High-Throughput Crystallography Reveals Boron-Containing Inhibitors of a Penicillin-Binding Protein with Di- And Tricovalent Binding Modes

Hector Newman, Alen Krajnc, Dom Bellini, Charles J. Eyermann,* Grant A. Boyle, Neil G. Paterson,

Article

High-Throughput Crystallography Reveals Boron-Containing

Inhibitors of a Penicillin-Binding Protein with Di- and Tricovalent

Binding Modes

Katherine E. McAuley, Robert Lesniak, Mukesh Gangar, Frank von Delft, Jurgen Brem, Kelly Chibale,

Christopher J. Scho ﬁ eld,* and Christopher G. Dowson*

Cite This: https://doi.org/10.1021/acs.jmedchem.1c00717

Read Online

ACCESS

Metrics & More

Article Recommendations

sı Supporting Information

ABSTRACT: The eﬀectiveness of β-lactam antibiotics is increas-

ingly compromised by β-lactamases. Boron-containing inhibitors

are potent serine-β-lactamase inhibitors, but the interactions of

boron-based compounds with the penicillin-binding protein (PBP)

β-lactam targets have not been extensively studied. We used high-

throughput X-ray crystallography to explore reactions of a boron-

containing fragment set with the Pseudomonas aeruginosa PBP3

(

PaPBP3). Multiple crystal structures reveal that boronic acids

react with PBPs to give tricovalently linked complexes bonded to

Ser294, Ser349, and Lys484 of PaPBP3; benzoxaboroles react with

PaPBP3 via reaction with two nucleophilic serines (Ser294 and

Ser349) to give dicovalently linked complexes; and vaborbactam

reacts to give a monocovalently linked complex. Modiﬁcations of

the benzoxaborole scaﬀold resulted in a moderately potent inhibition of PaPBP3, though no antibacterial activity was observed.

Overall, the results further evidence the potential for the development of new classes of boron-based antibiotics, which are not

compromised by β-lactamase-driven resistance.

INTRODUCTION

while those employing a zinc ion-mediated mechanism are

Ambler class B BLs (MBLs).

The ability of penicillins to treat infections is enhanced by

■

β-Lactam antibacterials, that is, penicillins, carbapenems,

monobactams, and cephalosporins, target the penicillin-bind-

ing protein (PBP) family of transpeptidases. In Gram-negative

bacteria, inhibition of high-molecular mass (HMM) class A

and class B PBPs (PBP1a/b, PBP2, and PBP3) is typically

their combination with a class A SBL inhibitor (e.g.,

6,7,9

clavulanate);

however, the rise of Class A (i.e., K.

pneumoniae carbapenemases) and class D BLs has compro-

mised this approach. Non-β-lactam-based carbapenemase

inhibitors, that is, diazabicyclooctanes (DBOs), for example,

lethal. The class B PBP3 is a monofunctional peptidoglycan

transpeptidase, which is associated with cell division, where it

cross-links stem peptides of polymerized molecules of lipid II

to create the peptidoglycan mesh essential for bacterial

,7,10

avibactam, which inhibits class A, C, and some D BLs;

relebactam, which inhibits class A and C BLs,

(

and

show utility

avibactam and relebactam) and promise (other DBOs) in

11−13

survival. β-Lactams inhibit PBPs, initially by competing

with the D-Ala-D-Ala terminus of the stem peptide substrate to

give a non-covalent complex, which then reacts with the active-

site catalytic serine [in Pseudomonas aeruginosa PBP3

restoring the clinical eﬃcacy of β-lactams against some

resistant strains though they are not potent MBL inhibitors.

There is thus a need for new antibiotics to treat a wider

spectrum of resistant Gram-negative infections. The develop-

ment of antibiotics with novel modes of action avoiding

(

PaPBP3): Ser294] to give an acyl−enzyme complex which

is stable over a biologically relevant timescale.

The eﬀectiveness of β-lactams to treat Gram-negative

infections caused by Escherichia coli, Klebsiella pneumoniae,

Enterobacter cloacae, Acinetobacter baumannii, and P. aeruginosa

is increasingly compromised by serine- and/or metallo-β-

Received: April 20, 2021

lactamases (SBLs and MBLs, respectively), with >1800 BL

variants identiﬁed. Ambler classes A, C, and D BLs are SBLs,

XXXX The Authors. Published by

American Chemical Society

J. Med. Chem. XXXX, XXX, XXX−XXX

Article

Figure 1. BCIs are proposed to act as mimics of “tetrahedral” transition states arising from enzyme-catalyzed hydrolysis pathways of natural

substrates and β-lactams with PBPs. Outline of general mechanisms of (a) transpeptidase reactions catalyzed by a class B PBP, for example, PaPBP3

and (b) the reaction of β-lactam with PBP3, exempliﬁed with a penicillin. Hydrolysis of the acyl−enzyme complex is typically slow in PBPs but is

rapid in serine-BL (SBL) catalysis. (c) sp form of BCIs may mimic the proposed “tetrahedral” transition states and thus bind tightly to the PBP

active site. The ability of boron to “morph” between sp and sp hybridization states is important in the context of SBL and likely MBL inhibition.

existing resistance mechanisms is time-consuming. An

alternative is to identify new chemo types for validated targets

and to incorporate new features that limit resistance

mechanisms to existing antibiotics.

that bicyclic boron-based PBP inhibitors might be analogously

developed.

In addition to acting as “transition-state” analogues of BLs,

crystallography reveals the potential for BCIs to undergo

unexpected reactions, for example, formation of a tricyclic

Following the pioneering clinical development of boronic

structure with the NDM-1 MBL and of a tricovalent binding

acids as protease inhibitors for multiple myeloma treatment,

mode with D,D-peptidases. We are interested in exploring the

reactions of boron-based compounds with PBP inhibitors, with

a view of enabling them as non-β-lactam-based inhibitors that

are not susceptible to BLs.

interest in boron-containing inhibitors (BCIs) of nucleophilic

enzymes has been growing. The ability of boronates to inhibit

SBLs has long been known. However, it is only recently that

monocyclic (i.e., vaborbactam, which inhibits class A and C

6−18

Here, we report the use of high-throughput protein

BLs

) and bicyclic boronates (e.g., taniborbactam and

crystallography to investigate the binding modes of a

fragment library enriched with boron-based compounds to

PaPBP3. The extensive structural results reveal that diﬀerent

types of potential boron-based inhibitors react diﬀerently with

PBPs, in particular boronic acids react to form tricovalent

complexes, while benzoxaboroles form dicovalent complexes.

structurally related bicyclic boronates, which inhibit members

19−21

of all four Ambler classes of BLs including MBLs

) suitable

for clinical use/development have been reported. These

compounds bind in a manner that mimics the proposed

,22

“tetrahedral” transition states in SBL and MBL BL catalysis

(Figure 1b). By contrast with the work on SBLs and MBLs,

there are relatively limited reports on the interactions of boron-

RESULTS

22−33

■

based inhibitors with PBPs.

Although there are reports of

X-ray Fragment Screen. Our previous attempt at high-

throughput (XChem) fragment screening with crystals of

PaPBP3 with a diverse library of >1300 fragments yielded only

a single, covalently reacted hit, a much lower hit rate than

3,24,26−31,34−36

boronic acids reacting with PBPs,

and with

antibacterial activity, there are no reports of potent bicyclic

boron-based PBP inhibitors/antibacterials in the peer-reviewed

literature. The development of dual-action PBP/BL boron-

based inhibitors is also of interest. The bicyclic BCI scaﬀolds

are of particular interest in part because their rigid nature

compared to acyclic inhibitors may make it easier to develop

selectivity for bacterial PBP and BL targets over nucleophilic

human enzymes. Work with the DBO scaﬀold, which was

originally developed for SBL inhibition but which was

typically expected for such screens. We therefore elected to

focus on a covalent fragment library enriched with boron-based

compounds, given their demonstrated reactivity toward serine

nucleophiles. In total, 262 compounds (most from Enamine’s

“Serine focused Covalent Fragments” library ), of which 152

were boron-containing compounds, were tested. For compar-

ison, other electrophilic compounds including epoxides and

sulfonyl ﬂuorides and vaborbactam

37−39

16−18

subsequently developed for antibacterial use,

suggests

(a monocyclic

J. Med. Chem. XXXX, XXX, XXX−XXX

side chains (contoured at 1 σ ), as calculated by “comit ” in the ccp4 suite.

Article

Figure 2. Boronates 1 and 2 react with PaPBP3 in a tricovalent manner. Boronates (a) 1 and (b) 2 react with Ser294, Ser349, and Lys484 (PDB:

ATM and 7ATO respectively). A dimethyl sulfoxide molecule at the active site in both structures is not shown for clarity (see Figure S10).

Hydrogen bonds are shown as black dashed lines. Unbiased omit Fo-Fc maps are shown (light blue mesh) for the ligand and covalently attached

Figure 3. Benzoxaboroles react with PaPBP3 in a dicovalent manner. (a) 3 and (b) 4 react covalently with Ser294 and Ser349 and hydrogen bond

with Lys484 (PDB: 7ATW and 7ATX, respectively). Hydrogen bonds are shown as black dashed lines. Unbiased omit Fo-Fc maps (light blue

mesh) are shown for the ligand and covalently attached residue side chains (contoured at 1σ), as calculated by “comit” in the ccp4 suite.

boronate SBL inhibitor, which reacts in a monocovalent

presence of a low pK lysine residue (generally the equivalent

residue to Lys297) in their active sites.

29,46−48

manner ) were also included. The library was screened

against PaPBP3 crystals (Figure S4 ) at both pH 6 and pH 8

because BCIs can interact diﬀerently with proteins depending

The phenyl rings of 1 and 2 occupy similar, but distinct,

regions at the active site. Water occupies the oxyanion hole

that is occupied by the β -lactam-derived carbonyl of PBP

inhibitors (Figure 4 and Figure S5). The tetrazole of 1 makes

hydrogen-bonds to the backbone NHs of Gly534 and Gly535

and is positioned to hydrogen bond with the Thr487 and

Ser485 side chains (Figure 2). The imidazole of 2 also forms a

hydrogen bond with the backbone NH of Gly535 and the

Ser485 side chain. The direct hydrogen-bonds to Gly534 and

Gly535 are unusual for PBP3 inhibitors that occupy the

carboxylic acid binding pocket, that is, the pocket binding the

C-3 penicillin carboxylate (Figure S5). These structures

therefore provide support for the future design of PBP3

inhibitors incorporating either weakly acidic non-carboxylates

on the pH. Soaking of crystals, harvesting, and data

collection were completed within 24 h. Thirty-four boron-

containing fragments (boronates and benzoxaboroles) and

vaborbactam were determined to be “hits” with various levels

of electron density observed at the active site. The fragments

showed a clear pattern of either tricovalent or dicovalent

bonding depending on whether the compound was based on a

boronate or a benzoxaborole scaﬀold.

Structures of boronates 1 and 2 with PaPBP3 show the

boron atom is sp -hybridized and tricovalently bonded to

Ser294 (the catalytic serine), Ser349 (from the conserved SxN

motif), and Lys484 (in the KS(T)G motif). The reaction of

or neutral groups that interact with the “acid binding pocket”

of PBP3, which is typically occupied by the C-3 carboxylate of

β-lactams (penicillin nomenclature) (Figure S5 ).

Lys484 is notable and may in part reﬂect a low pK for this

residue. Previous studies with PBPs have proposed the

J. Med. Chem. XXXX, XXX, XXX−XXX

Article

Figure 4. Binding mode of PaPBP3 inhibited by piperacillin compared with those for benzoxaboroles 7 and 11. (a) PaPBP3 inhibited by

piperacillin (PDB: 6R3X); (b) predicted binding mode of 11 as determined by docking, showing how groups 1 and 2 may engage in the same

manner as analogous groups in piperacillin; (c) binding mode of 7 complexed with PaPBP3 (PDB: 7AU0). Hydrogen bonds are shown as black

dashed lines. An unbiased omit Fo-Fc map is shown (light blue mesh) of the ligand and covalently attached residue side chains (contoured at 1σ),

as calculated by “comit” in the ccp4 suite. (d) Structure of piperacillin colored according to its functional groups: β-lactam (orange), C-3

carboxylate (pink), group 1 (blue), and group 2 (green); (e) benzoxaborole 11 was designed to mimic the piperacillin binding mode. Colors match

analogous groups within piperacillin which were hoped would engage the same parts of the protein in the case of 11.

Scheme 1. Synthesis of Benzoxaborole Derivatives (A) 5−10 and 14, (B) 11, and (C) 15

Reagents and conditions: (a) 1,1′-carbonyldiimidazole, N,N-dimethylformamide (DMF), 40 °C, 4−16 h; (b) LiOH·H O, 1,4-dioxane/H O

(

3:1), 40 °C, 60 min; (c) HCl (4 M soln. in 1,4-dioxane), CH Cl , rt,16 h; (d) Br , CH Cl , rt, 18 h; (e) benzyl bromide, K CO , DMF, rt, 4 h; (f)

2 2 2 2 2 2 3

bis(pinacolato)diboron, Pd(dppf)Cl , KOAc, dioxane, 80 °C, 12 h; (g) EtOAc, LDA (1 M solution in THF), −78 to −20 °C, 2 h; (h) LiOH·H O,

THF/H O (1:1), rt, 2 h. Note that low isolation yields for substituted benzoxaboroles (e.g., 8−10) in part reﬂect signiﬁcant losses during

puriﬁcation on the silica gel and provide scope for further optimization. For B1, R = H and R = CH(R )CO Me. The complete structures of 5−7,

14, and 8−10 are shown in Table 1. Dppf: 1,1′-bis (diphenylphosphino)ferrocene; Boc, tert-butoxycarbonyl; Bn, benzyl.

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

A similar tricovalent binding mode is reported for an alkyl

Table 1. Residual Activities and K s of Compounds Tested

boronate bound to a D,D-peptidase from Actinomadura sp. R39

in the BOCILLIN FL Competition Assay for Binding to

PaPBP3 and Residual Activities Measured in S2d Turnover

(

PDB: 3ZVT) though our work represents the ﬁrst time

these binding modes have been observed with a HMM PBP.

Notably, the boron atoms of both structures are positioned

similarly within the active site and the side chains of both

nucleophilic serines and lysine align closely, independent of the

nature of the boron-bonded functional group, that is, alkyl

versus phenyl.

Assays

The screen also revealed that benzoxaboroles bind to the

PBP active site, but with the boron dicovalently bound to

Ser349 and Ser294 and a hydrogen bond via its “endocyclic”

oxygen to Lys484. Both 3 and 4 form hydrogen bonds to the

side chain NH group of Asn351. Like the phenyl boronates, a

water molecule occupies the oxyanion hole and forms

hydrogen bonds to the backbone NH of Ser294 and Thr487.

In the structures for 3 and 4, Tyr409 is hydrogen-bonded to

the backbone carbonyl of Thr487 (Figure 3).

Hit Expansion Design and Synthesis. Due to the

chemically interesting nature of their reaction with the PBP

active site, we investigated the potential of benzoxaboroles for

PBP inhibition. The PaPBP3/3 complex structure, along with

50,51

reported structural studies on β-lactams,

was used to

design a set of 6-substituted amides incorporating the features

of the piperacillin C-6 side chain, that is, D-phenylglycine and

diketopiperazine moieties; for synthetic simplicity, a ketopiper-

azine was used to mimic the diketopiperazine. Ideas were

evaluated using docking into the PaPBP3 piperacillin binding

site (Figure 4).

Benzoxaboroles 5-11 with a C-6 acylamino side chain were

synthesized (Scheme 1) and structures of their PaPBP3

complexes solved. As demonstrated by comparison of Figure

c,e, despite their covalent reaction with Ser294 and Ser349,

−11 generally failed to engage the active site as proposed and

did not show inhibition (Table 1). Analogues 12, 13 (from

Wuxi AppTec), and 15 (Scheme 1C) with a C-3 carboxylic

acid group were then synthesized. The C-3 carboxylate group

was modeled to align with the analogous C-3 carboxylate

present in penicillins (pink in Figure 4), with the aim of

improving the aﬃnity.

To explore alternatives to amino acid-derived benzoxabor-

oles, 14 and 15 were synthesized (Scheme 1). Once again, a

dicovalent reaction of the boron with both Ser294 and Ser349

was observed. The phenyl groups within the side chains of

both these compounds are situated in a region close to where

the reacted piperacillin phenyl binds (PDB: 6R3X) (Figure

S8 ).

For both BOCILLIN FL FA and S2d assays, residual activities in the

presence of a 1 mM inhibitor are a percentage of the activity of the

Inhibition Assays. 1−15 were screened for PaPBP3

untreated control. Errors are standard errors (n = 3) from

inhibition at 1 mM using an established ﬂuorescence

independent measurements. Residual activity measured at 100 μM

anisotropy (FA) assay. For selected compounds, inhibition

due to solubility issues at 1 mM. K , calculated with 11 concentrations

,24,53,54

of hydrolysis of the thioester substrate analogue S2d

of each compound and determined using global ﬁtting in Kintek

52,59

was measured, with good correlation between the results with

the two assays (Table 1). Most compounds manifested

minimal inhibition despite clear evidence of binding in

crystallo. Pre-incubation of 12 with PaPBP3 (0, 30, or 60

min) prior to assay initiation had no signiﬁcant eﬀect on

Global Explorer.

ND, not determined.

Given the sequence (Figure S6) and structural (Figure S7)

similarities of the studied PBP3 variants, the apparent variation

in their activity is interesting, possibly indicating a degree of

selectivity.

inhibition (data not shown). K values were determined using

the BOCILLIN FL assay for the four compounds with the

lowest residual activities (12, 13, 15, and vaborbactam), of

To investigate the reversibility of the reaction, we used the

colorimetric substrate nitroceﬁn, which acylates PaPBP3 and

then rapidly deacylates, with a concurrent color change.

which 12 was the most potent with a K of 73.9 ± 0.8 μM

22,57,58

(Table 1). 12 was also tested against puriﬁed PBP3s from E.

coli, A. baumannii, and Haemophilus inﬂuenzae, and PBP2 from

Upon twofold dilution of the assay, the half−maximal

55,56

Neisseria gonorrhoeae (equivalent to PBP3)

(Table 2).

inhibitory concentration (IC ) value relative to the unin-

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

BL-expressing strains (data not shown).

Article

Table 2. Inhibition Properties of Benzoxaborole 12 against

Various PBP variants

Consistent with this is the observation that the progress curves

can be ﬁ t well using a simple reversible, one-step binding

model (Figure S2). While the inhibition is weak, these results

are consistent with other investigations of boronate binding in

HMM PBPs, which typically show IC s in the μM

protein tested

residual activity by BOCILLIN FL FA (%)

PBP3 from P. aeruginosa

PBP3 from H. Inﬂuenzae

PBP3 from A. baumannii

PBP3 from E. coli

16 ± 6

32 ± 3

73 ± 3

>90

23−25,32,33

range.

Selected compounds were screened against E. coli, P.

aeruginosa, H. inﬂuenzae, A. baumannii, and N. gonorrhoeae

and a P. aeruginosa strain engineered to remove the outer

membrane permeability barrier to investigate their antimicro-

PBP2 from N. gonorrhoeae

>90

Residual activities (treatment with 12 (1 mM) for 1 h, the activity

determined by BOCILLIN FL competition assay) are given as a

percentage of the untreated control, with errors as standard deviations

of three technical replicates. Note that the N. gonorrhoeae PBP2 was a

transpeptidase-only construct.

bial activity (Table S2 ). All were ineﬀective (MICs ≥64 μg/

mL). Synergy with piperacillin, a Gram-negative PBP3 speciﬁc

β-lactam, was not observed in non-BL-expressing E. coli and

P. aeruginosa, although weak synergy was observed in class C

hibited control was doubled, indicative of rapid (equilibrium

established in <30 s) reversibility. The IC₅₀of irreversibly

bound ceftazidime was not signiﬁcantly aﬀected (Figure S3 ).

Figure 5. Structures of benzoxaboroles with a C-3 acid group (12, 13, and 15) and 14 complexed with PaPBP3. (a) PaPBP3/12 complex (PDB:

AU1); (b) PaPBP3/13 complex (PDB: 7AU8); (c) PaPBP3/14 (PDB: 7AU9); and (d) PaPBP3/15 complex (PDB: 7AUB). For clarity, only one

of the two reﬁned conformations of Tyr409 is shown in (a). Hydrogen bonds: dashed black lines. Unbiased omit Fo-Fc maps are shown (light blue

mesh) of the ligand and covalently attached residue side chains (contoured at 1σ), as calculated by “comit” in the CCP4 suite.

J. Med. Chem. XXXX, XXX, XXX−XXX

Article

Figure 6. Boron-containing compounds react with a PBP in three distinct modes. The formation of diﬀerent complexes with exemplary

crystallographically observed complexes are shown. (a) Vaborbactam reacts monocovalently with Ser294; (b) Benzoxaboroles (3−15) bind

dicovalently with Ser294 and Ser349; (c) phenylboronic acids (e.g., 1, 2, and alkyl boronic acids ) bind tricovalently to Ser294, Ser349, and

Lys484. The order of nucleophilic residue reactions is unknown; (d) overlay of crystallographically observed states. The position of the boron can

be described by rotations of the chi angles of the Ser294 side chain. In the monocovalently and dicovalently bound crystal structures [and in the

piperacillin-reacted structure (PDB: 6R3X), the chi1 angle of Ser294 is gauche- (∼−60°) relative to the serine amine. In contrast, in tricovalent

mode (e.g., 1), the chi1 angle of Ser294 is trans (∼−161°) relative to the amine. There is an 80° diﬀerence between the chi2 angles of dicovalent

mode and monocovalent mode (Figure S9 ).

DISCUSSION

correlate with the crystallization conditions but is probably

related to the form of the warhead (Table S3).

Structures of PaPBP3 with 3, 4, 7, 9, 12, 13, 14, and 15

■

(

The 10 structures of boron compounds reacted with PaPBP3

Table S2) reveal three distinct binding modes: (i)

monocovalent reaction with Ser294 (Figure 6a), (ii) dicovalent

reaction with Ser294 and Ser349 (Figure 6b), and (iii)

tricovalent reaction with Ser294, Ser349, and Lys484 (Figure

show the benzoxaborole core binding in a conserved manner

(

Figure S8). In the dicovalent benzoxaborole binding mode,

the chi1 angle of the Ser294 side chain is the same as in the

monocovalently piperacillin-reacted structure, but the benzox-

aborole and piperacillin-reacted structures have a diﬀerence in

the chi2 angle of ∼80° (Figure S9). These observations reveal a

scope for variations in the Ser294 side chain conformation on

covalent inhibitor reaction. By contrast, in all our structures,

Ser349 is not signiﬁcantly displaced compared to the β-lactam-

c). The observed binding mode depends on the nature of the

boron compound, with vaborbactam reacting monocovalently,

benzoxaboroles (3−15) reacting dicovalently, and phenyl

boronates (1 and 2) reacting tricovalently. Tricovalent and

2,25−27,29,30

monocovalent

bonding of boron compounds with

reacted structures, for example, piperacillin-reacted PaPBP3

PBPs are known, but to our knowledge, this is the ﬁrst time

that benzoxaborole compounds have been shown to bind to a

PBP in a dicovalent manner. Although we cannot rule out the

possibility that some of our structures are a consequence of

crystallization, analogous data were collected at pH 6 and pH

(

PDB: 6R3X). Similarly, in the tricovalent binding mode of

, the Ser294 chi1 angle rotates by ∼100 °(relative to the

monocovalently reacted structure) (Figure S9), but the Ser349

and Lys484 side chains are not signiﬁcantly displaced (with O

and N being displaced <1 Å relative to their positions in the

piperacillin-reacted structure (PDB: 6R3X) ).

; the observed covalent binding mode thus does not appear to

J. Med. Chem. XXXX, XXX, XXX−XXX

Article

Figure 7. Binding of 2 causes changes in the PaPBP3 active-site conformation. Views of benzoxaborole PaPBP3/2 (orange, PDB: 7ATO) and

piperacillin-reacted PaPBP3 (black, PDB: 6R3X) (protein backbone colored blue and gray, respectively) are shown. A clear unprecedented

diﬀerence in the α10-β3 loop conformation (residues 466−473) is observed. See also Figure S14 .

It has been suggested that, at least in the context of BL

inhibition, binding to the catalytic serine occurs ﬁrst, followed

by (where appropriate) reaction with the other nucleophilic

contributing to the lack of inhibition shown for some

compounds (5−10). The position of the C-6 amide group

did enable hydrogen bonding with the highly conserved

Asn351, a hydrogen bond observed in many β-lactam-reacted

PBP complexes (Figure 4a).

Exploration of C-5-substituted benzoxaboroles employing

structural insights revealed here could lead to further potency

improvements. Thus, for example, while 14 did not inhibit

PaPBP3, similar to the other benzoxaboroles without an acid

group, one of its benzyl groups is positioned near the D-Phe

position of reacted piperacillin and the other is positioned

residues. In the structures where boron reacts in the

dicovalent manner (see, e.g., Figure 5), substitution of the

strong B−O bond in the ﬁve-membered ring of the

2,63

benzoxaborole

by the primary amine of Lys484 is likely

unfavorable, explaining why tricovalent binding does not occur.

Our kinetic data indicate that the equilibrium between PBP

and boronate is achieved at least within minutes, supporting a

relatively fast, but weak binding model for boronates to PBPs,

as suggested previously. Our method does not allow for

determination of binding rates on the seconds timescale, but

detailed kinetic analysis has been carried out to investigate the

observation of biphasic inhibition curves with some

close to key active-site residues. The structure with 15 (K =

78.1 ± 0.9 μM) places the phenyl of the C-5 benzylether

benzoxaborole substituent in the same PaPBP3 active-site

location as the phenyl rings in piperacillin- and amoxicillin-

derived adducts.

tricovalently binding BCIs. Zervosen et al. concluded that

the biphasic curves were caused by rapid non-covalent

association, followed by slow formation of monocovalent,

Importantly, structures of 1 and 2 demonstrate the potential

for non-acid groups to engage the PBP3 acid binding pocket

optimized binding of these groups may be used to combat

tetrahedral complexes. The formation of tricovalent com-

plexes was suggested to be in rapid equilibrium with the

monocovalent complex (Figure 6c). Boronates that react with

PBPs monocovalently are proposed to act as analogues of the

point mutation-mediated resistance in this region.

50,51,66−69

Examination of the reported PaPBP3 structures

reveals the β5-α11 loop (residues 528−539, Figure S10)

adopts diﬀerent conformations. The β-lactam-containing

28−30

tetrahedral transition state (Figure 1);

however, the tri-

and dicovalent binding modes are less obvious transition state

mimics. It remains unclear if the tricovalent binding mode will

be important for developing boron-based PBP inhibitors with

increased potency, or if it is, at least in part, a crystallographic

artifact.

One conclusion from the structural−activity relationships is

that the presence of a C-3 acid group increases potency.

PaPBP3 inhibitors (e.g., amoxicillin, aztreonam, ceftazidime,

50,51

and piperacillin

) induce formation of a “hydrophobic wall”

composed of Tyr533, Phe532, and residues within the β5-α11

loop together with Tyr503. Meropenem, which lacks a

hydrophobic C-6 group, by contrast, does not induce the

formation of this “hydrophobic wall” (PDB: 3PBR), while

apo PaPBP3 adopts a third conformation for the β5-α11 loop

Comparisons of the aﬃnities of 13 (K = 172.0 ± 3.0 μM) and

(

Figure S10, PDB: 6HZR).

0 (residual activity by BOCILLIN FL FA > 90%)

In most of our PaPBP3 boron-based compound structures

demonstrate the beneﬁt of a C-3 acid (Table 1). This result

is consistent with the conservation of the C-3 (or equivalent)

carboxylate group across most β-lactam PBP inhibitors and

studies showing addition of a carboxyl group to benzoxabor-

oles leads to a 10−100-fold increase in binding aﬃnity for

(

e.g., 12), the electron density maps in the region of the β5-

α11 loop were of insuﬃcient quality to enable residues to be

ﬁ tted. However, although the observed densities were weak

(

Figure S10), in the complexes with 1, 3, and 13, unique

conformations of the β5-α11 loop were observed. Benzox-

aboroles 9, 11, 12, 14, and 15 were designed to have a phenyl

group, binding of which may mimic that of the D-Phe of

reacted piperacillin in interactions with the hydrophobic wall

(Figure S11). While the phenyl group of 14 and 15 is

positioned close to the position of the D-Phe of reacted

piperacillin, the hydrophobic wall was not formed on binding,

perhaps in part reﬂecting the poor potency of these

compounds.

some BLs.

Our structure-based design of boron inhibitors aimed to

utilize analogous hydrogen bond interactions to those made by

reacted piperacillin (Figure 4). However, the constrained

position of the benzoxaborole ring (Figure S8), imposed by the

dicovalent bond formation and rigid 6−5 ring system, meant

key hydrogen bond interactions, particularly those to the

backbone of the β3 strand (Thr487 and Arg489), could not be

made by the C-6 amide group (blue in Figure 4d), likely

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

The conformation of the β3 β-strand is associated with

hydrophobic wall formation possibly mediated by inhibitor

interactions with Thr487 and Arg489 . The overall fold of

PaPBP3 bound to 12, which lacks the ketopiperazine terminal

substituent, is similar to that in the meropenem-reacted

PaPBP3 complex (Figure S12 , PDB: 3PBR). In both these

structures, Tyr409 forms a hydrogen-bond to the backbone

carbonyl of Thr487, whereas with amoxicillin, aztreonam,

ceftazidime, and piperacillin, the inhibitor-derived adducts

form a direct hydrogen-bond to the backbone carbonyl of

Thr487 (Figure S12). The rigid dicovalent binding mode of

the benzoxaborole core (Figure S8 ) prevents the C-6 amide

from directly forming hydrogen-bonds to the backbone

carbonyl of Thr48. A structure of PaPBP3 complexed with

Figure 8. Binding of vaborbactam to PaPBP3. Hydrogen bonds: black

dashed lines. An unbiased omit Fo-Fc map is shown (light blue mesh)

of the ligand and covalently attached residue side chains (contoured

3, which has a ketopiperazine substituent but which lacks a

phenyl analogous to that of the D-Phe of piperacillin, has β5-

α11 loop and β3 β-strand conformations similar to those

observed in the piperacillin-reacted PaPBP3 (Figure S13 ).

Interestingly, both the ketopiperazine of 13 and the

diketopiperazine of reacted piperacillin form hydrogen bonds

at 1 σ), as calculated by “comit” in the ccp4 suite. PDB code 6AUH.

dicovalent complexes (additionally reacting with Ser349),

and the boronic acids form tricovalent complexes (additionally

reacting with Lys484). While these binding modes are

with the O oxygens of Tyr328 and Tyr407 (Figure S13).

Further optimization of the side chains of the benzoxaboroles

may be able to better exploit the interactions possible with the

β3 strand backbone, Tyr328 and Tyr407, and the hydrophobic

wall regions, leading to improved binding.

27,70,71

precedented with other nucleophilic serine enzymes,

the generality of the reactions is notable, in particular for the

tricovalent reactions observed for the boronic acids. Given the

binding modes as reﬂected by crystallography (though not as

yet validated in solution) and that SBL/MBL inhibition by

boron-containing compounds has been demonstrated, it is

perhaps surprising that the compounds described here are not

Variations in the α10-β3 loop were observed in the PaPBP3/

complex (Figures 7 and S14 ), that is, it adopts a

conformation not previously observed in PaPBP3 structures.

The largest displacements, relative to the piperacillin-reacted

structure, are for residues Gly469, Gly470, and Val471 (7.2,

more potent PaPBP3 inhibitors though aﬃnities for boronates

0.2, and 9.9 Å, respectively). The signiﬁcance of these

22−25

against essential PBPs have typically been poor.

Recent

movements is unclear. The structure of PaPBP3/12 was

patent reports concerning compounds with a bicyclic boronate

^r^e^ﬁⁿ^e^d^wⁱ^t^h^t^w^o^c ^o ⁿ ^f ^o ^r ^m ^a ^t ⁱ ^o ⁿ ^s^f^o^r^V^a^l⁴76^-^P^r^o^-^G^l^y^-^T^y^r^-^Hⁱ^s480

of the α10-β3 loop (Figure S15). These structures appear to

(taniborbactam-like) scaﬀold show promising activity against

32,33

PBPs,

but a clear explanation for the diﬀerential potencies

nicely exemplify how, when combined with high-throughput

crystallography, the ability of boron-based compounds to

of boronate inhibition of BLs versus PBPs has yet to be

established. Our combined results imply that the rigid nature

interchange between diﬀerent forms (including sp and sp

of the di- and tricovalent complexes formed may require rather

precise derivatization of the boron scaﬀold to obtain potent

inhibition. Consistent with this, our attempt to combine the

benzoxaborole PBP3 structures with those of β-lactams (e.g.,

piperacillin) did not give potent inhibitors. An interesting

observation from the crystallography was of the inﬂuence of

the boron inhibitors on the conformation of the β5-α11 and

α10-β3 loops. The results presented here and elsewhere

suggest that conformational changes during active-site ligand

binding are important for inhibition and (likely) also induced

ﬁt substrate binding. The “morphing” ability of boron-

containing ligands, combined with high-throughput crystallog-

raphy to reveal otherwise latent structural changes, is notable.

However, it is presently diﬃcult to predict how speciﬁc active-

site interactions may inﬂuence conformational changes for the

active site. Indeed, the results suggest that the lack of potency

of the PBP inhibition of compounds described here and

elsewhere may relate to non-optimal interactions with mobile

regions in and around the PBP active site. Thus, from a

medicinal chemistry perspective, it may be most eﬃcient to

employ a combination of structure-guided (core scaﬀold) and

empirical (side chain) approaches.

forms) during their reactions with proteins can be employed to

reveal otherwise latent conformations.

Vaborbactam, an approved monocyclic SBL inhibitor

without reported antimicrobial activity, was included in our

studies to investigate its potential interactions with PaPBP3

compared to the other BCIs. Our PaPBP3/vaborbactam

structure shows a monocovalent reaction, as observed with

SBLs (Figure 8). The structures of vaborbactam and

amoxicillin in complex with PaPBP3 superimpose well,

including with respect to the oxyanion hole, the C-3

carboxylate, their amide bonds, and in the position of their

aromatic rings (Figure S16 ). Similar to amoxicillin, vaborbac-

tam is positioned to form hydrogen bonds with residues

including with Asn351, Tyr409, Thr487, and Ser485 as well as

well to engage with the hydrophobic wall (residues Tyr503,

Tyr532, and Phe533). Despite these favorable interactions,

vaborbactam binds poorly to PaPBP3 in solution (Table 1),

consistent with its lack of antimicrobial activity.

CONCLUSIONS

■

The multiple PBP3 structures reported here with three

diﬀerent classes of boron-based compounds, that is, boronic

acids, and predominantly mono- (vaborbactam) and bicyclic

Insights from our PaPBP3/benzoxaborole structures com-

bined with reported structures for piperacillin-, aztreonam- and

ceftazidime-reacted complexes provide a strong knowledge

base that will enable further explorations of benzoxaboroles as

non-β-lactam-based PBP inhibitors. The PaPBP3 structures

(

benzoxaboroles) boronates, reveal three distinct binding

modes. Vaborbactam forms a covalent bond reacting with

the catalytic serine, whereas the benzoxaboroles form

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

with 14 and 15 are interesting as they indicate vectors to target

key PBP active-site interaction hotspots for improving aﬃnity.

If benzoxaborole PaPBP3 potency can be improved, there is

patent literature precedent that boron-based PBP inhibitors

can achieve MICs of 8−16 μg/mL in CTX-M-15-expressing E.

done to ensure the correct parameters for BOCILLIN FL binding

were used. Progress curves were analyzed with KinTek, USA, as

52,69

described;

reported errors represent the standard error. The

acylation of a PBP by BOCILLIN FL (Figure S1 ) can be described by

a simple one-step model. However, unlike the previous studies, we

found it was necessary to include a term (k ) to account for the

2,33

coli strains.

Given the pressing need for novel antibiotics

deacylation of BOCILLIN FL to release a hydrolyzed product (Figure

S1). The ﬂuorescence intensity was constant throughout the reaction.

Inhibitor binding was modeled as a reversible reaction to form the

against resistant Gram-negative bacteria, further exploration of

benzoxaboroles and related PBP inhibitors is a promising

approach to new antibiotics targeting PBPs.

enzyme−inhibitor complex (Figure S1 ). K was calculated as the ratio

of the oﬀ-rate k to the on-rate k (Figure S1 ). Inhibitor binding

off

ﬁttings are shown in Figure S2 .

S2d Assays. Residual activities were measured by the ability of a

potential inhibitor to hinder hydrolysis of the substrate analogue S2d

EXPERIMENTAL SECTION

■

Protein Synthesis and Crystallography. Genes encoding for

recombinant PBP3s from P. aeruginosa, E.coli, A. baumannii, and H.

inﬂuenzae were expressed as soluble fragments to produce proteins

with PBP residues 50−579, 60−588, 64−609, and 80−610,

respectively, which lack the N-terminal transmembrane helix anchor.

A soluble construct expressing only the C-terminal transpeptidation

domain of N. gonorrhoeae PBP2 (residues 237−581) (a class B PBP

with close analogy to the PaPBP3 protein, with diﬀerent

(

2-((benzoyl-D-alanyl)thio)acetic acid) turnover by PaPBP3 as

4,24,53,54

described for other PBPs.

Assays were conducted in 50 μL

in a 384-well, clear-bottom, black-walled microplate (Greiner Bio-

One). PaPBP3 (400 nM) was incubated with 2 mM of each

compound for 1 h at 30 °C in 100 mM sodium phosphate, pH 7,

supplemented with 0.01% (v/v) triton. A solution of 5,5′-dithiobis-(2-

nitrobenzoic acid) and S2d diluted in the same buﬀer (to give a ﬁnal

concentration in the assay of 1 mM for both reagents) was added to

each well to initiate the reaction. The ﬁnal concentration of protein

was 200 nM and the ﬁnal inhibitor concentration was 1 mM. A

ClarioStar plate reader (BMG Labtech) was used to follow the

reaction by observing the change in absorbance at 412 nm at 30 °C.

The same assay was conducted in the absence of the inhibitor and

additionally with an excess of aztreonam (1 mM), which completely

inhibits PaPBP3; the results of this were used as a control to

determine the rate of spontaneous S2d hydrolysis in the absence of an

enzyme. The initial rate of S2d turnover was calculated using

GraphPad prism (Prism 8 for macOS, GraphPad Software LLC) and

the standard error calculated from three independent technical

replicates. The rate of non-enzymatic S2d hydrolysis was subtracted

from each rate to ﬁnd the corrected rate. For each compound, the

ratio of the corrected rate to the untreated control corrected rate was

expressed as a percentage to give the residual activity.

55,56

nomenclatures for historical reasons)

from the clinical muta-

tion-mediated penicillin-resistant N. gonorrhoeae (FA6140) was used.

All proteins were produced and puriﬁed as described using reverse

nickel aﬃnity chromatography utilizing the HRV 3C protease to

cleave the N-terminal His6 expression tags.

PaPBP3 crystals were obtained at 294 K using the hanging drop

vapor diﬀusion method by mixing equal volumes of 10 mg/mL

protein with the following precipitant solutions: 25% (w/v)

polyethylene glycol 3 350, 1% (w/v) protamine sulfate, and 0.1 M

Bis-Tris propane at pH 6 or 8. Crystals were cryoprotected with

0% (v/v) glycerol prior to ﬂash-freezing in liquid nitrogen.

Inhibitor−protein complexes were obtained by soaking crystals

overnight with a 250 mM solution of the requisite compound.

Diﬀraction data were collected on Diamond beamlines I03, I04, and

I04-1. All data were processed using autoPROC and STARANISO

due to signiﬁcant anisotropy of the data, with the exception of the

complexes of PaPBP3 with 7 and 15, which were isotropic and which

were processed with autoPROC. Structures were phased by

residual activity(%) = 100

Phaser_MR using a high-resolution structure of PaPBP3 (PDB:

[initial rate in the presence of inhibitor] − [non‐enzymatic rate]

HZR).₇Manual building and ligand ﬁtting were performed with

COOT. Reﬁnement was carried out primarily using REFMAC5

[initial rate of untreated control] − [non‐enzymatic rate]

within the CCP4 suite as well as in phenix.reﬁne. Structures were

Methods for conducting the nitroce ﬁ n assay and microbiology are

provided in the Supporting Information.

validated with MolProbity. Figures of structures were prepared

using PyMOL (The PyMOL Molecular Graphics System, Schro

Interference Considerations. The activity of all the compounds

1−15) belonging to the same class were analyzed by two orthogonal

dinger, LLC) or CCP4mg.

(

BOCILLIN FL Assays. Unless otherwise stated, assays were run in

triplicate using a 60 nM puriﬁed protein (see above), 30 nM

BOCILLIN FL in pH 7 100 mM sodium phosphate buﬀer, with

methods (FA and absorbance) with good correlation between the

results of both assays. Triton (0.01% v/v) was added to the buﬀer for

both assays, which should reduce interference by aggregators.

Moreover, a nitroceﬁn dilution assay indicates that the interaction

of 12 with PaPBP3 is reversible and well behaved (Figure S3 ). The

weak interaction of 12 with closely related PBPs from other bacteria

indicates that the reaction is selective (Table 2). Lastly, for boronic

acids (PaPBP3/1 and PaPBP3/2, Figure 2) and benzoxaboroles

.01% Triton to reduce promiscuous ligand binding and reduce

binding of the protein to the plate. Assays were run in triplicate in a

volume of 50 μL, in black, ﬂat-bottom, 384-well microplates (Grenier

Bio-One, Austria) at 30 °C. The change in FA was measured using a

ClarioStar plate reader (BMG Labtech) with polarized ﬁlters at

excitation: 482−16, emission: F: 530−4, and calculated using MARS

software v3.32 (BMG Labtech) using the equation

(PaPBP3 in complex with 3, 4, 7, 12, 13, 14, and 15, Figure S8 ),

multiple crystal structures demonstrate a conserved binding mode,

speciﬁcally engaging active-site residues. The interaction is consistent

with examples from the literature of nucleophilic residue engagement

by boron-based compounds (Figure S17 ).

1 −15 were also screened in silico (http://zinc15.docking.org/

patterns/home ) for predicted PAINS and aggregator functional

groups, but none were identiﬁed. During the design of the compounds

(1−15), we aimed to obtain polar compounds to limit the chance of

aggregation and avoided functional groups which can cause redox

activity, ﬂuorescence, protein reactivity, singlet-oxygen quenching,

and so forth. The purity of each compound was >95%, minimizing the

presence of impurities.

fluorescence anisotropy = (F_p_a_r_a− F_p_e_r_p)/(F_p_a_r_a+ 2F_p_e_r_p

)

where F_p_a_r_ais the ﬂuorescence intensity parallel to the excitation plane

and F_p_e_r_pis ﬂuorescence intensity perpendicular to the excitation

plane.

Residual activities were determined by pre-incubating the test

compound (1024 μM) and protein for 1 h at 30 °C before the

reaction was initiated by the addition of BOCILLIN FL. The change

in FA after 30 min was compared to the uninhibited control to

determine the residual activity. In order to calculate K , the compound

at 11 concentrations: 1−1024 μM) and BOCILLIN FL were mixed

and the reaction was initiated by the addition of PBP. Also included in

(

Synthetic Chemistry. All reagents were from Sigma-Aldrich,

Fisher Scientiﬁc, Combi-Blocks, Enamine, or Fluorochem and were

used without further puriﬁcation. 1 ((3-(1H-tetrazol-5-yl)phenyl)-

the model were progress curves of nine concentrations of PBP (40−

10 nM). Inhibition plots with meropenem and ceftazidime were

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

boronic acid) and 3 (1-hydroxy-1,3-dihydrobenzo[c][1,2]oxa-borole-

9.31 (br s, 1H, OH), 8.45 (d, J = 8.5 Hz, 1H, NH), 8.18−8.15 (m,

1H, Ar−H), 7.89 (dd, J = 8.0, 2.0 Hz, 1H, Ar−H), 7.59−7.52 (m, 1H,

-carboxylic acid) were from Combi-Blocks, Inc. 2 ((5-methyl-1H-

indazol-6-yl)boronic acid)) and 4 (4-(1-hydroxy-1,3-dihydrobenzo-

c][1,2]oxaborole-6-carbonyl)-1,3,3-trimethylpiperazin-2-one) were

Ar−H), 7.46 (d, J = 8.0 Hz, 1H, Ar−H), 7.33−7.30 (m, 2H, NH ),

[

7.24 (t, J = 7.5 Hz, 2H, Ar−H), 7.20−7.13 (m, 1H, Ar−H), 7.12−

from Enamine. These were used without further puriﬁcation in the

X-ray fragment screen. 3 was purchased from Combi-Blocks, Inc.; 12

and 13 were purchased from Wuxi Apptec; vaborbactam was

purchased from MedChemExpress. 2-((Benzoyl-D-alanyl)thio)acetic

acid (S2d) was synthesized as reported; the spectroscopic data were

7.08 (m, 1H, Ar−H), 5.02 (s, 2H, −CH OB), 4.66 (ddd, J = 10.5, 8.5,

4.0 Hz, 1H, −CHNH), 3.12 (dd, J = 14.0, 4.0 Hz, 1H, −CH Ph), 2.99

(dd, J = 14.0, 4.0 Hz, 1H, −CH Ph); C NMR (151 MHz, DMSO-

d ): δ 173.8, 166.9, 157.3, 139.0, 133.5, 130.3, 130.2, 129.6, 128.5,

126.7, 121.6, 70.4, 55.2, 37.7; LCMS (ESI , m/z), 325 [M + H] ;

HRMS (ESI−TOF) calculated for C H N O B [M + Na] ,

consistent with the ones previously reported.

Solvents were used as received. Flash column chromatography was

performed using a Teledyne ISCO ﬂash puriﬁcation system using a

Silicycle SiliaSep C18 cartridge. Purity of all ﬁnal derivatives for

biological testing was conﬁrmed to be >95% as determined using an

Agilent ultra-performance liquid chromatograph−mass spectrometer

347.1174; found, 347.1176.

Methyl (2R)-2-[(1-hydroxy-3H-2,1-benzoxaborole-6-

carbonyl)amino]-2-phenyl-Acetate (7). General Protocol 1 was

followed using the following quantities of reagents: 3 (100 mg, 0.56

mmol, 1 equiv); N,N-DMF (2 mL); 1,1-carbonyldiimidazole (182

mg, 1.12 mmol, 2 equiv); methyl (2R)-2-amino-2-phenyl-acetate (139

(

Agilent Technologies 6150 quadrupole, ES ionization) coupled with

an Agilent Technologies 1290 Inﬁnity II series UPLC system Agilent

290 series high-performance LC (HPLC) at two wavelengths of 254

and 280 nm using the following conditions: Kinetex 1.7 μm Evo C18

00A, LC column 50 × 2.1 mm, solvent A of 0.1% (v/v) (formic

acid) water, and solvent B of 0.1% (v/v) (formic acid) in acetonitrile.

mg, 0.84 mmol, 1.5 equiv). Product: crystalline solid (38 mg, 20%).

Purity: >96% (by HPLC). H NMR (600 MHz, DMSO-d

): δ 9.32

(br s, 1H, OH), 9.20 (d, J = 7.0 Hz, 1H, NH), 8.26 (t, J = 1.0 Hz, 1H,

Ar−H), 7.99 (dd, J = 8.0, 2.0 Hz, 1H, Ar−H), 7.52−7.45 (m, 3H,

Ar−H), 7.43−7.31 (m, 3H, Ar−H), 5.68 (d, J = 7.0 Hz, 1H,

H and C nuclear magnetic resonance (NMR) spectra were

−NHCH), 5.04 (s, 2H, −CH

OB), 3.66 (s, 3H, −OCH

); C NMR

(

151 MHz, DMSO-d ): δ 171.6, 167.4, 157.7, 137.6, 136.7, 133.0,

recorded using a Varian Mercury 300 MHz spectrometer or a Bruker

AVIII 600 MHz instrument. Deuterated solvents were used as

supplied. Chemical shifts (δ), referenced using residual solvent peaks,

are reported in parts per million downﬁeld from residual solvent peak

as an internal standard. Multiplicity is given as s (singlet), d (doublet),

t (triplet), q (quartet), m (multiplet), br (broad), or a combination of

these. Coupling constants, J, are reported in hertz (Hz) to the nearest

30.7, 129.3, 128.7, 127.6, 70.4, 57.4, 52.8; LCMS (ESI , m/z), 326

[

M + H] ; HRMS (ESI−TOF) calculated for C H N O B [M +

17 16

Na] , 348.1014; found, 348.1016.

N,N-Dibenzyl-1-hydroxy-3H-2,1-benzoxaborole-6-carboxa-

mide (14). General Protocol 1 was followed using the following

quantities of reagents: 3 (100 mg, 0.56 mmol, 1 equiv); N,N-DMF (2

mL); 1,1-carbonyldiimidazole (182 mg, 1.12 mmol, 2 equiv);

dibenzylamine (133 mg, 0.67 mmol, 1.2 equiv). Product: crystalline

.5 Hz. High-resolution mass spectra were recorded using a Bruker

MicroTOF instrument with an electrospray ionization source and

solid (20 mg, 10%). Purity: >99% (by HPLC). H NMR (300 MHz,

time of ﬂight (TOF) analyzer. The parent ion is quoted with the

−

DMSO-d ): δ 9.26 (br s, 1H, OH), 7.85 (s, 1H, Ar−H), 7.61−7.52

indicated ion: [M − H] or [M + Na] .

(

m, 1H, Ar−H), 7.47 (d, J = 8.0 Hz, 1H, Ar−H), 7.44−7.08 (m, 10H,

General Protocol 1: Amide Coupling. To a solution of the

appropriate carboxylic acid (1 equiv) in N,N-DMF (2 mL) was added

Ar−H), 5.02 (s, 2H, −CH OB), 4.75−4.24 (m, 4H, 2× −CH Ph).

C NMR (151 MHz, DMSO-d ): δ 171.6, 167.4, 157.7, 136.7, 133.0,

,1′-carbonyldiimidazole (CDI) (2 equiv). The reaction was stirred

for 5 min at room temperature; the appropriate amine (1, 1.2, or 1.5

130.7, 130.6, 129.3, 129.0, 128.7, 121.7, 70.4, 57.4, 52.7; LCMS

equiv) was then added, and the reaction was stirred for 4−16 h at 40

(ESI , m/z), 358 [M + H] ; HRMS (ESI−TOF) calcd for

10 +

°C. The solvent was removed in vacuo, and the crude product was

B [M + Na] , 380.1429; found, 380.1429.

puriﬁed using a Teledyne ISCO CombiFlash chromatography system

eluting with a reverse phase solvent gradient of MeOH in 0.1% (v/v)

CH CO H/water and a C18 column. The product-containing

General Protocol 2: Synthesis of 8, 9, and 10 (Scheme 1A).

Step (i): General Protocol 1 was followed to aﬀord an appropriate

methyl ester intermediate, B1, which was then directly subjected to

saponiﬁcation. Step (ii): To a solution of B1 (1 equiv) in 1,4-

dioxane/water (3:1; 10 mL) was added lithium hydroxide

monohydrate (either 4 or 6 equiv.) in one portion. The reaction

mixture was then stirred for 1 h at 40 °C before the volatiles were

removed in vacuo. The residue thus obtained was then lyophilized to

aﬀord corresponding free carboxylic acids (conﬁrmed by LC−MS

analysis) as solids. Step (iii): Crude carboxylic acids were immediately

coupled with selected piperazin-2-one derivatives using conditions

outlined in General Protocol 1, giving target benzoxaboroles as solids.

1-Hydroxy-N-[2-oxo-2-(3-oxopiperazin-1-yl)ethyl]-3H-2,1-

benzoxaborole-6-carboxamide (8). General Protocol 2 was

followed with the following quantities of reagents: step (i): 3 (250

mg, 1.40 mmol, 1 equiv); N,N-DMF (2 mL); 1,1-carbonyldiimidazole

(455 mg, 2.8 mmol, 2 equiv.) methyl glycinate HCl (133 mg, 0.67

mmol, 1.2 equiv). Step (ii): methyl 2-[(1-hydroxy-3H-2,1-benzox-

aborole-6-carbonyl)amino]acetate (349 mg, 1.40 mmol, 1 equiv); 1,4-

dioxane/water (3:1; 10 mL); lithium hydroxide monohydrate (235

mg, 5.61 mmol, 4 equiv). Step (iii): 2-[(1-hydroxy-3H-2,1-

benzoxaborole-6-carbonyl)amino]acetic acid (50 mg, 0.21 mmol, 1

equiv); N,N-DMF (2 mL); 1,1-carbonyldiimidazole (69 mg, 0.69

mmol, 2 equiv); piperazin-2-one (32 mg, 0.32 mmol, 1.5 equiv).

fractions were then combined, and the organic solvent was removed

in vacuo. When amide coupling yielded a target intermediate

compound (e.g., methyl esters of general structure B1, Scheme 1),

it was used in the next step without further puriﬁcation, else

lyophilization was used to aﬀord the desired products as solids (i.e.,

for 5−10 and 14).

-Hydroxy-N-[2-(methylamino)-2-oxo-ethyl]-3H-2,1-ben-

zoxaborole-6-carboxamide (5). General Protocol 1 was followed

using the following quantities of reagents: 3 (100 mg, 0.56 mmol, 1

equiv); N,N-DMF (2 mL); 1,1-carbonyldiimidazole (182 mg, 1.12

mmol, 2 equiv); 2-amino-N-methylacetamide HCl (104 mg, 0.84

mmol, 1.5 equiv). Product: crystalline solid (47 mg, 32%). Purity:

96% (by HPLC). H NMR (600 MHz, DMSO-d ): δ 9.55 (br s, 1H,

OH), 8.71 (t, J = 6.0 Hz, 1H, NH), 8.18 (s, 1H, Ar−H), 7.93 (dd, J =

.0, 2.0 Hz, 1H, Ar−H), 7.85 (q, J = 5.0 Hz, 1H, NHCH ), 7.49 (d, J

8.0 Hz, 1H, Ar−H), 5.02 (s, 2H, −CH OB), 3.85 (CH , obscured

by the solvent peak), 2.59 (d, J = 5.0 Hz, 3H, CH ); C NMR (151

MHz, DMSO-d ): δ 170.3, 167.8, 157.6, 133.2, 130.3, 130.1, 121.9,

0.4,43.1, 26.1; LCMS (ESI , m/z), 249 [M + H] ; HRMS (ESI−

10 −

TOF) calcd for C H N O B; [M − H] , 247.0896; found,

47.0895.

N-[(1R)-2-Amino-1-benzyl-2-oxo-ethyl]-1-hydroxy-3H-2,1-

Product: crystalline solid [14 mg, 4% (over three steps)]. Purity:

benzoxaborole-6-carboxamide (6). General Protocol 1 was

followed using the following quantities of reagents: 3 (100 mg, 0.56

mmol, 1 equiv); N,N-DMF (2 mL); 1,1-carbonyldiimidazole (182

mg, 1.12 mmol, 2 equiv); (R)-2-amino-3-phenylpropanamide HCl

>97% (by HPLC). H NMR (600 MHz, DMSO-d ): δ 9.37 (br s, 1H,

OH), 8.63−8.56 (m, 1H, NH), 8.33−8.20 (m, 1H, Ar−H), 7.96 (dd,

J = 8.0, 2.0 Hz, 1H, Ar−H), 7.51 (d, J = 8.0 Hz, 1H, Ar−H), 5.05 (s,

2H, −CH OB), 4.21−4.10 (m, 3H, CH ), 3.96 (s, 1H, CH ), 3.72−

(

169 mg, 0.84 mmol, 1.5 equiv). Product: crystalline solid (80 mg,

3.61 (m, 2H, CH ), 3.32−3.18 (m, 2H, CH ); C NMR (151 MHz,

3%). Purity: >98% (by HPLC). H NMR (600 MHz, DMSO-d ): δ

DMSO-d ): δ 172.1, 155.5, 135.4, 129.2, 129.2, 129.2, 129.1, 129.1,

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

The Supporting Information is available free of charge at

Article

27.9, 127.6, 127.5, 122.1, 70.4, 52.1; LCMS (ESI , m/z), 318 [M +

(151 MHz, DMSO-d ): δ 170.5, 167.3, 167.3, 157.8, 137.7, 132.1,

H] ; HRMS (ESI−TOF) calcd for C H N O B [M + Na] ,

131.9, 130.3, 129.6, 128.7, 127.1, 122.5, 70.4, 51.5, 46.1, 42.3, 40.4,

40.1076; found, 340.1076.

R)-1-Hydroxy-N-(1-(4-methyl-3-oxopiperazin-1-yl)-1-oxo-

-phenylpropan-2-yl)-1,3-dihydrobenzo[c][1,2]oxaborole-6-

37.5; LCMS (ESI , m/z), 408 [M + H] ; HRMS (ESI−TOF) calcd

10 +

(

for C H N O B [M + Na] , 430.1543; found, 430.1546.

21 22 3 5

Synthesis of 15 (Scheme 1C). Intermediates 15a and 15b were

81,82

carboxamide (9). General Protocol 2 was followed using the

following quantities of reagents: Step (i): 3 (100 mg, 0.56 mmol, 1

equiv); DMF (2 mL); 1,1-carbonyldiimidazole (182 mg, 1.12 mmol,

equiv); methyl D-phenylalaninate hydrochloride (121 mg, 0.56

mmol, 1 equiv). Step (ii): methyl(1-hydroxy-1,3-dihydrobenzo[c]-

1,2]oxaborole-6-carbonyl)-D-phenylalanine (187 mg, 0.55 mmol, 1

prepared as previously described

Synthesis of (4-(benzyloxy)-2-formylphenyl)boronic Acid Pinacol

Ester (15c). Bis(pinacolato)diboron (0.618 g, 2.43 mmol) (15b), 5-

(benzyloxy)-2-bromobenzaldehyde (0.500 g, 1.71 mmol), and

potassium acetate (0.478 g, 4.82 mmol) were dissolved in 1,4-

dioxane (15 mL); the suspension was then degassed with nitrogen,

[

equiv); 1,4-dioxane/water (3:1; 10 mL); lithium hydroxide

monohydrate (139 mg, 3.31 mmol, 6 equiv). Step (iii): (1-hydroxy-

after which PdCl (dppf) (0.119 g, 0.16 mmol) was added. The

reaction mixture was reﬂuxed for 12 h before it was allowed to cool to

room temperature. Solids were then removed by ﬁltration through

Celite washing with EtOAc (100 mL). The ﬁltrate was concentrated

in vacuo and the residue obtained was puriﬁed by Teledyne ISCO

CombiFlash automated chromatography (gradient of EtOAc in

hexane, 0−30%) to give a white solid (490 mg, 86%). Purity: >99%

,3-dihydrobenzo[c][1,2]oxaborole-6-carbonyl)-D-phenylalanine (30

mg, 0.09 mmol, 1 equiv); N,N-DMF (2 mL); 1,1-carbonyldiimidazole

30 mg, 0.18 mmol, 2 equiv); 1-methylpiperazin-2-one (16 mg, 0.14

(

mmol, 1.5 equiv). Product: crystalline solid [16 mg, 8% (over 3

steps)]. Purity: >97% (by HPLC). H NMR (300 MHz, DMSO-d ):

δ 8.85 (br s, 1H, OH), 8.24−8.16 (m, 1H, NH), 7.98−7.86 (m, 1H,

Ar−H), 7.48 (d, J = 8.0 Hz, 1H, Ar−H), 7.36−7.14 (m, 6H, Ar−H),

(by HPLC). H NMR (300 MHz, chloroform-d): δ 10.69 (s, 1H),

7.90 (d, J = 8.5 Hz, 1H), 7.62 (d, J = 2.5 Hz, 1H), 7.44 (tt, J = 4.0, 2.0

Hz, 2H), 7.41 − 7.37 (m, 3H), 7.22 (dd, J = 8.5, 2.5 Hz, 1H), 5.17 (s,

.20−4.95 (m, 2H, −CH OB), 4.20−3.93 (m, 2H), 3.34−2.96 (m,

+ +

H), 2.80 (s, 3H); LCMS (ESI , m/z), 422 [M + H] .

2H), 1.39 (s, 13H), 1.29 (s, 12H). LCMS (ESI , m/z): mass not

-Hydroxy-N-[2-(4-methyl-3-oxo-piperazin-1-yl)-2-oxo-

observed.

ethyl]-3H-2,1-benzoxaborole-6-carboxamide (10). General Pro-

tocol 2 was followed with the following quantities of reagents: Step

Synthesis of Ethyl 2-(5-(Benzyloxy)-1-hydroxy-1,3-dihydrobenzo-

[c][1,2]oxaborol-3-yl)acetate (15d). To a stirred solution of ethyl

acetate (0.32 g, 3.63 mmol, 1.5 equiv) in dry tetrahydrofuran (THF)

was added lithium diisopropylamide (LDA, 3.55 mL, 3.63 mmol, 1.5

equiv, 1 M solution in THF) dropwise at −78 °C. The mixture was

stirred for 30 min at −78 °C before (4-(benzyloxy)-2-formylphenyl)-

boronic acid pinacol ester (0.8 g, 2.36 mmol, 1 equiv) was added.

Next, the reaction mixture was stirred for 2 h at −20 °C before it was

(

i): 3 (250 mg, 1.40 mmol, 1 equiv); N,N-DMF (2 mL); 1,1-

carbonyldiimidazole (455 mg, 2.8 mmol, 2 equiv.) methyl glycinate

HCl (133 mg, 0.67 mmol, 1.2 equiv). Step (ii): methyl 2-[(1-hydroxy-

H-2,1-benzoxaborole-6-carbonyl)amino]acetate (349 mg, 1.40

mmol, 1 equiv); 1,4-dioxane/water (3:1; 10 mL); lithium hydroxide

monohydrate (235 mg, 5.61 mmol, 4 equiv). Step (iii): 2-[(1-

hydroxy-3H-2,1-benzoxaborole-6-carbonyl)amino]acetic acid (47 mg,

quenched by addition of a saturated solution of NH Cl (10 mL).

EtOAc (10 mL) was added and the layers were separated. The

.2 mmol, 1 equiv); N,N-DMF (2 mL); 1,1-carbonyldiimidazole (65

mg, 0.4 mmol, 2 equiv); 1-methylpiperazin-2-one (34 mg, 0.30 mmol,

.5 equiv). Product: crystalline solid [17 mg, 5% (over 3 steps)].

aqueous layer was extracted with EtOAc (3 × 10 mL). The combined

organic extracts were washed with brine, dried (MgSO ), and then

concentrated in vacuo. The residue obtained was then puriﬁed by

Teledyne ISCO CombiFlash automated chromatography (gradient of

EtOAc in hexane in 0−30%) to give the desired product as a white

Purity: >96% (by HPLC). H NMR (300 MHz, DMSO-d ): δ 8.58

(

br s, 1H, OH), 8.26 (s, 1H, Ar−H), 7.97 (d, J = 6.0 Hz, 1H, Ar−H),

.51 (d, J = 8.0 Hz, 1H, Ar−H), 5.05 (s, 2H, CH ), 4.24−4.07 (m,

solid (160 mg, 57%). Purity >99% (by HPLC). H NMR (300 MHz,

H, −CH OB and CH ), 4.05−3.94 (m, 2H, CH ), 3.48−3.30 (m,

methanol-d ): δ 7.62−7.53 (m, 1H), 7.51−7.30 (m, 5H), 7.04−6.97

H, CH ), 2.89 (s, 3H, −NCH ); C NMR (151 MHz, DMSO-d ):

(

m, 2H), 5.53 (dd, J = 8.5, 4.5 Hz, 1H), 5.14 (s, 2H), 4.18 (qd, J =

δ167.6, 167.3, 165.2, 157.5, 133.4, 130.2, 130.0, 121.8, 70.3, 48.0,

7.0, 1.5 Hz, 2H), 2.95 (dd, J = 15.5, 4.5 Hz, 1H), 2.55 (dd, J = 15.5,

H] .

7.5, 46.2, 41.2, 33.9; LCMS (ESI , m/z), 332 [M + H] .

.5 Hz, 1H), 1.26 (t, J = 7.0 Hz, 3H). LCMS (ESI , m/z), 327 [M +

Synthesis of 11 (Scheme 1B). Step (i). To a solution of (11a)

tert-butoxycarbonyl)-D-phenylalanine) (300 mg, 1.13 mmol) in N,N-

(

Synthesis of 2-(5-(Benzyloxy)-1-hydroxy-1,3-dihydrobenzo[c]-

1,2]oxaborol-3-yl)acetic Acid (15). To a stirred solution of ethyl

DMF (3 mL) was added 1,1-carbonyldiimidazole (367 mg, 2.26

mmol); the reaction was stirred for 5 min at room temperature. To

the reaction mixture was then added piperazin-2-one (170 mg, 1.70

mmol) and the resultant solution was stirred for 4 h at 40 °C. Ethyl

acetate (10 mL) and water (10 mL) were added. The layers were

separated. The aqueous layer was extracted with ethyl acetate (3 × 20

[

-(5-(benzyloxy)-1-hydroxy-1,3-dihydrobenzo[c][1,2]oxaborol-3-yl)-

acetate (0.3 g, 0.92 mmol, 1 equiv) in THF/H O (10 mL, 1:1) was

added LiOH·H O (77 mg, 1.84 mmol, 2 equiv); the mixture was then

stirred for 2 h at room temperature. Water (20 mL) and ethyl acetate

(

20 mL) were then added and the layers were separated. The aqueous

layer was acidiﬁed to pH 2−3 and then extracted with ethyl acetate (3

20 mL). The layers were separated and the combined organic

mL). Combined organic layers were dried (MgSO ), ﬁltered, and

concentrated in vacuo to aﬀord 11b (tert-butyl N-[(1R)-1-benzyl-2-

oxo-2-(3-oxopiperazin-1-yl)ethyl]carbamate) as a yellow oil (350 mg,

extracts were washed with brine, dried (MgSO ), and then

8%), which was used in the next step without further puriﬁcation.

Step (ii). To a stirred solution of (11b) (350 mg, 1.01 mmol, 1

concentrated in vacuo. On standing at room temperature overnight,

the product solidiﬁed. Further washing with diethyl ether gave the

equiv) in CH Cl (10 mL) was added HCl (4 M in 1,4-dioxane, 0.76

puriﬁed product as a white solid (210 mg, 78%). Purity >97% (by

mL, 3.02 mmol, 3 equiv) at room temperature under nitrogen. The

reaction mixture was stirred overnight to aﬀord an insoluble

precipitate. The precipitate, 11c, was washed three times with

CH Cl and then ﬁltered, dried in air, and used in the next step

HPLC). H NMR (300 MHz, methanol-d ): δ 7.58 (d, J = 8.0 Hz,

H), 7.49−7.32 (m, 5H), 7.07−6.98 (m, 2H), 5.54 (dd, J = 8.5, 4.5

Hz, 1H), 5.14 (s, 2H), 2.89 (dd, J = 15.5, 4.5 Hz, 1H), 2.51 (dd, J =

5.5, 8.5 Hz, 1H). C NMR (151 MHz, DMSO-d ): δ 172.4, 161.4,

without further puriﬁcation (54 mg, 19%).

58.8, 137.3, 132.2, 128.9, 128.4, 128.4, 115.5, 107.6, 77.3, 69.8, 42.3.

Step (iii). General Protocol 1 was followed using the following

quantities of reagents: 3 (54 mg, 0.3 mmol, 1 equiv); N,N-DMF (3

mL); 1,1-carbonyldiimidazole (98 mg, 0.600 mmol, 2 equiv); 11c

LCMS (ESI , m/z), 299 [M + H] .

ASSOCIATED CONTENT

sı Supporting Information

https://pubs.acs.org/doi/10.1021/acs.jmedchem.1c00717 .

(

103 mg, 0.360 mmol, 1.2 equiv). Product (11): white powder (20

■

mg, 16%). Purity: >99% (by HPLC). H NMR (300 MHz, methanol-

d ): δ 8.31 (br s, 1H, OH), 8.14 (dd, J = 8.0, 2.0 Hz, 2H, Ar−H), 7.93

(

d, J = 8.0 Hz, 1H, Ar−H), 7.56−7.45 (m, 2H, Ar−H), 7.36−7.22 (m,

H, Ar−H), 5.15 (m, 5H, 2× CH and CH), 4.33−3.96 (m, 2H,

CH ), 3.88−3.40 (m, 3H, CH ), 3.28−3.06 (m, 3H, CH ); C NMR

PDB model of a docked PaPBP3:11 complex (PDB)

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Complete contact information is available at:

Article

Methods for antimicrobial and nitroceﬁn assays and

details of the fragment selection process; computational

chemistry protocols; model and inhibition curves for the

BOCILLIN FL and nitroceﬁn assays; microbiological

data; crystallography statistics and crystallization pHs for

each model; additional crystal structure views; examples

of boron-based inhibition of other targets via reaction

with nucleophilic residues; and HPLC traces for lead

compounds (PDF)

Jurgen Brem − Department of Chemistry and the Ineos

Oxford Institute of Antimicrobial Research, Chemistry

Research Laboratory, Oxford OX1 3TA, U.K.; orcid.org/

0000-0002-0137-3226

Kelly Chibale − Drug Discovery and Development Centre

(H3D) and South African Medical Research Council Drug

Discovery and Development Research Unit, Department of

Chemistry and Institute of Infectious Disease and Molecular

Medicine, University of Cape Town, Rondebosch 7701, South

Africa; orcid.org/0000-0002-1327-4727

Molecular formula strings (CSV)

https://pubs.acs.org/10.1021/acs.jmedchem.1c00717

AUTHOR INFORMATION

■

Corresponding Authors

Author Contributions

Charles J. Eyermann − Drug Discovery and Development

Centre (H3D), University of Cape Town, Rondebosch 7701,

South Africa; Email: eyermanncj@gmail.com

H.N., C.J.E., F.v.D., K.C., J.B., C.J.S., and C.G.D. conceived

and designed the study. H.N., A.K., D.B., C.J.E., and G.A.B

carried out the experiments. R.L. and M.G. synthesized

compounds. H.N., A.K., C.J.E., J.B., C.J.S, and C.G.D wrote

the manuscript. C.J.E., K.E.M., and N.G.P. helped supervise

the project. K.C., C.J.S., and C.G.D. supervised the project.

Christopher J. Schoﬁeld − Department of Chemistry and the

Ineos Oxford Institute of Antimicrobial Research, Chemistry

Research Laboratory, Oxford OX1 3TA, U.K.; orcid.org/

000-0002-0290-6565 ; Email: christopher.scho ﬁ eld@

chem.ox.ac.uk

Notes

Christopher G. Dowson − School of Life Sciences, University

The authors declare no competing ﬁnancial interest.

PaPBP3 in complex with 1 (7ATM), 2 (7ATO), 3 (7ATW), 4

(7ATX), 7 (7AU0), 12 (7AU1), 13 (7AU8), 14 (7AU9), 15

of Warwick, Coventry CV4 7AL, U.K.; orcid.org/0000-

002-8294-8836 ; Email: C.G.Dowson@warwick.ac.uk

(

7AUB) and vaborbactam (7AUH). The authors will release

Authors

the atomic coordinates and experimental data upon article

publication.

Hector Newman − School of Life Sciences, University of

Warwick, Coventry CV4 7AL, U.K.; Diamond Light Source

Ltd, Didcot OX11 0DE, U.K.

Alen Krajnc − Department of Chemistry and the Ineos Oxford

Institute of Antimicrobial Research, Chemistry Research

Laboratory, Oxford OX1 3TA, U.K.; Present

Address: Faculty of Pharmacy, University of Ljubljana,

Askerceva cesta 7, 1000 Ljubljana, Slovenia; orcid.org/

ACKNOWLEDGMENTS

■

This work was supported by the Newton Fund (grant number

MR/P007503/1); a Collaborative Postgraduate award be-

tween the University of Warwick and Diamond Light Source

(

grant number STU0212); a GCRF Networking grant (grant

number GCRFNG\100347; the Medical Research Council

grant numbers MR/N002679/1, and MR/P007503/1). This

000-0001-7822-1944

(

Dom Bellini − School of Life Sciences, University of Warwick,

Coventry CV4 7AL, U.K.; Present Address: X-ray

Crystallography Facility, MRC Laboratory of Molecular

Biology, Francis Crick Avenue, Cambridge Biomedical

Campus, CB2 0QH, United Kingdom

research was funded in whole, or in part, by the Wellcome

Trust (grant number 106244/Z/14/Z). The Novartis

Research Foundation, South African Medical Research

Council, and South African Research Chairs Initiative of the

Department of Science and Innovation administered through

the South African National Research Foundation are gratefully

acknowledged for support (K.C.). The help of the XChem

facility at Diamond in the crystallographic screening is

gratefully acknowledged. We thank Wuxi AppTec for the

synthesis of 12 and 13. Thank you to HI Zgurskaya for

providing the kind gift of the permeabilized P. aeruginosa

strain. C.J.S. thanks the Innovative Medicines Initiative

Grant A. Boyle − Drug Discovery and Development Centre

H3D), University of Cape Town, Rondebosch 7701, South

Africa

Neil G. Paterson − Diamond Light Source Ltd, Didcot OX11

DE, U.K.

Katherine E. McAuley − Diamond Light Source Ltd, Didcot

OX11 0DE, U.K.; Present Address: Paul Scherrer Institute,

Forschungsstrasse 111, 5232 Villigen PSI, Switzerland

Robert Lesniak − Department of Chemistry and the Ineos

Oxford Institute of Antimicrobial Research, Chemistry

Research Laboratory, Oxford OX1 3TA, U.K.; Present

Address: Stanford ChEM-H, Medicinal Chemistry

Knowledge Center, 290 Jane Stanford Way, Stanford, CA

(

European Lead factory and ENABLE components), the

Medical Research Council, the Wellcome Trust, Cancer

Research UK, and the Ineos Institute for Antimicrobial

Research for funding our work on antibiotics, MBL fold/

metalloenzymes, and BL inhibitors. A.K. thanks The Public

Scholarship, Development, Disability and Maintenance Fund

of the Republic of Slovenia (Ad Futura). For the purpose of

open access, the author has applied a CC BY public copyright

license to any Author Accepted Manuscript version arising

from this submission.

4305, United States

Mukesh Gangar − Drug Discovery and Development Centre

H3D), University of Cape Town, Rondebosch 7701, South

Africa

Frank von Delft − Diamond Light Source Ltd, Didcot OX11

DE, U.K.; Structural Genomics Consortium (SGC),

ABBREVIATIONS

University of Oxford, Oxford, U.K.; Department of

■

Biochemistry, University of Johannesburg, Auckland Park

BL, β-lactamase; CDI, 1,1′-carbonyldiimidazole; DBO, dia-

zabicyclooctane; FA, ﬂuorescence anisotropy; MBL, metallo-β-

lactamase; ND, not determined; PaPBP3, Paeruginosa PBP3;

006, South Africa; Research Complex at Harwell, Didcot

OX11 0FA, U.K.; orcid.org/0000-0003-0378-0017

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

of Beta-Lactamase Inhibition and Impact of Resistance Mechanisms

Article

PBP, penicillin-binding protein; PDB, protein data bank; S2d,

on Activity in Enterobacteriaceae. Antimicrob. Agents Chemother.

-((benzoyl-D-alanyl)thio)acetic acid; SBL, serine β-lactamase

(

017, 61, No. e01443-17.

18) Tsivkovski, R.; Lomovskaya, O. Biochemical Activity of

Vaborbactam. Antimicrob. Agents Chemother. 2019, 64, e01935−19.

19) Krajnc, A.; Brem, J.; Hinchliffe, P.; Calvopina, K.;

REFERENCES

■

1) Spratt, B. G. Distinct Penicillin Binding Proteins Involved in the

(

Division, Elongation, and Shape of Escherichia Coli K12. Proc. Natl.

Acad. Sci. U.S.A. 1975 , 72 , 2999 −3003.

2) Spratt, B. G. Temperature-Sensitive Cell Division Mutants of

Panduwawala, T. D.; Lang, P. A.; Kamps, J. J. A. G.; Tyrrell, J. M.;

Widlake, E.; Saward, B. G.; Walsh, T. R.; Spencer, J.; Schofield, C. J.

Bicyclic Boronate VNRX-5133 Inhibits Metallo- and Serine-β -

Lactamases. J. Med. Chem. 2019, 62, 8544−8556.

Escherichia Coli with Thermolabile Penicillin-Binding Proteins. J.

Bacteriol. 1977, 131, 293−305.

(20) Liu, B.; Trout, R. E. L.; Chu, G.-H.; McGarry, D.; Jackson, R.

3) Sauvage, E.; Kerff, F.; Terrak, M.; Ayala, J. A.; Charlier, P. The

W.; Hamrick, J. C.; Daigle, D. M.; Cusick, S. M.; Pozzi, C.; De Luca,

F.; Benvenuti, M.; Mangani, S.; Docquier, J.-D.; Weiss, W. J.; Pevear,

D. C.; Xerri, L.; Burns, C. J. Discovery of Taniborbactam (VNRX-

Penicillin-Binding Proteins: Structure and Role in Peptidoglycan

Biosynthesis. FEMS Microbiol. Rev. 2008, 32, 234−258.

(

4) Adam, M.; Fraipont, C.; Rhazi, N.; Nguyen-Disteche, M.;

133): A Broad-Spectrum Serine- and Metallo-β -Lactamase Inhibitor

for Carbapenem-Resistant Bacterial Infections. J. Med. Chem. 2020,

3, 2789−2801.

21) Hamrick, J. C.; Docquier, J.-D.; Uehara, T.; Myers, C. L.; Six,

Lakaye, B.; Frere, J. M.; Devreese, B.; Van Beeumen, J.; van

Heijenoort, Y.; van Heijenoort, J.; Ghuysen, J. M. The Bimodular

G57-V577 Polypeptide Chain of the Class B Penicillin-Binding

Protein 3 of Escherichia Coli Catalyzes Peptide Bond Formation from

Thiolesters and Does Not Catalyze Glycan Chain Polymerization

from the Lipid II Intermediate. J. Bacteriol. 1997 , 179 , 6005 −6009.

(

D. A.; Chatwin, C. L.; John, K. J.; Vernacchio, S. F.; Cusick, S. M.;

Trout, R. E. L.; Pozzi, C.; De Luca, F.; Benvenuti, M.; Mangani, S.;

Liu, B.; Jackson, R. W.; Moeck, G.; Xerri, L.; Burns, C. J.; Pevear, D.

C.; Daigle, D. M. VNRX-5133 (Taniborbactam), a Broad-Spectrum

Inhibitor of Serine- and Metallo-β -Lactamases, Restores Activity of

Cefepime in Enterobacterales and Pseudomonas Aeruginosa.

Antimicrob. Agents Chemother. 2019, 64, No. e01963.

(22) Brem, J.; Cain, R.; Cahill, S.; McDonough, M. A.; Clifton, I. J.;

Jiménez-Castellanos, J.-C.; Avison, M. B.; Spencer, J.; Fishwick, C. W.

G.; Schofield, C. J. Structural Basis of Metallo-β -Lactamase, Serine-β -

Lactamase and Penicillin-Binding Protein Inhibition by Cyclic

Boronates. Nat. Commun. 2016, 7, 12406.

5) Tipper, D. J.; Strominger, J. L. Mechanism of Action of

Penicillins: A Proposal Based on Their Structural Similarity to Acyl-

D-Alanyl-D-Alanine. Proc. Natl. Acad. Sci. U.S.A. 1965, 54, 1133−

141.

An Overview. Cold Spring Harbor Perspect. Med. 2016 , 6 , a025247.

6) Bush, K.; Bradford, P. A. β -Lactams and β -Lactamase Inhibitors:

7) Bush, K. Resurgence of β -Lactamase Inhibitor Combinations

8) Krajnc, A.; Lang, P. A.; Panduwawala, T. D.; Brem, J.; Schofield,

Effective against Multidrug-Resistant Gram-Negative Pathogens. Int. J.

Antimicrob. Agents 2015, 46, 483−493.

(

C. J. Will Morphing Boron-Based Inhibitors Beat the β -Lactamases?

(23) Zervosen, A.; Bouillez, A.; Herman, A.; Amoroso, A.; Joris, B.;

Sauvage, E.; Charlier, P.; Luxen, A. Synthesis and Evaluation of

Boronic Acids as Inhibitors of Penicillin Binding Proteins of Classes

A, B and C. Bioorg. Med. Chem. 2012, 20, 3915−3924.

Curr. Opin. Chem. Biol. 2019, 50, 101−110.

9) Drawz, S. M.; Papp-Wallace, K. M.; Bonomo, R. A. New β -

Lactamase Inhibitors: A Therapeutic Renaissance in an MDR World.

Antimicrob. Agents Chemother. 2014, 58, 1835−1846.

(24) Inglis, S. R.; Zervosen, A.; Woon, E. C. Y.; Gerards, T.; Teller,

(

10) Wang, D. Y.; Abboud, M. I.; Markoulides, M. S.; Brem, J.;

N.; Fischer, D. S.; Luxen, A.; Schofield, C. J. Synthesis and Evaluation

of 3-(Dihydroxyboryl)Benzoic Acids as d , d -Carboxypeptidase R39

Inhibitors. J. Med. Chem. 2009, 52, 6097−6106.

Schofield, C. J. The Road to Avibactam: The First Clinically Useful

Non-β -Lactam Working Somewhat like a β -Lactam. Future Med.

Chem. 2016, 8, 1063−1084.

(25) Contreras-Martel, C.; Amoroso, A.; Woon, E. C. Y.; Zervosen,

(

11) Hirsch, E. B.; Ledesma, K. R.; Chang, K.-T.; Schwartz, M. S.;

A.; Inglis, S.; Martins, A.; Verlaine, O.; Rydzik, A. M.; Job, V.; Luxen,

A.; Joris, B.; Schofield, C. J.; Dessen, A. Structure-Guided Design of

Cell Wall Biosynthesis Inhibitors That Overcome β -Lactam

Resistance in Staphylococcus Aureus (MRSA). ACS Chem. Biol.

Motyl, M. R.; Tam, V. H. In Vitro Activity of MK-7655, a Novel β -

Lactamase Inhibitor, in Combination with Imipenem against

Carbapenem-Resistant Gram-Negative Bacteria. Antimicrob. Agents

Chemother. 2012, 56, 3753−3757.

(

011, 6, 943−951.

(

12) Blizzard, T. A.; Chen, H.; Kim, S.; Wu, J.; Bodner, R.; Gude, C.;

26) Woon, E. C. Y.; Zervosen, A.; Sauvage, E.; Simmons, K. J.;

Imbriglio, J.; Young, K.; Park, Y.-W.; Ogawa, A.; Raghoobar, S.;

Hairston, N.; Painter, R. E.; Wisniewski, D.; Scapin, G.; Fitzgerald, P.;

Sharma, N.; Lu, J.; Ha, S.; Hermes, J.; Hammond, M. L. Discovery of

MK-7655, a β -Lactamase Inhibitor for Combination with Primaxin ®.

Bioorg. Med. Chem. Lett. 2014, 24, 780−785.

Zivec, M.; Inglis, S. R.; Fishwick, C. W. G.; Gobec, S.; Charlier, P.;

Luxen, A.; Schofield, C. J. Structure Guided Development of Potent

Reversibly Binding Penicillin Binding Protein Inhibitors. ACS Med.

Chem. Lett. 2011, 2, 219−223.

(

27) Zervosen, A.; Herman, R.; Kerff, F.; Herman, A.; Bouillez, A.;

13) Livermore, D. M.; Warner, M.; Mushtaq, S. Activity of MK-

655 Combined with Imipenem against Enterobacteriaceae and

Pseudomonas Aeruginosa. J. Antimicrob. Chemother. 2013, 68, 2286−

14) Theuretzbacher, U.; Bush, K.; Harbarth, S.; Paul, M.; Rex, J. H.;

Prati, F.; Pratt, R. F.; Frére, J.-M.; Joris, B.; Luxen, A.; Charlier, P.;

Sauvage, E. Unexpected Tricovalent Binding Mode of Boronic Acids

within the Active Site of a Penicillin-Binding Protein. J. Am. Chem.

Soc. 2011, 133, 10839−10848.

(

(28) Pechenov, A.; Stefanova, M. E.; Nicholas, R. A.; Peddi, S.;

Tacconelli, E.; Thwaites, G. E. Critical Analysis of Antibacterial

Agents in Clinical Development. Nat. Rev. Microbiol. 2020, 18, 286−

Gutheil, W. G. Potential Transition State Analogue Inhibitors for the

Penicillin-Binding Proteins. Biochemistry 2003, 42, 579−588.

98.

(29) Nicola, G.; Peddi, S.; Stefanova, M.; Nicholas, R. A.; Gutheil,

15) Adams, J.; Kauffman, M. Development of the Proteasome

Inhibitor Velcade (Bortezomib). Canc. Invest. 2004, 22, 304−311.

16) Hecker, S. J.; Reddy, K. R.; Totrov, M.; Hirst, G. C.;

W. G.; Davies, C. Crystal Structure of Escherichia Coli Penicillin-

Binding Protein 5 Bound to a Tripeptide Boronic Acid Inhibitor: A

(

†

Role for Ser-110 in Deacylation . Biochemistry 2005, 44, 8207−8217.

Lomovskaya, O.; Griffith, D. C.; King, P.; Tsivkovski, R.; Sun, D.;

Sabet, M.; Tarazi, Z.; Clifton, M. C.; Atkins, K.; Raymond, A.; Potts,

K. T.; Abendroth, J.; Boyer, S. H.; Loutit, J. S.; Morgan, E. E.; Durso,

S.; Dudley, M. N. Discovery of a Cyclic Boronic Acid β -Lactamase

Inhibitor (RPX7009) with Utility vs Class A Serine Carbapenemases.

J. Med. Chem. 2015, 58, 3682−3692.

(30) Dzhekieva, L.; Rocaboy, M.; Kerff, F.; Charlier, P.; Sauvage, E.;

Pratt, R. F. Crystal Structure of a Complex between the Actinomadura

R39 Dd-Peptidase and a Peptidoglycan-Mimetic Boronate Inhibitor:

Interpretation of a Transition State Analogue in Terms of Catalytic

Mechanism. Biochemistry 2010, 49, 6411−6419.

(

17) Lomovskaya, O.; Sun, D.; Rubio-Aparicio, D.; Nelson, K.;

(31) Dzhekieva, L.; Kumar, I.; Pratt, R. F. Inhibition of Bacterial

DD-Peptidases (Penicillin-Binding Proteins) in Membranes and in

Tsivkovski, R.; Griffith, D. C.; Dudley, M. N. Vaborbactam: Spectrum

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

(46) Stefanova, M. E.; Davies, C.; Nicholas, R. A.; Gutheil, W. G.

Article

Vivo by Peptidoglycan-Mimetic Boronic Acids. Biochemistry 2012, 51,

Overview of the CCP 4 Suite and Current Developments. Acta

Crystallogr. Sect. D Biol. Crystallogr. 2011, 67, 235−242.

Inhibitor, and Substrate Specificity Studies on Escherichia Coli

Penicillin-Binding Protein 5. Biochim. Biophys. Acta Protein Struct. Mol.

Enzymol. 2002, 1597, 292−300.

(

804−2811.

32) Burns, J.; Daigle, D.; Chu, G.-H.; Jackson, R. W.; Hamrick, J.;

Boyd, S. A.; Zulli, A. L.; Mesaros, E. F. Penicillin-Binding Protein

Inhibitors. WO 2018218154 A1, 2018.

(

33) Burns, J.; Daigle, D.; Chu, G.-H.; Hamrick, J.; Lucas, M.; Boyd,

S. A.; Zulli, A. L.; Mesaros, E. F.; Condon, S. M.; Trout, R. E. L.;

Myers, C. L. Penicillin-Binding Protein Inhibitors. WO 2018218190

A1, 2018.

(47) Davies, C.; White, S. W.; Nicholas, R. A. Crystal Structure of a

Deacylation-Defective Mutant of Penicillin-Binding Protein 5 at 2.3-Å

Resolution. J. Biol. Chem. 2001, 276, 616−623.

(48) Thomas, B.; Wang, Y.; Stein, R. L. Kinetic and Mechanistic

Studies of Penicillin-Binding Protein 2x from Streptococcus Pneumo-

niae. Biochemistry 2001, 40, 15811−15823.

(49) Dave, K.; Palzkill, T.; Pratt, R. F. Neutral β -Lactams Inactivate

High Molecular Mass Penicillin-Binding Proteins of Class B1,

Including PBP2a of MRSA. ACS Med. Chem. Lett. 2014, 5, 154−157.

(50) Han, S.; Zaniewski, R. P.; Marr, E. S.; Lacey, B. M.; Tomaras, A.

P.; Evdokimov, A.; Miller, J. R.; Shanmugasundaram, V. Structural

Basis for Effectiveness of Siderophore-Conjugated Monocarbams

against Clinically Relevant Strains of Pseudomonas Aeruginosa. Proc.

Natl. Acad. Sci. U.S.A. 2010, 107, 22002−22007.

(51) Bellini, D.; Koekemoer, L.; Newman, H.; Dowson, C. G. Novel

and Improved Crystal Structures of H. Influenzae, E. Coli and P.

Aeruginosa Penicillin-Binding Protein 3 (PBP3) and N. Gonorrhoeae

PBP2: Toward a Better Understanding of β -Lactam Target-Mediated

Resistance. J. Mol. Biol. 2019, 431, 3501−3519.

34) Inglis, S. R.; Strieker, M.; Rydzik, A. M.; Dessen, A.; Schofield,

C. J. A Boronic-Acid-Based Probe for Fluorescence Polarization

Assays with Penicillin Binding Proteins and β -Lactamases. Anal.

Biochem. 2012, 420, 41−47.

(

35) Nemmara, V. V.; Dzhekieva, L.; Subarno Sarkar, K.; Adediran,

S. A.; Duez, C.; Nicholas, R. A.; Pratt, R. F. Substrate Specificity of

Low-Molecular Mass Bacterial DD-Peptidases. Biochemistry 2011, 50,

0091−10101.

36) Dzhekieva, L.; Adediran, S. A.; Pratt, R. F. Interactions of

Bora-Penicilloates ” with Serine β -Lactamases and DD-Peptidases.

Biochemistry 2014, 53, 6530−6538.

37) Moya, B.; Barcelo, I. M.; Bhagwat, S.; Patel, M.; Bou, G.; Papp-

Wallace, K. M.; Bonomo, R. A.; Oliver, A. WCK 5107 (Zidebactam)

“

(

and WCK 5153 Are Novel Inhibitors of PBP2 Showing Potent “β -

Lactam Enhancer ” Activity against Pseudomonas Aeruginosa,

Including Multidrug-Resistant Metallo-β -Lactamase-Producing High-

Risk Clones. Antimicrob. Agents Chemother. 2017, 61, No. e02529.

(52) Shapiro, A. B.; Gu, R.-F.; Gao, N.; Livchak, S.; Thresher, J.

Continuous Fluorescence Anisotropy-Based Assay of BOCILLIN FL

Penicillin Reaction with Penicillin Binding Protein 3. Anal. Biochem.

2013, 439, 37−43.

(

38) Levy, N.; Bruneau, J.-M.; Le Rouzic, E.; Bonnard, D.; Le Strat,

F.; Caravano, A.; Chevreuil, F.; Barbion, J.; Chasset, S.; Ledoussal, B.;

Moreau, F.; Ruff, M. Structural Basis for E. Coli Penicillin Binding

Protein (PBP) 2 Inhibition, a Platform for Drug Design. J. Med. Chem.

(53) Jamin, M.; Damblon, C.; Millier, S.; Hakenbeck, R.; Frere, J. M.

(

019, 62, 4742−4754.

Penicillin-Binding Protein 2x of Streptococcus Pneumoniae: Enzymic

Activities and Interactions with Beta-Lactams. Biochem. J. 1993, 292,

735.

39) Durand-Réville, T. F.; Guler, S.; Comita-Prevoir, J.; Chen, B.;

Bifulco, N.; Huynh, H.; Lahiri, S.; Shapiro, A. B.; McLeod, S. M.;

Carter, N. M.; Moussa, S. H.; Velez-Vega, C.; Olivier, N. B.;

McLaughlin, R.; Gao, N.; Thresher, J.; Palmer, T.; Andrews, B.;

Giacobbe, R. A.; Newman, J. V.; Ehmann, D. E.; de Jonge, B.;

O ’Donnell, J.; Mueller, J. P.; Tommasi, R. A.; Miller, A. A. ETX2514

Is a Broad-Spectrum β -Lactamase Inhibitor for the Treatment of

Drug-Resistant Gram-Negative Bacteria Including Acinetobacter

Baumannii. Nat. Rev. Microbiol. 2017, 2, 17104.

(54) Boes, A.; Olatunji, S.; Breukink, E.; Terrak, M. Regulation of

the Peptidoglycan Polymerase Activity of PBP1b by Antagonist

Actions of the Core Divisome Proteins FtsBLQ and FtsN. mBio 2019,

10 (). DOI: 10.1128/mBio.01912-18

(55) Spratt, B. G. Hybrid Penicillin-Binding Proteins in Penicillin-

Resistant Strains of Neisseria Gonorrhoeae. Nature 1988, 332, 173−

176.

(

40) McIntyre, P. J.; Collins, P. M.; Vrzal, L.; Birchall, K.; Arnold, L.

(56) Barbour, A. G. Properties of Penicillin-Binding Proteins in

Neisseria Gonorrhoeae. Antimicrob. Agents Chemother. 1981, 19, 316−

322.

H.; Mpamhanga, C.; Coombs, P. J.; Burgess, S. G.; Richards, M. W.;

Winter, A.; Veverka, V.; Delft, F. v.; Merritt, A.; Bayliss, R.

Characterization of Three Druggable Hot-Spots in the Aurora-A/

TPX2 Interaction Using Biochemical, Biophysical, and Fragment-

Based Approaches. ACS Chem. Biol. 2017, 12, 2906−2914.

(57) O ’Callaghan, C. H.; Morris, A.; Kirby, S. M.; Shingler, A. H.

Novel Method for Detection of -Lactamases by Using a Chromogenic

Cephalosporin Substrate. Antimicrob. Agents Chemother. 1972 , 1 , 283.

(58) Graves-Woodward, K.; Pratt, R. F. Reaction of Soluble

Penicillin-Binding Protein 2a of Methicillin-Resistant Staphylococcus

Aureus with β -Lactams and Acyclic Substrates: Kinetics in

Homogeneous Solution. Biochem. J. 1998 , 332 , 755 −761.

(59) Johnson, K. A. New Standards for Collecting and Fitting Steady

State Kinetic Data. Beilstein J. Org. Chem. 2019, 15, 16−29.

(60) Krishnamoorthy, G.; Wolloscheck, D.; Weeks, J. W.; Croft, C.;

Rybenkov, V. V.; Zgurskaya, H. I. Breaking the Permeability Barrier of

Escherichia Coli by Controlled Hyperporination of the Outer

Membrane. Antimicrob. Agents Chemother. 2016 , 60 , 7372.

(61) Kocaoglu, O.; Carlson, E. E. Profiling of β -Lactam Selectivity

for Penicillin-Binding Proteins in Escherichia Coli Strain DC2.

Antimicrob. Agents Chemother. 2015, 59, 2785−2790.

(

41) Kidd, S. L.; Fowler, E.; Reinhardt, T.; Compton, T.; Mateu, N.;

Newman, H.; Bellini, D.; Talon, R.; McLoughlin, J.; Krojer, T.;

Aimon, A.; Bradley, A.; Fairhead, M.; Brear, P.; Díaz-Sáez, L.;

McAuley, K.; Sore, H. F.; Madin, A.; O’Donovan, D. H.; Huber, K. V.

M.; Hyvönen, M.; von Delft, F.; Dowson, C. G.; Spring, D. R.

Demonstration of the Utility of DOS-Derived Fragment Libraries for

Rapid Hit Derivatisation in a Multidirectional Fashion. Chem. Sci.

43) Enamine Serine focused Covalent Fragments Library. https://

020, 11, 10792−10801.

42) Martin, J. S.; MacKenzie, C. J.; Fletcher, D.; Gilbert, I. H.

Characterising Covalent Warhead Reactivity. Bioorg. Med. Chem.

019, 27, 2066−2074.

enamine.net/index.php?option=com_content&view=article&id=

16&Itemid=720. (Accessed 17/06/2020).

44) Alterio, V.; Cadoni, R.; Esposito, D.; Vullo, D.; Fiore, A. D.;

(62) Cummings, W. M.; Cox, C. H.; Snyder, H. R. Arylboronic

Acids. Medium-Size Ring-Containing Boronic Ester Groups. J. Org.

Chem. 1969, 34, 1669−1674.

Monti, S. M.; Caporale, A.; Ruvo, M.; Sechi, M.; Dumy, P.; Supuran,

C. T.; Simone, G. D.; Winum, J.-Y. Benzoxaborole as a New

Chemotype for Carbonic Anhydrase Inhibition. Chem. Commun.

(63) Vshyvenko, S.; Clapson, M. L.; Suzuki, I.; Hall, D. G.

Characterization of the Dynamic Equilibrium between Closed and

Open Forms of the Benzoxaborole Pharmacophore. ACS Med. Chem.

Lett. 2016, 7, 1097−1101.

(

016, 52, 11983−11986.

45) Winn, M. D.; Ballard, C. C.; Cowtan, K. D.; Dodson, E. J.;

Emsley, P.; Evans, P. R.; Keegan, R. M.; Krissinel, E. B.; Leslie, A. G.

W.; McCoy, A.; McNicholas, S. J.; Murshudov, G. N.; Pannu, N. S.;

Potterton, E. A.; Powell, H. R.; Read, R. J.; Vagin, A.; Wilson, K. S.

(64) Dunbrack, R. L.; Karplus, M. Conformational Analysis of the

Backbone-Dependent Rotamer Preferences of Protein Sidechains.

Nat. Struct. Biol. 1994, 1, 334−340.

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Prince, D. B.; Breen, J.; Giacobbe, R. A.; Lahiri, S.; Verheijen, J. C.

Article

(

65) McKinney, D. C.; Zhou, F.; Eyermann, C. J.; Ferguson, A. D.;

,5-Disubstituted 6-Aryloxy-1,3-Dihydrobenzo[ c ][1,2]Oxaboroles

Are Broad-Spectrum Serine β -Lactamase Inhibitors. ACS Infect. Dis.

015, 1, 310−316.

66) Sainsbury, S.; Bird, L.; Rao, V.; Shepherd, S. M.; Stuart, D. I.;

Anderson, J.; Kozikowski, A. P. Rational Design, Synthesis, and

Biological Evaluation of Rigid Pyrrolidone Analogues as Potential

Inhibitors of Prostate Cancer Cell Growth. Bioorg. Med. Chem. Lett.

2001, 11, 955−959.

(82) Ku, A. F.; Cuny, G. D. Access to 6a-Alkyl Aporphines:

Synthesis of (±)- N -Methylguattescidine. J. Org. Chem. 2016, 81,

10062−10070.

(

Hunter, W. N.; Owens, R. J.; Ren, J. Crystal Structures of Penicillin-

Binding Protein 3 from Pseudomonas Aeruginosa: Comparison of

Native and Antibiotic-Bound Forms. J. Mol. Biol. 2011, 405, 173−184.

(

67) van Berkel, S. S.; Nettleship, J. E.; Leung, I. K. H.; Brem, J.;

Choi, H.; Stuart, D. I.; Claridge, T. D. W.; McDonough, M. A.;

Owens, R. J.; Ren, J.; Schofield, C. J. Binding of (5 S )-Penicilloic Acid

to Penicillin Binding Protein 3. ACS Chem. Biol. 2013, 8, 2112−2116.

68) Ren, J.; Nettleship, J. E.; Males, A.; Stuart, D. I.; Owens, R. J.

Crystal Structures of Penicillin-Binding Protein 3 in Complexes with

Azlocillin and Cefoperazone in Both Acylated and Deacylated Forms.

FEBS Lett. 2016, 590, 288−297.

(

69) Sacco, M. D.; Kroeck, K. G.; Kemp, M. T.; Zhang, X.; Andrews,

L. D.; Chen, Y. Influence of the α -Methoxy Group on the Reaction of

Temocillin with Pseudomonas Aeruginosa PBP3 and CTX-M-14 β -

Lactamase. Antimicrob. Agents Chemother. 2019, 64, e01473 −19.

70) Transue, T. R.; Gabel, S. A.; London, R. E. NMR and

Crystallographic Characterization of Adventitious Borate Binding by

Trypsin. Bioconjugate Chem. 2006, 17, 300−308.

(

71) Stoll, V. S.; Eger, B. T.; Hynes, R. C.; Martichonok, V.; Jones, J.

B.; Pai, E. F. Differences in Binding Modes of Enantiomers of 1-

Acetamido Boronic Acid Based Protease Inhibitors: Crystal Structures

of γ -Chymotrypsin and Subtilisin Carlsberg Complexes. Biochemistry

(

998, 37, 451−462.

72) Tickle, I. J.; Flensburg, C.; Keller, P.; Paciorek, W.; Sharﬀ, A.;

Vonrhein, C.; Bricogne, G. STARANISO; Global Phasing Ltd.:

Cambridge, United Kingdom, 2018.

(

73) Vonrhein, C.; Flensburg, C.; Keller, P.; Sharff, A.; Smart, O.;

Paciorek, W.; Womack, T.; Bricogne, G. Data Processing and Analysis

with the AutoPROC Toolbox. Acta Crystallogr. Sect. D Biol.

Crystallogr. 2011, 67, 293−302.

(

74) McCoy, A. J.; Grosse-Kunstleve, R. W.; Adams, P. D.; Winn, M.

D.; Storoni, L. C.; Read, R. J. Phaser Crystallographic Software. J.

Appl. Crystallogr. 2007, 40, 658−674.

McNicholas, S.; Long, F.; Murshudov, G. N. REFMAC 5 Dictionary:

75) Emsley, P.; Lohkamp, B.; Scott, W. G.; Cowtan, K. Features

and Development of Coot. Acta Crystallogr. Sect. D Biol. Crystallogr.

010, 66, 486−501.

76) Vagin, A. A.; Steiner, R. A.; Lebedev, A. A.; Potterton, L.;

(

Organization of Prior Chemical Knowledge and Guidelines for Its

Use. Acta Crystallogr. Sect. D Biol. Crystallogr. 2004, 60, 2184−2195.

(

77) Adams, P. D.; Afonine, P. V.; Bunkóczi, G.; Chen, V. B.; Davis,

I. W.; Echols, N.; Headd, J. J.; Hung, L.-W.; Kapral, G. J.; Grosse-

Kunstleve, R. W.; McCoy, A. J.; Moriarty, N. W.; Oeffner, R.; Read, R.

J.; Richardson, D. C.; Richardson, J. S.; Terwilliger, T. C.; Zwart, P.

H. PHENIX: A Comprehensive Python-Based System for Macro-

molecular Structure Solution. Acta Crystallogr. Sect. D Biol. Crystallogr.

(

010, 66, 213−221.

78) Williams, C. J.; Headd, J. J.; Moriarty, N. W.; Prisant, M. G.;

Videau, L. L.; Deis, L. N.; Verma, V.; Keedy, D. A.; Hintze, B. J.;

Chen, V. B.; Jain, S.; Lewis, S. M.; Arendall, W. B.; Snoeyink, J.;

Adams, P. D.; Lovell, S. C.; Richardson, J. S.; Richardson, D. C.

MolProbity: More and Better Reference Data for Improved All-Atom

Structure Validation: PROTEIN SCIENCE.ORG. Protein Sci. 2018,

79) McNicholas, S.; Potterton, E.; Wilson, K. S.; Noble, M. E. M.

7, 293−315.

Presenting Your Structures: The CCP 4 Mg Molecular-Graphics

Software. Acta Crystallogr. Sect. D Biol. Crystallogr. 2011, 67, 386−394.

80) Simon, J. F.; Bouillez, A.; Frere, J.-M.; Luxen, A.; Zervosen, A.

Synthesis of an Enantiopure Thioester as Key Substrate for Screening

the Sensitivity of Penicillin Binding Proteins to Inhibitors. Arkivoc

(

016, 2016, 22−31.

81) Qiao, L.; Zhao, L.-Y.; Rong, S.-B.; Wu, X.-W.; Wang, S.; Fujii,

T.; Kazanietz, M. G.; Rauser, L.; Savage, J.; Roth, B. L.; Flippen-