous potassium dihydrogen phosphate
(1.13 L) followed by ethyl acetate (2.26 L). Aqueous layer was washed with dichloromethane:
ethyl acetate mixture (1:2 v/v, 2.26 L twice). The aqueous layer was stirred with solid tetrabutyl
ammonium hydrogen sulfate (84 gm, 0.247 mol) for 3 hr. The mixture was extracted with
dichloromethane (1.13 L). The organic layer was evaporated under vacuum to provide crude
TBA salt. It was purified on a short silica gel column to provide pure intermediate 2d as a foamy
ACS Paragon Plus Environment

Page 23 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
white solid in 122 gm (60%) quantity. Analysis of analytical sample: H NMR (400 MHz,
CDCl ) : δ 8.54 (br s, 2H), 4.31 (br s, 1H), 3.98 (d, J = 8 Hz, 2H), 3.18ꢀ3.37 (m, 12H), 2.45 (br s,
3
1H), 2.32ꢀ2.37 (m, 1H), 2.10ꢀ2.20 (br m, 1H), 1.83ꢀ1.96 (m, 5H), 1.62ꢀ1.73 (m, 10H), 1.26ꢀ1.49
1
3
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
m, 18H), 1.00 (t, 12H); C NMR (100 MHz, CDCl ) : δ 172.32, 168.50, 165.82, 154.78, 60.22,
3
5
4
5
9.30, 58.40, 57.79, 47.86, 40.68, 28.22, 23.96, 23.70, 20.49, 19.49, 17.28. 13.49; Mass: (Mꢀ1):
90.4 as a free sulfonic acid for C H N O S.N(C H ) ; HPLC purity: 96.3%. Stepꢀ4 2: (2S,
18 28
5
9
4
9 4
R)ꢀ6ꢀsulfooxyꢀ7ꢀoxoꢀ2ꢀ[((3R)ꢀpiperidineꢀ3ꢀcarbonyl)ꢀhydrazinocarbonyl]ꢀ1,6ꢀdiazaꢀ
bicyclo[3.2.1]octane (2e): To the clear solution of TBA salt intermediate 2d (113 gm, 0.154 mol)
in dichloromethane (280 ml) was added trifluoroacetic acid (280 ml) between 0°C to 5°C via
addition funnel. The excess trifluoroacetic acid and solvent was evaporated under vacuum below
4
0°C to provide pale yellow oily residue. Oily residue was triturated with methyl tertꢀbutyl ether
(
2.25 L X 2). The precipitate was filtered and was suspended in acetone (1.130 L) and the pH of
suspension was adjusted to 5.5 to 6.5 (aliquot slurry was taken out and made a clear solution by
adding water prior to measure pH) by adding 10% solution of sodiumꢀ2ꢀethyl hexanoate in
acetone. The suspension was filtered and dried under vacuum below 40°C to furnish 65 gm
crude 2. The crude 2 was purified in 15% aqueous isopropanol to provide white crystalline 2 (2e)
1
in 48 gm quantity in 80% yield. Analysis of analytical sample: H NMR (400 MHz, DMSOꢀd ) :
6
δ 9.99 (br s, 2H), 8.39 (br s, 2H), 4.03 (br s, 1H), 3.84 (d, J = 6.8 Hz, 1H), 3.12ꢀ3.23 (m, 3H),
2
.99ꢀ3.05 (m, 2H), 2.91 (t, J = 10 Hz, 1H), 2.68 ( br m, 1H), 1.99ꢀ2.05 (m, 1H), 1.57ꢀ1.90 (m,
1
3
7H). C NMR (100 MHz, DMSOꢀd ): δ 171.1, 168.9, 165.9, 58.3, 57.5, 46.9, 44.3, 43.2, 36.9,
6
ꢀ1
2
5.8, 20.9, 20.5. 18.4. IR (cm ): 3509, 1749, 1715, 1676, 1279, 1242, 1030. Analysis calculated
for 2 dihydrate C H N O S.2H O C, 36.53; H, 5.89; N, 16.39. Found: C, 36.26; H, 5.92; N,
1
3
21
5
7
2
16.30. Mass: (M+1): 392.3 for C H N O S. HPLC purity 98.84%; Column: 5 µm particle size
1
3
21
5
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Unisol C18 column (Bonna Agela Technologies); mobile phase: buffer (0.01 M ammonium
dihydrogen phosphate water and pH adjusted to 7.2 ± 0.1 with dilute ammonia solution) and
acetonitrile in gradient mode; detection: at 225 nm. column temperature: 25°C; flow rate: 1
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
25
mL/min. 2 solution was maintained at 10 °C. Specific rotation: [α] D: ꢀ32.6° (c 0.5, water)
Synthesis of (2S, 5R)ꢀ1,6ꢀdiazaꢀbicyclo[3.2.1]octaneꢀ2ꢀcarbonitrileꢀ7ꢀoxoꢀ6ꢀ(sulfooxy)ꢀ mono
sodium salt (3)ꢀ Step 1: Preparation of (2S,5R)ꢀ6ꢀ(benzyloxy)ꢀ7ꢀoxoꢀ1,6ꢀdiazaꢀ
bicyclo[3.2.1]octaneꢀ2ꢀcarboxamide (3a): Triethylamine hydrochloride (104 g, 0.755 mole) was
added to a suspension of Sodium (2S, 5R)ꢀ6ꢀ(benzyloxy)ꢀ7ꢀoxoꢀ1,6ꢀdiazabicyclo[3.2.1]octaneꢀ2ꢀ
carboxylate A (150 gm, 0.503 mol) in DCM (1.5 L) under stirring at 30°C and after 1 h
triethylamine (70 ml, 0.503 mol) was added followed by drop wise addition of pivaloyl chloride
(74 ml, 0.603 mol) at 0°C to 5°C and stirring continued further for 1h. The reaction mixture was
cooled to –20°C and aqueous ammonia (103 ml, 1.51 mol) was slowly added. After 30 min of
stirring at –20°C, water (1.5 L) was added and the two layers separated. The aqueous layer was
extracted with fresh DCM (750 ml). The combined organic layer was dried over anhydrous
Na SO and evaporated under reduced pressure (200 mm Hg). The concentrate was diluted with
2
4
nꢀbutyl chloride (450 ml) and the mixture stirred for 2 h. The separated solid was filtered and the
solid residue washed with fresh nꢀbutyl chloride (100 ml). The solid was dried under reduced
pressure (4 mm Hg) to obtain the product (3a), as white solid, 55 g, (yield 40 %). Mp: 170ꢀ172
o
1
C. Analysis of analytical sample: HꢀNMR (500MHz, CDCl ): δ 7.42ꢀ7.27 (m, 5H), 6.60 (1H,
3
s), 5.71( s, 1H), 5.06 (d, 1H, J = 10 Hz), 4.91 (d, 1H, J =10 Hz), 4.95 (d, 1H, J =10 Hz), 3.32 (s,
1
H), 3.04 (d, 1H, J = 10 Hz), 2.77 (d, 1H, J = 10 Hz ), 2.37ꢀ2.33 (1H, m), 2.02ꢀ1.92 (m, 2H),
ACS Paragon Plus Environment

Page 25 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
.62 (1H, m). Mass: (M+1): 276 for C H N O . HPLC purity: 99.11 %. Step 2: Synthesis of
1
4
17
3
3
(
2S, 5R)ꢀ6ꢀ(benzyloxy)ꢀ7ꢀoxoꢀ1,6ꢀdiazaꢀbicyclo[3.2.1]octaneꢀ2ꢀcarbonitrile (3b): Trifluoroacetic
anhydride (48 ml, 0.340 mol) was slowly added to a solution of (2S,5R)ꢀ6ꢀ(benzyloxy)ꢀ7ꢀoxoꢀ
,6ꢀdiazaꢀbicyclo[3.2.1]octaneꢀ2ꢀcarboxamide, 3a (47.0 g, 0.170 mol), in DCM (1430 ml)
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
containing triethylamine (107 ml, 0.765 mol) at ꢀ5°C, under stirring. After 2 h of stirring at ꢀ5°C,
water (1450 ml) was added to the reaction mixture and stirring continued for 15 minutes. The
organic layer was separated and washed with saturated sodium bicarbonate solution (470 ml),
with brine (470 ml) dried over anhydrous Na SO and evaporated under reduced pressure (200
2
4
mm Hg). The crude thus obtained was purified by column chromatography over silica gel (60ꢀ
20 mesh) using 10ꢀ20% v/v mixtures of acetone: hexane as an eluent. Evaporation of the
solvent from the combined fractions, gave the product (3b), 32 g as a buff coloured solid (yield
1
o
1
7
4%). Mp: 80ꢀ82 C. Analysis of analytical sample: H NMR (500MHz, DMSOꢀd ):
δ
7.41ꢀ
6
7
.38 (m, 5H), 5.06ꢀ5.04 (d, 1H, J = 8 Hz), 4.91ꢀ4.89 (d, 1H, J =8 Hz), 4.38ꢀ4.37 (d, 1H, J =4
Hz), 3.37ꢀ3.36 (t, 1H, J = 4 Hz), 3.30ꢀ3.27 (d,1H, J =12 Hz), 3.16ꢀ3.16 (m, 1H),2.31ꢀ2.27 (m,
1
H), 2.13ꢀ2.12 (m, 1H), 1.89ꢀ1.81 (m, 2H). Mass: (M+1): 258 for C H N O . HPLC purity:
14 15 3 2
1
00%.
Step 3: Synthesis of (2S,5R)ꢀ6ꢀhydroxyꢀ7ꢀoxoꢀ1,6ꢀdiazaꢀbicyclo[3.2.1]octaneꢀ2ꢀ
carbonitrile (3c): A solution of (2S,5R)ꢀ6ꢀ(benzyloxy)ꢀ7ꢀoxoꢀ1,6ꢀdiazaꢀbicyclo[3.2.1]octaneꢀ2ꢀ
carbonitrile 3b (32 g, 0.124 mol) in a mixture of DMF: DCM (1:1, 160 ml:160 ml) containing
10% Pd/C (4.6 g, 50% wet) was hydrogenated at 50ꢀ55 psi, for 2 h, at 25°C. The resulting
mixture was filtered through a celite pad and residue was washed with DMF: DCM (1:1, 25 ml:
2
5 ml). The solvent from the combined filtrate was evaporated under reduced pressure to obtain
the product as oil, which was used as such for the next reaction without further purification.
(20.66 g; yield ca ~100%). Small quantity was purified by column chromatography over silica
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
gel (60ꢀ120 mesh) and using 30ꢀ35% v/v mixture of acetone: hexane as an eluent. Evaporation of
the solvent from the combined fractions gave the product (3c) as a pale yellow solid. Mp: 135ꢀ
o
1
140 C. (Dec) Analysis of analytical sample: H NMR (500 MHz, DMSOꢀd ): δ 9.97 (s, 1H),
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
.52ꢀ4.50 (d, 1H, J =8 Hz), 3.69 (s, 1H), 3.19 (s, 2H), 2.01ꢀ2.00 (m, 2H), 1.86ꢀ1.83 (m, 2H).
Mass: (Mꢀ1): 166.1 for C H N O . HPLC purity: 99.7%. Step 4: Synthesis of (2S, 5R)ꢀ6ꢀ
7
9
3
2
(sulfooxy)ꢀ7ꢀoxoꢀ1,6ꢀdiazaꢀbicyclo[3.2.1]octaneꢀ2ꢀcarbonitrile, tetrabutylammonium salt (3d):
o
To a stirred and cooled (5 ꢀ 10 C) solution of (2S,5R)ꢀ6ꢀhydroxyꢀ7ꢀoxoꢀ1,6ꢀdiazaꢀ
bicyclo[3.2.1]octaneꢀ2ꢀcarbonitrile 3c (20.66 g, 0.124 mol) in DMF (160ml) DMFꢀSO complex
3
(22.8 g, 0.149 mol) was added in one portion and stirring continued further. After 1h of stirring,
to the resulting reaction mass was added slowly, a solution of tetrabutylammonium acetate (48.6
g, 0.161 mol) in water (160 ml). After 1 h of further stirring, the solvent from the reaction
mixture was evaporated under reduced pressure to obtain an oily residue. The oily mass was coꢀ
evaporated with xylene (2 x 200 ml), to obtain a thick mass. This mass was partitioned between
dichloromethane (320 ml) and water (320 ml). The organic layer was separated and the aqueous
layer reꢀextracted with dichloromethane (160 ml). The combined organic extracts were washed
with water (3 x 160 ml), dried over anhydrous Na SO and the solvent evaporated under reduced
2
4
pressure at 35 °C. The residual oily mass was triturated with ether (3 x 160 ml), each time the
ether layer was decanted and finally the residue was dried under reduced pressure, to obtain the
1
product (3d) as pale yellow oil, 52.5 g (86% yield). Analysis of analytical sample: H NMR
(
400 MHz, CDCl ): δ 4.42 (s, 1H), 4.34 (d, 1H, J = 6.8 Hz), 3.45 (d, 1H, J = 12.4 Hz), 3.36 (d,
3
1
H, J =12.4 Hz), 3.31ꢀ3.23 (m, 8H), ), 2.35ꢀ2.20 (2H, m), 1.94ꢀ1.81 (2H, m), 1.69ꢀ1.56 (8H, m),
1.48ꢀ1.39 (8H, m), 1.02ꢀ0.98 (12H, m). Mass: (Mꢀ1): 246 as a free sulfonic acid for C H N O S.
7
8
3
5
N(C H ) HPLC purity: 95.24%. Step 5: Synthesis of (2S, 5R)ꢀ1,6ꢀdiazaꢀbicyclo[3.2.1]octaneꢀ
4
9 4.
ACS Paragon Plus Environment

Page 27 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2
ꢀcarbonitrileꢀ7ꢀoxoꢀ6ꢀ(sulfooxy)ꢀ mono sodium salt (3): A solution of (2S, 5R)ꢀ6ꢀ(sulfooxy)ꢀ7ꢀ
oxoꢀ1,6ꢀdiazaꢀbicyclo[3.2.1]octaneꢀ2ꢀcarbonitrile, tetrabutylammonium salt 3d (51.5 g, 0.105
mol) in THF (50 ml) was loaded on column packed with Indion 225 Na resin (1200 gm) and
was eluted by using 10% THF in water. The pure fractions were collected and THF was
evaporated under reduced pressure. The aqueous solution was extracted with ethyl acetate (5 x
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
50 ml), treated with charcoal to decolourise and filtered. The filtrate was lyophilised, and the
powder obtained was recrystallized using isopropanol: water, to obtain the product (3), as a white
o
1
solid, 20.5 g (72% yield). Mp: 198ꢀ201 C. (Dec) Analysis of analytical sample: H NMR
(400MHz, DMSOꢀd6):
δ
4.56ꢀ4.54 (d, 1H, J =6.8 Hz), 4.09 (s, 1H), 3.31ꢀ3.22 (s, 2H), 1.98ꢀ1.86
1
3
(m, 4H). C NMR (400MHz, DMSOꢀd ): 163.5 (C=O), 119.1 (CN), 57.3 (CH), 48.6 (CH),
δ
6
ꢀ1
4
8.0 (CH ), 22.0 (CH ), 20.2 (CH ). IR (cm ): 3543, 2249, 1751, 1292, 1072, 1020, 1009. Mass:
2 2 2
(Mꢀ1): 246 as a free sulfonic acid for C H N O SNa. HPLC purity: 99.6 %. Water content
7
8
3
5
4
.18%. Analysis calculated for 3, C H N O SNa, C, 31.23; H, 3.0; N, 15.61. Found: C, 30.94; H,
7
8
3
5
25
3
.4; N, 15.45. Specific rotation: [α] D: ꢀ28.7° (c 0.5, water)
Plasmids and Strains. The cloning and/or origins for blaKPCꢀ2, blaPDCꢀ3, and blaADCꢀ7 genes for
susceptibility testing and blaKPCꢀ2, blaPDCꢀ3, blaADCꢀ7, blaOXAꢀ23, and blaOXAꢀ24/40 genes (missing
the nucleotides encoding their signal peptides) for protein expression were described in the
1
1, 20, 35ꢀ37
following references
.
The methods for the generation of the A. baumannii
38
ATCC17978 knockout strains is previously described .
For MIC analysis, the blaOXAꢀ48 gene was cloned into the pBC SK (ꢀ) vector by
amplifying the blaOXAꢀ48 coding and upstream promoter regions from K. pneumoniae CAV1543
1
9
containing blaOXAꢀ48, sequence verified, and transformed into the E. coli DH10B . For large
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
scale protein expression, the blaOXAꢀ48 gene without its leader peptide sequence was cloned into
the pET24a(+) vector using NdeI and XhoI restriction sites and electroporated into E. coli
DH10B. The resulting construct was sequence verified and transformed into E. coli BL21(DE3)
cells for protein expression and purification.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
For MIC testing, as OXAꢀ23 and OXAꢀ24/40 do not express well in an E. coli
background, presumably due to codon usage, another expression strategy was utilized. For
blaOXAꢀ23, the bla gene and its upstream ISAba1 was amplified from strain AB0057, and for
blaOXAꢀ24/40 the bla gene and its upstream XerC/XerD was amplified from strain A. baumannii
NM55. The amplified genes were cloned into the Xba1 site of the modified pWH1266 vector (a
vector that replicates in A. baumannii) in which ampicillin resistance was eliminated (ꢂblaTEMꢀ1
)
by inverse PCR, and an Xba1 site was engineered into the vector. Once the proper sequence was
confirmed in the constructs, the clone was transformed into A. baumannii OM2 clinical isolate.
To reduce the ampicillin MIC of the OM2 strain, the native blaADC gene was knocked out. A 3
kb fragment including the blaADC gene and about 1 kb flanking sequence upstream and
downstream of blaADC was amplified out of OM2 and cloned into pCRꢀXLꢀTOPO. The blaADC
gene in the pCRꢀXLꢀTOPO clone was removed by inverse PCR, and a tobramycin resistance
gene was put in its place. This plasmid was used to generate the blaADC knockout. Linearized
blaADC knockout plasmid was electroporated into OM2 cells with selection on tobramycin.
Colonies that grew on tobramycin were PCR screened for the tobramycin resistance gene to
assure that they were not breakthrough colonies. PCR using a primer upstream of the area used
to make the blaADC construct and a second primer residing in the tobramycin resistance gene was
performed to show integration into the desired location. Finally, PCR to amplify the blaADC gene
was negative, demonstrating that the blaADC gene was successfully knocked out. Note: The
ACS Paragon Plus Environment

Page 29 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
blaADC knockout of OM2 still contains blaOXAꢀ51. Numerous attempts to knock out the blaOXAꢀ51
gene were unsuccessful. It is unknown if other βꢀlactamases are present in OM2.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Protein Purification for Kinetic Assays and Mass Spectrometry. The purification of KPCꢀ2,
11, 20, 36, 37
ADCꢀ7, PDCꢀ3, and OXAꢀ23 was previously described
were purified according to the following protocols.
. OXAꢀ24/40,and OXAꢀ48
The OXAꢀ24/40 βꢀlactamase was purified as follows. E. coli BL21(DE3) containing
blaOXAꢀ24/40 pET24 (+) construct was grown in SOB containing 50 ꢁg/ml kanamycin at 37 °C
shaking to achieve an OD600 of 0.8. IPTG was added to the culture to a final concentration of 0.2
mM and the culture was grown for three more hours. The cells were centrifuged and frozen at ꢀ
2
0 °C. Pellets were thawed and suspended in 50 mM TrisꢀHCl, pH 7.4 with lysozyme (40
µg/ml), benzonuclease, and 1 mM MgSO . Cellular debris was removed by centrifugation, and
4
the lysate was subjected to preparative isoelectric focusing (pIEF) overnight. Location of OXAꢀ
2
4/40 on the pIEF gel was determined using a nitrocefin (NCF; Becton, Dickinson and
Company) overlay. OXAꢀ24/40 further purified by SEC using a HiLoad 16/60 Superdex 75
column (GE Healthcare Life Science). The OXAꢀ48 βꢀlactamase was purified from E. coli
BL21(DE3) containing blaOXAꢀ48/ pET24 (+) construct as described above for OXAꢀ24/40.
SteadyꢀState Inhibitor Kinetics. To determine the inhibitory potential of compounds 1, 2, and 3
against select class A, C, and D βꢀlactamases, steadyꢀstate inhibition kinetics were conducted
using an Agilent 8453 Diode Array spectrophotometer. Class A and C βꢀlactamases were tested
in 10 mM phosphateꢀbuffered saline, pH 7.4, while 50 mM sodium phosphate buffer, pH 7.2
ACS Paragon Plus Environment