Download as pdf or txt
Download as pdf or txt
You are on page 1of 22

pubs.acs.

org/jmc Perspective

Targeting Metalloenzymes by Boron-Containing Metal-Binding


Pharmacophores
You-Cai Xiao,* Jun-Lin Yu, Qing-Qing Dai, Gen Li, and Guo-Bo Li*

Cite This: J. Med. Chem. 2021, 64, 17706−17727 Read Online

ACCESS Metrics & More Article Recommendations


See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

ABSTRACT: Metalloenzymes have critical roles in a wide range of


biological processes and are directly involved in many human
diseases; hence, they are considered as important targets for
Downloaded via NIPER KOLKATA on November 3, 2023 at 07:00:41 (UTC).

therapeutic intervention. The specific characteristics of metal


ion(s)-containing active sites make exploitation of metal-binding
pharmacophores (MBPs) critical to inhibitor development targeting
metalloenzymes. This Perspective focuses on boron-containing
MBPs, which display unique binding modes with metalloenzyme
active sites, particularly via mimicking native substrates or tetrahedral
transition states. The design concepts regarding boron-containing
MBPs are highlighted through the case analyses on five distinct
classes of clinically relevant nucleophilic metalloenzymes from
medicinal chemistry perspectives. The challenges (e.g., selectivity)
faced by some boron-containing MBPs and possible strategies (e.g., bioisosteres) for metalloenzyme inhibitor transformation are also
discussed.

■ INTRODUCTION
Metalloenzymes are a broad set of enzymes that use active site
fluconazole is positioned to make coordination with the heme-
iron of Saccharomyces cerevisiae CYP51 (ScCYP51) with the
metal ion(s) for catalysis; more strictly, metalloenzyme active triazole motif (Figure 1A). Several other representative MBPs
site metal ion(s) need to form stable coordination with active such as sulfonamide and hydroxamic acid often exist in approved
site residues and be involved in related chemical reactions.1−3 drugs (Figure 1A).14−16 MBPs are important for target binding
The metal ion(s) usually acts as a Lewis acid or redox partner affinity but are associated with metalloenzyme selectivity; a
and facilitates the catalyzed reaction by raising the ground state rational design or selection of proper MBPs is often the focal
energy of reaction partners, enhancing substrate binding or point for metalloenzyme inhibitor development. Our previous
stabilizing reactive intermediates.4 Metalloenzymes cover a wide analyses of metalloenzyme-ligand interactions yielded a diverse
range of enzyme types including oxidoreductases, hydrolases, set of 469 curated MBPs,8 which may help provide a wealth of
transferases, lyases, isomerases and ligases, and are involved in ideas for inhibitor design and MBP expansion. Of these curated
various physiological processes, e.g., metabolism and immune MBPs, some have been considered as privileged moieties, e.g.,
modulation. By data-mining analyses of the Protein Data Bank hydroxamic acid and imidazole, coupling in several clinically
(PDB), we found that there are more than 3300 structurally useful drugs, while some have not been widely explored such as
validated metalloenzymes. More importantly, a great number of cyano and oxirane groups. Deep understanding of more
metalloenzymes have been directly associated with various biocompatible, privileged MBPs is hence critical for drug
human illnesses, such as neoplastic, neurodegenerative, discovery targeting metalloenzymes.
inflammatory, cardiovascular, and infectious diseases, thereby The boron-containing MBPs are of particular interest to
representing a rich target space for drug discovery and researchers in developing new agents targeting metalloenzymes.
development.5 On the one hand, the boron atom with a vacant p orbital is
To date, about 70 small-molecule inhibitors for metal-
loenzymes have been approved for clinical use.5,6 Most of
these inhibitors bind to make direct coordination interactions Received: September 28, 2021
with the active site metal ion(s) by specific chemotypes, Published: December 8, 2021
commonly termed as metal-binding pharmacophores
(MBPs).7−12 For example, the antihypertensive drug captopril
binds to coordinate the two zinc ions of angiotensin converting
enzyme (ACE) with its thiol group;13 the antifungal drug

© 2021 American Chemical Society https://doi.org/10.1021/acs.jmedchem.1c01691


17706 J. Med. Chem. 2021, 64, 17706−17727
Journal of Medicinal Chemistry pubs.acs.org/jmc Perspective

Figure 1. (A) Representative metal-binding pharmacophores (MBPs) of FDA-approved drugs targeting metalloenzymes.13−17 (B) Boronic acid and
its diverse binding modes with protein targets, including metalloenzymes.

unique in its chameleonic capacity to interact with various promising targets for several human diseases, including heart
protein targets (Figure 1B),18 including relatively challenging failure, depression, asthma, inflammation, cancer, and so on.27,29
metalloenzymes for their flexibility, dynamics, and electronic The PDE family includes 11 isoenzyme classes. Taking PDE4
structure of metal ion(s)-containing active sites in metal−ligand isoenzyme, for example, it is encoded by four genes (PDE4A,
interactions.19 On the other hand, the boron-containing MBPs PDE4B, PDE4C, and PDE4D) that widely exist in the brain,
have diverse interaction modes with active site metal ion(s) leukocytes, and bones.30 These PDE4 enzymes are important
(Figure 1B); some have been clinically validated, e.g., the regulators for inflammation and have been the focus of extensive
benzoxaborole of Crisaborole (a drug approved for psoriasis drug development projects for the treatment of autoimmune
treatment in 2014) as an MBP to inhibit phosphodiesterase 4 by diseases.31−33 PDE4 consists of an N-terminal regulatory
reacting with the active site metal ion-activated water domain, a C-terminal domain, and a conserved catalytic domain
molecule.20 Besides, boron-containing MBPs are not prone to that lies at the interface of these two domains; each PDE4
being metal chelators and usually do not promiscuously destroy isoform has a unique N-terminal region (Figure 2A). The
metalloenzyme active sites. Moreover, the therapeutic potentials catalytic domain of PDE4 contains two metal ions; usually, one
of boron-containing MBP compounds remain to be ex- metal ion is Zn2+, and the other is Mn2+, Mg2+, Zn2+, or Ca2+.34
ploited.21−26 We herein provide an overview on boron- The two metal ions are bridged by an activated hydroxide, which
containing MBPs targeting five distinct types of clinically is likely to perform a nucleophilic attack on the phosphorus of
relevant nucleophilic metalloenzymes and their common cAMP, triggering the cAMP hydrolysis (Figure 2B). The
inhibition principles, with the hope to aid metalloenzyme hydrolysis product AMP was observed by crystallographic
inhibitor transformation and boost drug discovery targeting analysis to have coordination interactions with the active site
metalloenzymes.


metal ions (Figure 2C), offering a structural basis for the
hydrolytic mechanism and inhibitor design.
TARGETING HUMAN PHOSPHODIESTERASES Boron-containing compounds were first reported as PDE4
We describe work on boron-containing MBPs targeting inhibitors by Akama et al. in 2009.35 They found that 5-
metalloenzymes, beginning with disease-related phosphodies- phenoxybenzoxaboroles were particularly active against PDE4;
terases (PDEs). PDEs are a family of enzymes that are for example, compound 1 (Crisaborole, AN2728)) showed an
responsible for regulating the intracellular level of the second IC50 of 110 nM to PDE4 (Figure 3A). Crisaborole also has
messenger molecules cAMP and cGMP.27,28 PDEs are inhibitory activity against PDE1A3, PDE3C, and PDE7AZ but
17707 https://doi.org/10.1021/acs.jmedchem.1c01691
J. Med. Chem. 2021, 64, 17706−17727
Journal of Medicinal Chemistry pubs.acs.org/jmc Perspective

Figure 2. Overall architecture and catalytic mechanism of PDE4. (A) Domain organization of PDE4. (B) Chemical mechanism of PDE-mediated
cAMP hydrolysis. (C) View of a crystal structure of PDE4B complexed with AMP (PDB ID 1ROR),34 showing the binding features of AMP with metal
ions and catalytically important residues.

has no or very low inhibition to other PDE isozymes at 10 μM. compounds 3 and 4 showed 3- and 100-fold improvement
Crisaborole was approved by FDA for treatment of atopic activities compared with 1, respectively (Figure 4).20 The acetyl
dermatitis in 2014. The inhibitory activity could be increased to group of 4 may have additional hydrogen bonding interactions
60 nM by introducing another cyano group at the adjacent with Gln443, affording approximately 30-fold potency improve-
position (2, AN2898).20 Crystallographic analyses of the ment over 3. The tetrahedral character of the sp3 hybridized
PDE4:2 complex (PDB ID 3O0J)20 revealed that the boronate is essential for PDE4 inhibition, which was confirmed
benzoxaborole motif of 2 binds to interact with the dimetal by the observation that compound 5 (AN2913) lacking a boron
center, receiving the nucleophilic attack by the bridging atom has weak PDE4 inhibition (Figure 4).
hydroxide to form the sp3 hybridized boronate form. All the Zhang et al. proposed a soft-drug strategy for the design of
three boronate oxygen atoms are involved in coordination with benzoxaborole PDE4 inhibitors, with the aim to improve the
the dimetal ions, forming a classical trigonal-pyramidal geometry therapeutic index.36 Changing the cyano-substituted benzox-
(Figure 3B). The 3,4-dicyanophenoxy ring makes π−π aborole 6 (IC50 = 180 nM) to an ester group led to a more potent
interactions with Phe446, partly contributing to stabilize the inhibitor 7 (IC50 = 47 nM), whereas further hydrolysis to the
specific coordination mode. More interestingly, 2 was observed carboxylic acid form (8) yielded much less PDE4 inhibition
to have a similar binding mode as that of AMP, i.e., sp3 (IC50 > 10 μM) (Figure 4). In this case, the ester compound 7
hybridized boronate and phosphate making a highly similar exerts its therapeutic effects in the target area of the skin, and
coordination manner to the active site metal ions and 3,4- interestingly, it will be converted into inactive inhibitor (8)
dicyanophenoxy and adenosine making π−π interactions with when penetrating the skin and reaching the circulation system
Phe446, reflecting that 2 is likely to mimic the binding nature of (e.g., hydrolyzed by plasma esterases). The inhibitor 7 was
AMP (Figure 3C). This comparison also led to a conclusion that observed to have good skin bioavailability and stability, broad in
the sp3 hybridized boronate and phosphate are bioisostere vivo anti-inflammatory activity, and improved safety properties
MBPs, which can be used for related structural transformation with less systemic side effects.36 Although this strategy is able to
and drug design. decrease PDE inhibition in other tissues, it could not exclude the
Extended medicinal chemistry efforts were carried out for this possibility of this benzoxaborole series inhibiting other targets
benzoxaborole series to improve the inhibitory potency or other other than PDEs.
features, such as selectivity and toxicity. When introducing a side Recently, Xu’s group designed various bicyclic substituents at
chain onto 1 to occupy the PDE4 adenine-binding pocket the benzoxaborole 5-position to occupy the adenine pocket and
mainly formed by the side chains of Phe446, IIe410, and Gln443, discovered several potent PDE4 inhibitors. For example,
17708 https://doi.org/10.1021/acs.jmedchem.1c01691
J. Med. Chem. 2021, 64, 17706−17727
Journal of Medicinal Chemistry pubs.acs.org/jmc Perspective

Figure 3. Inhibition mode of benzoxaboroles to PDE4. (A) Chemical structures of 1 (AN2728) and 2 (AN2898). (B) View of a crystal structure of
PDE4:2 complex (PDB ID 3O0J)20 reveals a sp3 hybridized boronate as an MBP to coordinate with the two active site metal ions. (C) Superimposition
of PDE4:2 and PDE4:AMP (PDB ID 1ROR)34 revealing that 2 is likely to mimic the nature of AMP binding to PDE4.

Figure 4. Structural transformation of benzoxaborole PDE4 inhibitors.

Figure 5. Representative oxaborole PDE4 inhibitors.

compound 9 displayed 136-fold improved inhibitory activity induced mouse atopic dermatitis model. Moreover, 9 displayed
(IC50 = 0.42 nM) compared with Crisaborole (Figure 4).37 In a favorable safety property in repeated oral dose toxicity and
addition, 9 exhibited high selectivity over PDE3A, PDE3B, and phototoxicity studies, which is a good drug candidate worthy of
PDE10A. In vivo studies indicated that 9 manifested greater further clinical investigation.
efficacy than Crisaborole at the same dosage (P < 0.05) both in Interestingly, a series of novel 4-biaryl-substituted oxaboroles
the PMA-induced mouse ear edema model and the calcipotriol- were reported recently as highly potent PDE4 inhibitors (Figure
17709 https://doi.org/10.1021/acs.jmedchem.1c01691
J. Med. Chem. 2021, 64, 17706−17727
Journal of Medicinal Chemistry pubs.acs.org/jmc Perspective

Figure 6. Overall architecture and catalytic mechanism of ATX. (A) Domain organization of ATX. (B) ATX-medicated hydrolysis of LPC to LPA. (C)
A view of the mouse ATX:LPA complex structure (PDB ID 3NKM)41 revealing binding features of LPA to mATX active site zinc ions and long and
narrow subpockets.

5).38 The representative oxaboroles 10, 11, and 12 showed IC50 LPA1-6, to evoke a wide range of cellular responses, including
values of 0.05, 0.12, and 0.24 nM, respectively. These cell proliferation, survival, differentiation, and migration.
compounds could modulate inflammatory cytokine levels and As shown in Figure 6A, ATX consists of four domains, and the
could be used for treating inflammatory disorders such as atopic core catalytic domain has an α/β fold and binds two Zn2+ ions.41
dermatitis, arthritis asthma, and fibrosis (details regarding in vivo One Zn2+ ion is coordinated by His315, His474, and Asp311, and
studies are not disclosed). the other Zn2+ is coordinated by Asp171, Thr209, and His359. The
Compared with other MBPs (e.g., pyridazin-3-ol), the proposed catalysis mechanism for LPC hydrolysis by ATX
benzoxaboroles or oxaboroles are promising MBPs targeting involves a nucleophilic attack on the LPC phosphorus atom by
PDE4 and can uniquely mimic the AMP phosphate binding with Thr209 oxygen anion, resulting in a choline and a pentacovalent
PDE4 active site metal ions.20 Although almost all PDE reaction intermediate, and the latter undergoes a water-
isoenzymes have the dimetal active site, the benzoxaboroles mediated hydrolysis to give the product LPA (Figure 6B).42
manifest selectivity for PDE4, implying that the target selectivity LPA was observed to bind with mATX active site zinc ions by its
is not solely dependent on MBPs, but rather the entire phosphate group,41 providing insights for inhibitor design
compound chemical structure, particularly with respect to (Figure 6C).
interactions with unique N-terminal regions of PDE isoenzymes. Since ATX is highly associated with several human diseases
Notably, MBPs-containing inhibitors are not the only manner including cancer, great efforts have been made for the
for targeting PDE4; several non-MBP inhibitors have been development of new ATX inhibitors for therapeutic inter-
developed as highly potent, selective PDE4 inhibitors,8,28 and vention.43−45 Boron-containing compounds have played an
those, including Roflumilast, Cilomilast, and Apremilast, have important role in the early stage of ATX inhibitor development.
been approved for clinical uses. In 2010, Ovaa’s group campaigned a high-throughput screening


of 40 000 molecules and identified the thiazolidinedione 13 as a
selective ATX inhibitor (Figure 7A).46,47 Inspired by the case of
TARGETING HUMAN AUTOTAXIN
the proteasome inhibitor bortezomib which uses boronic acid to
Another example of using boronic acid as an MBP to mimic the react with N-terminal threonine oxygen nucleophile,48 they
substrate phosphate group is human autotaxin (ATX) replaced the carboxylic acid of 13 to boronic acid (14), leading
inhibitors. ATX belongs to the nucleotide pyrophosphatase to a significant increase in ATX inhibition; shifting boronic acid
and phosphodiesterase (NPP) family, and it is the only family from the meta-position to para-position (15) further improved
member that has lysophospholipase D activity, i.e., responsible the inhibitory activity to the single-digit nanomolar range (IC50
for the hydrolysis of lysophosphatidyl choline (LPC) to the = 5.67 nM), while to the ortho-position greatly decreased the
bioactive lysophosphatidic acid (LPA).39,40 LPA can activate the activity to an IC50 larger than 5 μM (Figure 7A). The 3-
LPA-specific G protein-coupled receptors (GPCRs), including benzyloxyboronic acid analogues 17 and 18 showed no
17710 https://doi.org/10.1021/acs.jmedchem.1c01691
J. Med. Chem. 2021, 64, 17706−17727
Journal of Medicinal Chemistry pubs.acs.org/jmc Perspective

Figure 7. Thiazolidinedione-boronic acid based ATX inhibitors. (A) Key structure−activity relationship for thiazolidinedione-boronic acid ATX
inhibitors. (B) View of the ATX:15 complex structure (PDB ID 2XRG)49 revealed the binding mode of 15, explaining the structure−activity
relationship shown in part A. (C) Comparison analyses reveals that 15 likely mimics the reaction intermediate, and boronate and phosphate groups are
high-quality MBP bioisosteres.

improvement in the inhibitory activity compared with 15 On the basis of the ATX:15 complex structure, a series of
(Figure 7A). Co-crystallography revealed that the boronic acid imidazolidine-2,4-dione based boronic acid inhibitors were
of 15 (PDB ID 2XRG) is transformed to a sp3 hybridized further developed (Figure 8).49 The representative compound
boronate by forming a covalent bond with the nucleophilic 19 showed similar inhibitory activity against ATX compared
hydroxyl group of Thr 209 and one boronate hydroxyl
with 15. Interestingly, the Z and E isomers of 19 showed similar
coordinated with the active site zinc ions, which are essential
for ATX catalytic activity (Figure 7B).49 This is unusual but is a nanomolar potency against ATX (Figure 8), reflecting the
particularly interesting binding mode for MBPs, reflecting the considerable tolerance of the ATX active site to accommodate
specific features of boronate to both nucleophilic residues and structurally distinct inhibitors.
metal ions. Further superimposition of ATX:15 and ATX:LPA
revealed that 15 perfectly mimics the binding mode of the
substrate reaction intermediate, e.g., sp3 hybridized boronate of
15 and phosphate of the substrate intermediate positioned to
form similar covalent bonding with Thr209, similar coordination
interactions to Zn1, and similar hydrogen bonding interactions
with Asn230 (Figures 6C and 7C). This comparison highlights
that the substrate or transition state-mimicking strategy is highly
effective to identify potent inhibitors and further reveals that the
boronate (in sp3 form) and phosphate groups are high-quality
bioisosteres, particularly in developing metalloenzyme inhib-
itors. Figure 8. Imidazolidine-2,4-dione-boronic acid ATX inhibitors.

17711 https://doi.org/10.1021/acs.jmedchem.1c01691
J. Med. Chem. 2021, 64, 17706−17727
Journal of Medicinal Chemistry pubs.acs.org/jmc Perspective

Figure 9. Thiazolinone-boronic acid ATX inhibitors.

In 2013, Nagano’s group developed a sensitive and specific responsible for converting L-arginine into ornithine and urea.58
ATX fluorescence probe, TG-mTMP, and used it to screen ATX The high expression or unusual activity of Arg I and Arg II has
inhibitors from a large chemical library.50 They identified the been observed in various pathological states, e.g., cancer,
thiazolinone 20 as a potent ATX inhibitor with an IC50 value of atherosclerosis, renal disease, inflammatory disease, and neuro-
180 nM (Figure 9). Substitution on the piperazine nitrogen of degenerative disease;59 in particular, recently both isoforms have
20 with benzylcarbonate and benzylboronate afforded 21 and been identified as major immunosuppressive players,60,61
22, respectively; both compounds showed improved inhibitory suggesting that they are promising targets for immunologic
potency. Compound 22 had promising pharmacologically therapy and related disease treatment.
related features in subsequent in vivo studies. Arg I and Arg II are typically homotrimeric enzymes stabilized
Recently, a new class of ATX inhibitors that use by two Mn2+ ions per monomeric structure, with an α/β fold
benzoxaborole as an MBP were reported (Figure 10).51 The consisting of an eight-strand β-sheet surrounded by several α-
helixes (Figure 11).62 Both arginase isoforms contain a dinuclear
Mn2+ active site. The main feature is the active site Mn2+ ions
bridged by an activated hydroxide, which is responsible for
substrate hydrolysis. The activated hydroxide is likely to attack
the guanidyl carbon of L-arginine, leading to production of L-
ornithine and urea, as observed by crystallography (Figure 11).
According to the active site features of arginases, the boron-
containing chainlike MBP was introduced to bind with
arginases, and several generations of boron-containing inhibitors
have been evolved, which are discussed as follows.64−66
Figure 10. Benzoxaborole ATX inhibitor. Christianson’s lab pioneered the use of boron-containing MBP
to inhibit arginases by mimicking the proposed “tetrahedral”
most potent compound 23 manifests IC50 of 0.13 μM to ATX, substrate transition state and reported the first boron-containing
comparable to 15 (IC50 = 0.088 μM) determined in the choline arginase inhibitor, 2-(S)-amino-6-boronohexanoic acid (24,
release assays. Notably, 23 has improved druggable properties, ABH) (Figure 12A).67 ABH acted as a “suicide” substrate and
including favorable plasma protein binding feature, pharmaco- showed a slow binding feature with an IC50 value of 1.45 μM to
kinetics, solubility, and rat/human plasma stability.51 Since ATX human arginase I (hArg I).68 The X-ray crystal structure of the
is a promising drug target, many medicinal chemistry efforts hArg I:ABH complex (PDB ID 2AEB)69 reveals that the boronic
have been involved in seeking potential drug candidates, partly acid of ABH undergoes nucleophilic attack by the hydroxide ion
including developing different MBPs, e.g., phosphate, phospho- that is activated by binuclear Mn2+ to yield a tetrahedral sp3
thioate, benzoxazolidinone, benoxazolone, and carboxylic acid, hybridized boronate form, which likely mimics the L-arginine
in addition to boronic acid.52 Further studies of boron- tetrahedral intermediates during the hydrolytic process (Figure
containing MBPs are needed to consider its in vivo efficacy 12A). The observed hydrogen bonding networks reflect the high
particularly in developing drug candidates for ATX-driven selectivity nature of arginase in recognition of the L-arginine
disease models. substrate as well as the substrate-mimicking inhibitor (e.g.,

■ TARGETING HUMAN ARGINASES


The boron-containing MBPs are also reported to target
ABH) (Figure 12B).
The success of boron-containing MBP inspired researchers to
develop new MBPs to interact with the arginase active site Mn2+
arginases, a class of dinuclear metalloenzymes, which use the ions (Figure 13). Shin et al. used sulfonamide as an MBP, which
nucleophilic hydroxyl for catalysis, similar to PDEs and ATX. is structurally similar to sp3 hybridized boronate; the obtained
Arginases are Mn2+-dependent metalloenzymes initially de- compound 25 showed inhibition to Arg I with a Ki value of 90
scribed by Kossel and Dankin in 1904;53 two isoforms of μM, significantly weaker than ABH (Figure 13). They later
arginases in mammals, arginase I (Arg I) and arginase II (Arg II), identified that using aldehyde as the MBP, i.e., 26, could increase
have been known to date.54,55 Both isoforms have similar kinetic Arg I inhibition (Figure 13). Interestingly, the rArg I:26 complex
and biochemical properties but differ from their quantitative structure (PDB ID 1T5F)70 revealed that the aldehyde of 26 is
requirement of Mn2+ for catalysis and subcellular localization. likely attacked by the hydroxide ion to form a gem-diol, similar
Arg I is a cytosolic enzyme primarily existing in hepatocytes, to that observed for boron-containing ABH, reflecting the
while Arg II is a mitochondrial enzyme expressed in small nucleophilic activity nature of the Arg I active site. Replacement
intestine, kidney, brain, monocytes, and macrophages.56,57 Arg I by a carboxylate or nitro group did not result in improved
and Arg II play critical roles in L-arginine metabolism, inhibitory activity to arginases (Figure 13). Notably, they
17712 https://doi.org/10.1021/acs.jmedchem.1c01691
J. Med. Chem. 2021, 64, 17706−17727
Journal of Medicinal Chemistry pubs.acs.org/jmc Perspective

Figure 11. Structural and catalytic mechanism of arginase-medicated L-arginine hydrolysis. The dotted-line box shows details of the binding modes of
the L-arginine hydrolytic products with Rattus norvegicus Arg I (rArg I) from a crystal structure of the rArg I:L-ornithine:urea complex (PDB ID
1HQG63).

expanded 2-aminoimidazole as the MBP to yield 27, which inhibitors, since it is solvent-exposed and oriented toward a
showed a Ki value of 4 μM to hArg I. The 2-aminoimidazole region of protein surface that is uncharted with regard to
MBP may have interactions with the carboxyl-rich, negatively inhibitor binding. To this end, a series of α,α-disubstituted ABH
charged active site of hArg I, thereby worthy of further analogues (33−38, Figure 15) were developed, which were
investigation not confined to hArg I. Despite these MBP called second-generation arginase inhibitors. Initial attempts to
bioisostere replacements, no more potent inhibitors were introduce simple alkyl substituents on the amino acid Cα atom,
obtained, suggesting that the boronic acid might be optimal such as CH3 and CHF2, led to 33 and 34;73 although these
MBP for arginases. Optimization for potency and pharmacoki- compounds did not show significantly enhanced binding affinity
netic properties for boronic acid containing compounds to to arginases, the work demonstrated that the arginase active site
target arginases is the main purposes in most medicinal can accommodate substituents on the ABH Cα-H position.
chemistry works. Later, by introducing larger substituents at the Cα-H position,
The early studies were mainly focused on the modification of the new ABH analoguess showed improved inhibitory activities.
the ABH alkyl chain (Figure 14).71,72 Replacement by sulfur in For example, compound 35 had IC50 values of 223 and 509 nM
the alkyl chain afforded 30, showing an IC50 of 5 μM to rat liver to hArg I and hArg II, respectively,68 and it had an oral
arginases (RLA). Introduction of a phenyl ring and a double bioavailability (F) of 28% and significantly reduced the infarct
bond in the alkyl chain yielded 31 and 32, respectively (Figure size in a rat model of myocardial ischemia/reperfusion injury.68
14), which did not have an improved inhibitory activity to RLA The hArg I:35 complex structure (PDB ID 4HWW)68 revealed
compared to ABH (rArg I IC50 = 0.8 μM) (Figure 14). that the Cα-H substituents occupied the hArg I solvent pocket
On the basis of the hArg I:ABH complex structure, the Cα-H without affecting the hydrogen bond formation between the
atom of ABH was considered to have the possibility for amino acid group and hArg I; additional hydrogen bonds were
structural modifications to obtain more potent arginase observed by a water-bridging the piperidine nitrogen atom with
17713 https://doi.org/10.1021/acs.jmedchem.1c01691
J. Med. Chem. 2021, 64, 17706−17727
Journal of Medicinal Chemistry pubs.acs.org/jmc Perspective

activity to arginases compared with ABH but still had poor oral
bioavailability.
Based on the second-generation Arg I inhibitors, Van Zandt’s
group developed a series of novel ring-constrained ABH
analogues, e.g., 41−43 (Figure 16A), which were called third-
generation arginase inhibitors.76 The representative pyrrolidine-
based compound 41 showed IC50 values of 110 and 440 nM to
hArg I and hArg II, respectively; the introduction of a
constrained pyrrolidine ring is able to reduce the conformational
entropy, and the pyrrolidine nitrogen has water molecule (W1)-
bridging hydrogen bonding interactions with Asp200 (Figure
16B). Further introducing a piperidine ring on the pyrrolidine
ring led to 42, which showed 40-fold increased inhibition
activities. The cocrystallography revealed that in addition to the
boronic acid involving coordination, the piperidine moiety of 42
is positioned to form direct electrostatic interactions with hArg
II Asp200 (Figure 16C). Further optimization of 42 generated 43,
which showed single-digit nanomolar activities against both
hArg I and hArg II.
Figure 12. Inhibition mode of ABH with hArg I. (A) Design of L- Very recently, further modifications of the monopyrrolidine
arginine-mimicking boron-containing chainlike MBP. (B) View of a ring to a fused 5,5-bicyclic ring system such as 44 and 45 were
crystal structure of hArg I:ABH (PDB ID 2AEB69) revealed the binding reported. Compound 44 had improved potency over ABH, with
mode of the sp3 hybridized boronate form of ABH with hArg I. an IC50 value of 29 nM to hArg I in a thioornithine generation
(TOGA) assay (Figure 17).77 X-ray crystallography revealed
that the 5,5-bicyclic pyrrolidine ring nitrogen atom of 44 makes
Asp181 and Asn183 (Figure 15). More interestingly, incorporation direct electrostatic interactions with Asp181 of hArg I (PDB ID
of a methyl on amine group (36) led to a nearly 4-fold increase in 6V7C).77 When dosed orally, 44 was demonstrated to have
potency compared with 35.68 This probably resulted from a gain serum arginase inhibition and concomitant arginine elevation in
in entropy due to the exclusion of a water molecule (W2) that a syngeneic mouse carcinoma model, albeit with low
was observed in the hArg I:35 structure (PDB ID 4HWW)68 but bioavailability (F = 7%, in mice). Greater oral bioavailability
not in the hArg I:36 structure (PDB ID 4HXQ)68 (Figure 15). was observed for 45 (F = 18%, in mice) with fluorinated
Further introduction of a two-carbon bridge in the piperidine pentalane bearing hydroxyl substitution (Figure 17).
ring led to 37 and 38, which showed improved activity against The nonexcellent pharmacokinetics, especially with low oral
both hArg I and hArg II.74 Unfortunately, these compounds had bioavailability and fast clearance of earlier inhibitors, prompted
low oral bioavailability in rats. Recently, the OncoArendi researchers to develop the fourth-generation arginase inhibitors.
Therapeutics company disclosed two new arginase inhibitors, 39 One of the successful methods to improve oral bioavailability
(with a sulfamido side chain) and 40 (with a guanidine side was based on the peptide transporter (PEPT1/2 transporter)
chain) (Figure 15).75 Both compounds had improved inhibitory targeted drug design. The peptide-based analogues with boronic

Figure 13. Several reported Arg I inhibitors with structurally different MBPs.

17714 https://doi.org/10.1021/acs.jmedchem.1c01691
J. Med. Chem. 2021, 64, 17706−17727
Journal of Medicinal Chemistry pubs.acs.org/jmc Perspective

Figure 14. Early research by modification of ABH.

Figure 15. Second-generation human arginase inhibitors and the binding modes of 35 (PDB ID 4HWW)68 and 36 (PDB ID 4HXQ)68 with hArg I.

acid moieties were reported by Calithera Biosciences.78 The nature of boron-containing MBP, i.e., an electron-deficient
representative inhibitors are compounds 46−48, which were boron atom, facilitates perfect interactions with arginase
obtained by coupling of the endocyclic nitrogen atom with dimanganese ions and catalytically important residues. The
various natural amino acids. Compound 46 (Numidargistat, CB- amino acid-mimicking strategy is particularly commendable,
1158) is a potent inhibitor for hArg I (IC50 = 86 nM) and hArg II which not only led to a drug candidate (i.e., Numidargistat) for
(IC50 = 296 nM) (Figure 18); the other two compounds, 47 and cancer treatment but also provided a new idea to address related
48 (without reported IC50 data), also showed good plasma-drug pharmacokinetic issues for boronic acids as well as other
concentrations (AUC). Numidargistat showed high oral chemotypes.


bioavailability in rats (52%) and mice (91%), moderated
clearance (10 mL/min/kg), and good tolerance in rats. Thus,
TARGETING CARBONIC ANHYDRASES
Numidargistat has been allowed for clinical trials in 2016 to treat
patients with metastatic solid tumors as monotherapy and in Boron-containing MBPs were also reported to target mono-
combination with chemo- and immunotherapy. nuclear metalloenzymes, for example, carbonic anhydrases
As summarized above, recent trends in developing new (CAs). CAs are ubiquitous zinc metalloenzymes, present in
arginase inhibitors have moved toward exploration of the core most living organisms (i.e., bacteria, archaea, and eukarya),
support scaffold rather than the boronic acid MBP. The unique which catalyze the reversible hydration of carbon dioxide and
17715 https://doi.org/10.1021/acs.jmedchem.1c01691
J. Med. Chem. 2021, 64, 17706−17727
Journal of Medicinal Chemistry pubs.acs.org/jmc Perspective

Figure 16. Third-generation human arginase inhibitors. (A) Representative third-generation human arginase inhibitors. (B) View of crystal structures
of hArg II:41 (PDB ID 6Q37)76 and hArg II:42 (PDB ID 6Q39)76 complexes reveals the common boron-involving interactions with manganese ions
and catalytically important residues such as Asp202, Asn149, and Ser156.

Figure 17. Bicyclic human arginase inhibitors.

Figure 18. Peptide-based human arginase inhibitor.

17716 https://doi.org/10.1021/acs.jmedchem.1c01691
J. Med. Chem. 2021, 64, 17706−17727
Journal of Medicinal Chemistry pubs.acs.org/jmc Perspective

Figure 19. Overall architecture and catalytic mechanism of hCA II: (A) Structure organization of hCA II (PDB ID 1CA2).86 (B) View of the hCA
II:HCO3− complex structure (PDB ID 2VVB)87 revealing the intermediate binding mode. (C) The proposed catalytic mechanism of hCA II-mediated
hydration of carbon dioxide and water to bicarbonate and protons.

Figure 20. Noncyclic boronic acid-based hCA inhibitors.

water to bicarbonate and protons (CO2 + H2O ⇌ HCO3− + consists of β-sheet with little α-helix content, and the active site
H+ ). 79,80 CAs have been classified into eight distinct contains a Zn2+ ion at the bottom of an ∼15 Å deep cleft, where
families,81,82 and all human CAs (hCAs) belong to the α- the Zn2+ is bound by three His residues and a water molecule in a
families. hCAs consist of 12 isoforms, including cytosolic hCA I, tetrahedral geometry form; the water molecule is further
hCA II, hCA III, hCA VII, and hCA XIII, membrane-bound stabilized by forming hydrogen bonding interactions to the
hCA IV, hCA IX, hCA XII, hCA XIV), mitochondrial (hCA VA
neighboring residue Thr199 that is important for catalysis (Figure
and hCA VB), and secretory hCA VI. These enzymes have been
extensively studied as potential therapeutic targets, e.g., for 19A).86 The proposed catalytic process for hCA II includes the
glaucoma, retinopathy, and hemolytic anemia cancer.83−85 nucleophilic attack of Zn2+-bound hydroxide ion to CO2,
We take hCA II as a canonical example for describing their producing a bicarbonate ion (HCO3−) by coordination with
structure and catalytic mechanism. The core structure of hCA II Zn2+ as observed by crystallography (Figure 19B), followed by
17717 https://doi.org/10.1021/acs.jmedchem.1c01691
J. Med. Chem. 2021, 64, 17706−17727
Journal of Medicinal Chemistry pubs.acs.org/jmc Perspective

attack of a water molecule to release the bicarbonate ion (Figure and interestingly, one oxygen atom makes additional hydrogen
19C).79 bonding interactions with Thr199 (Figure 21B), with a similar
Currently, the vast majority of hCA inhibitors approved by position as that of the Zn2+-bound hydroxide ion in hCA II
FDA use sulfonamide as an MBP to coordinate with the active active form (Figure 19), offering evidence to demonstrate the
site zinc ion, such as methzaolamide, dorzolamide, and advantage of benzoxaborole as an MBP engaging with the hCA
brinzolamide.83 In recent years, boronic acids have been II active sites.
reported as an alternative MBP to target hCAs. A series of In 2019, Larcher et al. reported the synthesis and character-
aromatic, arylalkenyl-, and arylalkyl boronic acids (e.g., 49−56) ization of a set of novel bis-benzoxaboroles from readily available
were reported as hCA inhibitors, mainly targeting the cytosolic 6-aminobenzoxaborole 58 (Figure 22).91 The bis-benzoxabor-
hCA I and hCA II, and the membrane-bound hCA IX and hCA
XII (Figure 20).88 In 2017, Supuran et al. reported that the
antitumor agent bortezomib showed micromolar inhibitory
activity to hCA isoforms, further offering insights for under-
standing the anticancer activity of bortezomib and suggesting
that it may be worth developing more potent boron-containing
hCA inhibitors for cancer treatment.89
Winum’s group reported a series of urea- and thiourea
substituted benzoxaboroles as a new chemotype for hCA Figure 22. Bis-benzoxaborole-based hCA inhibitors.
inhibition90 (Figure 21A). Among the reported benzoxaboroles,
oles could inhibit five hCA isoforms in the micromolar range,
including hCA I, hCA II, hCA IV, hCA IX, and hCA XII. The
bis-benzoxaborole 59 with a glutamic acid central core showed
the most potent inhibition against the tumor-associated isoform
hCA IX (Ki = 64 nM) compared with 6-aminobenzoxaborole 58
(Ki = 813 nM) (Figure 22).
The potential of CAs as anti-infective targets has also been
considered in recent years. Since the CA enzymes in the β-CA
genetic family are widespread and functionally important in
bacteria, fungi, and archaea, development of selective inhibitors
targeting β-CAs may lead to new anti-infective agents. The
above-mentioned boronic acids 49−56 and antitumor agent
bortezomib showed some inhibition but not strong activity to β-
CAs.92,93 In 2017, Winum’s group reported a series of 6-urea and
thiourea-substituted benzoxaboroles as β-CA inhibitors from
three pathogenic fungi (Can2 from Cryptococcus neoformans,
CgNce103 from Candida glabrata, and MgCA from Malassezia
globosa) (Figure 23).94 Most reported derivatives showed

Figure 23. Benzoxaborole-based β-CA inhibitors.


Figure 21. Benzoxaborole-based hCA inhibitors. (A) Representative
inhibitors 57 and 58. (B) View of the hCA II:57 complex structure
(PDB ID 4HWW)68 reveals the binding mode of 57 and the evidence of nanomolar inhibitory activities against Can2 and CgNce103
the advantage of benzoxaborole to interact with the hCA II active site. and micromolar inhibition against MgCA. Some benzoxaborole
derivatives, e.g., 60, showed selectivity to Can2 and CgNce over
compounds 57 (hCA I, Ki = 98 nM; hCA II, Ki = 89 nM; hCA IX hCAI/II. Recently, these types of benzoxaboroles were
Ki = 414 nM; hCA XII Ki = 69 nM) and 58 (hCA I, Ki = 417 nM; investigated for their inhibition against three CAs encoded by
hCA II, Ki = 1838 nM; hCA IX Ki = 93 nM; hCAX II Ki = 71 Vibrio cholera, including VchCAα, VchCAβ, and VchCAγ
nM) manifested potent inhibition to several hCAs. As observed (Figure 23).95 Compound 61 exhibited nanomolar inhibition
by X-ray crystallography, the benzoxaborole of 57 is likely to VchCAs, to a certain degree, with selectivity over hCA II,
attacked by hydroxide ion to form an sp3 hybridized state to suggesting that it is worthy of further optimization to develop
coordinate with the hCA II active site zinc ion (Figure 21B), new anti-infective agents.
similar to that described above for PDEs, ATX, and arginases. According to the CA active site features, boron-containing
The two oxygen atoms of sp3 hybridized benzoxaborole form a MBPs have the unique feature to interact with nucleophilic
trigonal bipyramidal coordination geometry with the zinc ion, water and have been used to successfully develop potent
17718 https://doi.org/10.1021/acs.jmedchem.1c01691
J. Med. Chem. 2021, 64, 17706−17727
Journal of Medicinal Chemistry pubs.acs.org/jmc Perspective

Figure 24. Proposed mechanism of MBL-mediated hydrolysis of carbapenem antibacterials.

inhibitors for human and pathogens’ CAs. Although current infection and ascending urinary tract infection models, which is
approved drugs are established based on the sulfonamide MBP, the first SBL/MBL dual-action inhibitor to enter clinical trials
boronic acids have shown great potential in inhibitor develop- (currently in phase III, ClinicalTrials.gov Identifier:
ment, which might provide an alternative way to identify new NCT03840148). Crystallographic studies revealed that the
treatments. dual SBL/MBL inhibition of 64 was achieved by mimicking the


common tetrahedral intermediates in the β-lactam hydrolysis
TARGETING BACTERIAL process (Figures 24 and 25).106 The CTX-M-15:64 (PDB ID
METALLO-β-LACTAMASES 6SP6)106 structure showed that the catalytic Ser70 is covalently
bound to the boron atom of 64, forming sp3 hybridized species.
In addition to the above-described human metalloenzymes, Similarly, the VIM-2:64 (PDB ID 6SP7)106 revealed that 64 was
boron-containing MBPs have been also used to target bacterial attacked by the zinc-bridged hydroxide anion to form sp3
metalloenzymes, e.g., metallo-β-lactamases (MBLs). The hybridized boronate, mimicking the tetrahedral intermediates.
production of MBLs in bacteria is considered as one of the Interestingly, recently, a cocrystal structure of NDM-1:64 (PDB
most important mechanisms of resistance to β-lactam ID 6RMF)107 revealed a tricyclic form of 64, i.e., the cyclization
antibacterials.96−99 A variety of MBLs have been identified in of the amide oxygen onto the boron atom, reflecting multiple
clinically isolated resistant bacteria, which were divided into possible binding modes for boronates in targeting metal-
Ambler classes B1, B2, and B3. The MBL enzymes utilize one or loenzymes.
two zinc ions to catalyze the hydrolysis of β-lactam antibacterials The above-mentioned bicyclic boronates have potent broad-
as summarized in Figure 24. Compared with serine β-lactamases spectrum inhibition to SBLs/MBLs and excellent druglike
(SBLs), MBLs have broader substrate specificity and can features, triggering efforts to identify more new bicyclic
hydrolyze penicillins, cephalosporins, and so-called “last-resort” boronates (Figure 26) with the aim of discovering better drug
carbapenems.100 MBLs also can destroy β-lactam SBL candidates. For example, introducing more rigid substituents at
inhibitors, including clavulanic acid, sulbactam, and tazobactam the 3-position (e.g., 65) or substituents at the 7-position (e.g.,
and even the non-β-lactam inhibitors, e.g., avibactam is slowly 66) of bicyclic boronates resulted in new patent chemotypes
hydrolyzed by some MBLs.100,101 The rapidly rising anti- with comparable inhibition to 64.108 Since the membrane
bacterial resistance prompted discovery of several broad- permeability for these bicyclic boronates is most likely related to
spectrum potent MBL inhibitors, including the typical their molecular size, it is desirable to identify new bicyclic
boronates.102,103 boronates with small molecular sizes but with potent SBL/MBL
The bicyclic boronate 62 (Figure 25) was initially reported by inhibitions (e.g., 67 and 68). In 2020, Qpex Biopharma reported
Ness et al. as a potent inhibitor to class A TEM1 (IC50 = 13 a very classic lead optimization case. They first investigated the
nM).104 Introduction of larger substituents (e.g., 63 and 64) at effects of different substituents at the 3- and 8-positions and
the 3-NH2 of bicyclic boronate resulted in improved inhibition discovered fluoroboronate 69 as inhibitors to SBLs only.
to both SBLs and MBLs (Figure 25).105 Notably, 64 exhibited Structural modifications led to 3-thiol analogues 70, which
nanomolar inhibition to a broad spectrum of SBLs and class B showed much improved MBL inhibition. The broader spectrum
MBLs (NDM-1, 10 nM; VIM-1, 7.9 nM; VIM-2, 0.5 nM; IMP-1, analogue 71 was obtained by removal of the 3-substituents.
2.5 μM) and could inhibit some PBPs.105 Compound 64 Moreover, further optimization resulted in a metabolically stable
showed high permeability to Gram-negative outer membrane tricyclic boronic acid QPX7728 (72), which displayed an
and periplasmic accumulation and showed in vivo activity ultrabroad spectrum of MBL/SBL inhibition, including class A
combined with cefepime in both the murine neutropenic lung KPC2 IC50 = 1.9 nM), class B NDM-1 (IC50 = 32 nM), class B
17719 https://doi.org/10.1021/acs.jmedchem.1c01691
J. Med. Chem. 2021, 64, 17706−17727
Journal of Medicinal Chemistry pubs.acs.org/jmc Perspective

Figure 25. Bicyclic boronate MBL/SBL inhibitors. (A) Representative bicyclic boronate inhibitors. Crystal structures of (B) CTX-M-15:64 (PDB ID
6SP6)106 and (C) VIM-2:64 (PDB ID 6SP7)106 revealed the dual-action inhibition by mimicking the common tetrahedral intermediates in the β-
lactam hydrolysis process.

Figure 26. Newly reported bicyclic boronate-based MBL/SBL inhibitors.

VIM-1 (IC50 = 8 nM), class C AmpC (IC50 = 8.5 nM), and class graphic studies revealed that 72 has a similar MBL inhibition
D OXA48 (IC50 = 0.28 nM), notably with little affects by porin mode as 64, e.g., the sp3 hybridized form coordinating with the
modification and the efflux pump (Figure 26).109 Crystallo- active site zinc ions. More interestingly, the cyclopropyl of 72 is
17720 https://doi.org/10.1021/acs.jmedchem.1c01691
J. Med. Chem. 2021, 64, 17706−17727
Journal of Medicinal Chemistry pubs.acs.org/jmc Perspective

positioned to occupy the small hydrophobic pocket near the zinc


ions in both NDM-1 and VIM-2, revealing new insights for drug
design and structural optimization for MBL inhibitors (Figure
27). As 72 has excellent pharmacokinetics in the rat similar to β-

Figure 28. Benzoxaborole-based β-lactamase inhibitors (IC50, μM).

Figure 27. View of crystal structures of 72 complexed with (A) NDM-1


(PDB ID 6V1M)109 and (B) VIM-2 (PDB ID 6V1P)109 reveals its
common binding features to NDM-1 and VIM-2.

lactam antibiotics and has good oral bioavailability without the


need for a prodrug (a common strategy to improve the
bioavailability of boronic acid-based inhibitors), this compound
has been advanced to phase I clinical trials (ClinicalTrials.gov
Identifiers: NCT04380207 and NCT05072444), providing a
new drug candidate for combating SBL/MBL-mediated
antibacterial resistance.
Recently, our group disclosed a series of 3-aryl substituted
benzoxaborole derivatives (e.g., 73 and 74) as dual SBL and
MBL inhibitors (Figure 28).110 Since such benzoxaborole
compounds do not have corresponding negatively charged
features to interact with the positively charged residue in the Figure 29. Benzo[b]thiophen-2-ylboronic acid-based MBL inhibitors.
MBL active site, which is important for recognition of β-lactam (A) Representative benzo[b]thiophen-2-ylboronic acids as NDM-1
antibiotics, we designed a series of C3-acrylic acid substituted inhibitors; (B) crystal structure of NDM-1:80 (PDB ID 6Q30)113
shows a similar inhibition mode as that described for cyclic boronate
benzoxaboroles by mimicking carbapenem antibiotics.111 Mean- MBL inhibitors.
while, in order to obtain the optical isomers, we presented an
asymmetric Morita−Baylis−Hillman (MBH) cascade reaction
and synthesized a series of optical C3 acrylic acid substituted showed good inhibition to class A KPC2 (IC50 = 4.9 μM), class
benzoxaboroles.111 Some of these compounds (e.g., 75 and 76) B NDM-1 (IC50 = 3.0 μM), class C AmpC (IC50 = 0.42 μM),
showed nanomolar inhibitory activities to KPC-2 SBL and and class D OXA24 (IC50 = 16 μM). Similar to above-described
micromolar inhibitory activity toward various class B1MBLs cyclic boronates, these compounds inhibit NDM-1 by
(Figure 28), providing a new starting point to develop coordinating with active site zinc ions using the sp3 hybridized
benzoxaborole-based MBL inhibitors. boronate form, as observed in the NDM-1:80 (PDB ID
In addition to the above-mentioned cyclic boronates, some 6Q30)113 crystal structure (Figure 29B).
acyclic boronates have also been investigated to target SBLs Recently, a series of acyclic boronic acid derivatives (e.g., 82
since the 1970s when boric acid (B(OH)3) was reported to and 83) were reported as dual-action MBL/SBL inhibitors by a
inhibit an B. cereus SBL.112 In 2019, Cendron et al. reported pharmacophore fusion strategy; for example, 83 has broad-
benzo[b]thiophen-2-ylboronic acids (e.g., 77−81, Figure 29A) spectrum inhibition to clinically relevant enzymes, including the
as inhibitors of several SBLs and MBLs.113 Compound 80 widespread KPC-2 (IC50 = 0.61 μM) and VIM-2 (IC50 = 0.44
17721 https://doi.org/10.1021/acs.jmedchem.1c01691
J. Med. Chem. 2021, 64, 17706−17727
Journal of Medicinal Chemistry pubs.acs.org/jmc Perspective

μM).114 As demonstrated by crystallography, the thiol of 82 is


positioned to coordinate with the active-site zinc ions of MBLs,
and the boronic acid is covalently bound with the catalytic serine
of SBLs (Figure 30), suggesting that pharmacophore fusion is

Figure 30. Acyclic boronic acid-based β-lactamase inhibitors.

Figure 31. Proposed bioisosteres of (A) sp2 and (B) sp3 hybridized
another useful strategy to develop dual-action inhibitors
boron MBPs.
particularly for structurally and mechanistically distinct targets.
Taken together, this metalloenzyme case analyses indicate that
development of proper boron-containing MBPs is quite selectivity, or achieving other purposes (e.g., intellectual
important for targeting MBLs or other metalloenzymes. property), its full potential remains largely unexploited. It is

■ BIOISOSTERES OF BORON-CONTAINING MBPS


The diverse metal coordination modes of boron-containing
highly desirable to exploit new MBP bioisosteres or their new
applications in metalloenzyme inhibitor development and,
particularly, to compare their respective advantages to
eventually promote drug discovery targeting metalloenzymes.


MBPs prompt researchers to consider bioisosteres in metal-
loenzyme inhibitor transformation. The MBP bioisosteres were
summarized with respect to the boron hybridization types. The CONCLUSION AND FUTURE PERSPECTIVES
sp2 boronic acid is possible to have different binding modes with The comprehensive analysis of various boron-containing MBPs
active site metal ions. For example, one hydroxyl group of sp2 targeting five distinct types of nucleophilic metalloenzymes
boronic acid could enable coordination with one or two metal clearly reveal their unique features in engaging with metal-
ions via a monodentate manner; the two hydroxyl groups of sp2 loenzyme active sites. Most notably, the boron-containing MBPs
boronic acid may be able to simultaneously form monodentate can receive a nucleophilic attack from an active site metal-
coordination with two metal ions or bidentate coordination with activated water molecule and enable their sp3 form to mimic the
one metal ion (Figure 1B). Several chemical motifs are proposed binding nature of native substrates or reaction transition states
to have similar metal binding features, such as carboxylic acid, (i.e., tetrahedral intermediates). For example, the benzoxaborole
nitrate, thioic acid, imidamide, imidazole, and tetrazole (Figure MBP was observed to mimic AMP binding with the PDE4 active
31A), which could be referred to as bioisosteres of sp2 site, and it was also successfully used to target ATX (e.g., 23),
hybridized boronic acid MBPs. Some of these bioisosteres CAs (e.g., 57−61), and MBLs (e.g., 73−76) by involving the
have been used in metalloenzyme inhibitor development, e.g., common sp3 boronate form to interact with these nucleophilic
the above-mentioned arginase inhibitors (Figure 13). More metalloenzymes. This offers the important basis for boron-
specifically, given the fact that the boron atom as a Lewis acid containing MBPs used in inhibitor transformation among
with a vacant p orbital could accept an electron from metal ion- structurally different metalloenzymes, and prompts them as
activated hydroxyl anion, the sp3 boronates display more bioisosteres to replace those, such as phosphoric acid and
attractive features in binding with metal ions, particularly via sulfonamide, to derive new metalloenzyme inhibitors.
forming more diverse metal coordination geometries (Figure Given that boron-containing MBPs have the possibility to
1B). Those chemical moieties, including sulfonamide, sulfonic interact with multiple nucleophilic metalloenzymes and other
acid, phosphoric acid, siliconic acid, and arsenic acid, could be nucleophilic enzymes (e.g., serine hydrolases), it is tempting to
used as bioisosteres of sp3 boronates since they have similar assume that such MBPs may face challenges such as selectivity in
tetrahedral pharmacophore features when binding with metal metalloenzyme inhibitor development. Some possible strategies
ions (Figure 31B). As highlighted above, the boronate and discussed below may be useful to address these challenges. The
phosphoric acid have been mutually used as bioisosteres in PDE first is to incorporate boron-containing MBPs into a target-
(Figure 3) and ATX (Figure 7) inhibitor transformation; the specific, substrate-mimicking chemical scaffold. As described
boronate, phosphoric acid, and sulfonamide have also been above, the boronic acid as an MBP was fused into arginine-
successfully applied as bioisosteres for MBL inhibitor develop- mimicking chemical scaffolds to develop arginase inhibitors by
ment.102,115 The unique advantage of these bioisosteres is that involving interactions with the active site Mn2+ ions and
they might be able to mimic native substrates or reaction catalytically important residues. The second is to introduce
transition states of nucleophilic metalloenzymes. Although such target-specific substituents close to boron-containing MBPs,
bioisostere-based strategy is useful in improving binding affinity, albeit with possibly high requirements for the chemical
17722 https://doi.org/10.1021/acs.jmedchem.1c01691
J. Med. Chem. 2021, 64, 17706−17727
Journal of Medicinal Chemistry pubs.acs.org/jmc Perspective

synthesis, to enhance binding affinity and metalloenzyme Gen Li − Key Laboratory of Drug-Targeting and Drug Delivery
specificity. The MBL inhibitor QPX-7728 is a classical example, System of the Education Ministry and Sichuan Province,
which has a magic cyclopropyl-fused cyclic boronate and Sichuan Engineering Laboratory for Plant-Sourced Drug and
manifests highly potent MBL inhibition. Coupling boron- Sichuan Research Center for Drug Precision Industrial
containing MBPs to target-specific peptide, antibody, or Technology, West China School of Pharmacy, Sichuan
aptamer could be an alternative to improve selectivity and University, Chengdu 610041, China
binding affinity to the metalloenzyme target. In addition, Complete contact information is available at:
emerging boronic ester prodrug strategies are likely to promote https://pubs.acs.org/10.1021/acs.jmedchem.1c01691
boron-containing drug development, including metalloenzyme
inhibitors. Some prodrugs have been validated in clinical Notes
settings, e.g., ixazomib citrate, which has improved pharmaco-
kinetics, pharmacodynamics, and therapeutic efficacy.116,117 The authors declare no competing financial interest.
The MBP bioisostere-based inhibitor transformation may be Biographies
also useful for addressing issues that are related to boron- You-Cai Xiao is an Associate Professor at the West China School of
containing inhibitor stability and other pharmaceutical proper- Pharmacy, Sichuan University. He completed his Bachelor’s degree in
ties. 2009 from Sichuan University and continued to pursue his Ph.D. in
Moreover, currently, boron-containing MBPs are mainly used Medicinal Chemistry at the same university. He has worked as a
to target nucleophilic metalloenzymes, and it remains underex-
Postdoctoral Fellow at KTH Royal Institute of Technology under the
plored how they target other types of metalloenzymes, which
supervision of Prof. Cristina Moberg between 2014 and 2015 and as
will be worthy of further study. The topic on how to exploit new
boron-containing MBPs and new MBP-metalloenzyme associ- Research Associate at The Wistar Institute in Prof. Joseph. M. Salvino’s
ations would also be worthwhile. Clearly, new computation tools group from 2015 to 2018. His research interests include the design and
such as metal ion-focused pharmacophore fitting,118 artificial synthesis of boron-based protease and metalloenzyme inhibitors,
intelligence algorithms,119 or new strategies120 are needed. proteolysis targeted chimera (PROTAC) technology, and asymmetric
Overall, this paper brings the knowledge of boron-containing reactions for drug synthesis.
MBPs targeting metalloenzymes and perspectives of how to Jun-Lin Yu is a postgraduate student, majoring in Medicine Chemistry
develop better metalloenzyme inhibitors. It is hoped that more at the West China School of Pharmacy, Sichuan University. He joined
efforts can be focused in this area to make new discoveries that Prof. Guo-Bo Li’s group in 2019 and has great interest in developing
not only advance understanding in inhibitor design principles by
new computational methods, including deep learning algorithms, for
integrating medicinal chemistry, bioinorganic chemistry, and
drug discovery.
structural biology but also contribute to developing new
clinically useful drugs targeting metalloenzymes. Qing-Qing Dai is a postgraduate student majoring in Medicinal

■ AUTHOR INFORMATION
Corresponding Authors
Chemistry at the West China School of Pharmacy, Sichuan University.
She joined Prof. Guo-Bo Li’s group in 2019 and mainly conducts
research on developing new computer-aided drug design methods and
Guo-Bo Li − Key Laboratory of Drug-Targeting and Drug deep learning models to aid in drug discovery.
Delivery System of the Education Ministry and Sichuan Gen Li is a Ph.D. candidate majoring in Medicinal Chemistry at the
Province, Sichuan Engineering Laboratory for Plant-Sourced West China School of Pharmacy, Sichuan University. He joined Prof.
Drug and Sichuan Research Center for Drug Precision Guo-Bo Li’s group in 2018 and moved to Prof. Fener Chen’s Group in
Industrial Technology, West China School of Pharmacy, 2021. He has interest in computer-aided drug design and medicinal
Sichuan University, Chengdu 610041, China; orcid.org/
chemistry.
0000-0002-4915-6677; Email: liguobo@scu.edu.cn
You-Cai Xiao − Key Laboratory of Drug-Targeting and Drug Guo-Bo Li is an Associate Professor at the West China School of
Delivery System of the Education Ministry and Sichuan Pharmacy, Sichuan University. He received his Ph.D. degree in 2014 at
Province, Sichuan Engineering Laboratory for Plant-Sourced Sichuan University supervised by Professor Shengyong Yang. He
Drug and Sichuan Research Center for Drug Precision worked as a postdoctoral researcher in Professor Yang’s group from
Industrial Technology, West China School of Pharmacy, 2014 to 2015, then in Professor Christopher J. Schofield’s group at
Sichuan University, Chengdu 610041, China; University of Oxford from 2015 to 2016, and joined in the West China
Email: xiaoliguo1987@scu.edu.cn School of Pharmacy in 2016. He is currently interested in developing
Authors new computational methods for medicinal chemistry and new
Jun-Lin Yu − Key Laboratory of Drug-Targeting and Drug inhibitors/chemical probes targeting clinically relevant enzymes,
Delivery System of the Education Ministry and Sichuan particularly metalloenzymes. He has authored or coauthored more
Province, Sichuan Engineering Laboratory for Plant-Sourced than 60 peer-reviewed publications.


Drug and Sichuan Research Center for Drug Precision
Industrial Technology, West China School of Pharmacy, ACKNOWLEDGMENTS
Sichuan University, Chengdu 610041, China
Qing-Qing Dai − Key Laboratory of Drug-Targeting and Drug This work was partly supported by the National Natural Science
Delivery System of the Education Ministry and Sichuan Foundation of China (Grants 82122065, 82073698, and
Province, Sichuan Engineering Laboratory for Plant-Sourced 81874291), Sichuan Science and Technology Program (Grant
Drug and Sichuan Research Center for Drug Precision 22GJHZ0253), 111 Project (Grant B18035), and Outstanding
Industrial Technology, West China School of Pharmacy, Interdiscipline Project of West China Hospital of Sichuan
Sichuan University, Chengdu 610041, China University (Grant ZYJC18024).
17723 https://doi.org/10.1021/acs.jmedchem.1c01691
J. Med. Chem. 2021, 64, 17706−17727
Journal of Medicinal Chemistry


pubs.acs.org/jmc Perspective

ABBREVIATIONS USED (15) Hai, Y.; Christianson, D. W. Histone deacetylase 6 structure and
molecular basis of catalysis and inhibition. Nat. Chem. Biol. 2016, 12,
ABH, 2-(S)-amnio-6-boronohexanoic acid; ACE, angiotensin 741−747.
converting enzyme; AMP, adenosine monophosphate; ArgI, (16) Cook, N. J.; Li, W.; Berta, D.; Badaoui, M.; Ballandras-Colas, A.;
arginase I; ArgII, arginase II; ATX, autotaxin; AUC, area under Nans, A.; Kotecha, A.; Rosta, E.; Engelman, A. N.; Cherepanov, P.
the curve; CA, carbonic anhydrase; cAMP, cyclic adenosine Structural basis of second-generation HIV integrase inhibitor action
monophosphate; Can2, Cryptococcus neoformans carbonic and viral resistance. Science 2020, 367, 806−810.
anhydrase; CgNce103, Candida globosa carbonic anhydrase; (17) Sagatova, A. A.; Keniya, M. V.; Wilson, R. K.; Monk, B. C.;
cGMP, cyclic guanosine monophosphate; GPCR, G protein- Tyndall, J. D. Structural insights into binding of the antifungal drug
coupled receptor; hCA, human carbonic anhydrase; KPC2, Fluconazole to saccharomyces cerevisiae Lanosterol 14α-demethylase.
Klebsiella pneumoniae carbapenemase 2; LPA, lysophosphatidic Antimicrob. Agents Chemother. 2015, 59, 4982−4989.
(18) Diaz, D. B.; Yudin, A. K. The versatility of boron in biological
acid; LPC, lysophosphatidyl choline; mATX, mouse autotaxin;
target engagement. Nat. Chem. 2017, 9, 731−742.
MBH, Morita−Baylis−Hillman; MBL, metallo-β-lactamase; (19) Riccardi, L.; Genna, V.; De Vivo, M. Metal-ligand interactions in
MBP, metal-binding pharmacophore; MgCA, Malassezia drug design. Nat. Rev. Chem. 2018, 2, 100−112.
globosa carbonic anhydrase; NDM, New Delhi metallo β- (20) Freund, Y. R.; Akama, T.; Alley, M. R. K.; Antunes, J.; Dong, C.;
lactamase; NPP, nucleotide pyrophosphatase and phosphodies- Jarnagin, K.; Kimura, R.; Nieman, J. A.; Maples, K. R.; Plattner, J. J.;
terase; PDB, Protein Data Bank; PDE, phosphodiesterase; Rock, F.; Sharma, R.; Singh, R.; Sanders, V.; Zhou, Y. Boron-based
PMA, phorbol ester; RLA, rat liver arginase; SBL, serine β- phosphodiesterase inhibitors show novel binding of boron to PDE4
lactamase; ScCYP51, Saccharomyces cerevisiae CYP51; TOGA, bimetal center. FEBS Lett. 2012, 586, 3410−3414.
thioornithine generation assay; VIM, Verona Integron encoded (21) Smoum, R.; Rubinstein, A.; Dembitsky, V. M.; Srebnik, M. Boron
metallo-β-lactamase containing compounds as protease inhibitors. Chem. Rev. 2012, 112,


4156−4220.
(22) Fernandes, G. F. S.; Denny, W. A.; Dos Santos, J. L. Boron in drug
REFERENCES design: Recent advances in the development of new therapeutic agents.
(1) Vallee, B. L.; Williams, R. J. Metalloenzymes: the entatic nature of Eur. J. Med. Chem. 2019, 179, 791−804.
their active sites. Proc. Natl. Acad. Sci. U. S. A. 1968, 59, 498−505. (23) Plescia, J.; Moitessier, N. Design and discovery of boronic acid
(2) Shriver, D.; Atkins, P. Chapter 19, Bioinorganic chemistry. In drugs. Eur. J. Med. Chem. 2020, 195, 112270.
Inorganic Chemistry, 3rd ed.; Oxford University Press, 1999. (24) Song, S.; Gao, P.; Sun, L.; Kang, D.; Kongsted, J.; Poongavanam,
(3) Hoppert, M. Metalloenzymes. In Encyclopedia of Geobiology; V.; Zhan, P.; Liu, X. Recent developments in the medicinal chemistry of
Reitner, J., Thiel, V., Eds.; Springer Netherlands: Dordrecht, The single boron atom-containing compounds. Acta Pharm. Sin. B 2021, 11,
3035−3059.
Netherlands, 2011; pp 558−563.
(25) Nocentini, A.; Supuran, C. T.; Winum, J.-Y. Benzoxaborole
(4) Suh, J. Model studies of metalloenzymes involving metal ions as
compounds for therapeutic uses: a patent review (2010- 2018). Expert
Lewis acid catalysts. Acc. Chem. Res. 1992, 25, 273−279.
Opin. Ther. Pat. 2018, 28, 493−504.
(5) Chen, A. Y.; Adamek, R. N.; Dick, B. L.; Credille, C. V.; Morrison,
(26) Leśnikowski, Z. J. Recent developments with boron as a platform
C. N.; Cohen, S. M. Targeting metalloenzymes for therapeutic for novel drug design. Expert Opin. Drug Discovery 2016, 11, 569−578.
intervention. Chem. Rev. 2019, 119, 1323−1455. (27) Boswell-Smith, V.; Spina, D.; Page, C. P. Phosphodiesterase
(6) Mullard, A. 2020 FDA drug approvals. Nat. Rev. Drug Discovery inhibitors. Br. J. Pharmacol. 2006, 147, S252−S257.
2021, 20, 85−90. (28) Jeon, Y. H.; Heo, Y. S.; Kim, C. M.; Hyun, Y. L.; Lee, T. G.; Ro, S.;
(7) Cohen, S. M. A Bioinorganic approach to fragment-based drug Cho, J. M. Phosphodiesterase: overview of protein structures, potential
discovery targeting metalloenzymes. Acc. Chem. Res. 2017, 50, 2007− therapeutic applications and recent progress in drug development. Cell.
2016. Mol. Life Sci. 2005, 62, 1198−1220.
(8) Li, G.; Su, Y.; Yan, Y.-H.; Peng, J.-Y.; Dai, Q.-Q.; Ning, X.-L.; Zhu, (29) Peng, T.; Gong, J.; Jin, Y.; Zhou, Y.; Tong, R.; Wei, X.; Bai, L.;
C.-L.; Fu, C.; McDonough, M. A.; Schofield, C. J.; Huang, C.; Li, G.-B. Shi, J. Inhibitors of phosphodiesterase as cancer therapeutics. Eur. J.
MeLAD: an integrated resource for metalloenzyme-ligand associations. Med. Chem. 2018, 150, 742−756.
Bioinformatics 2019, 36, 904−909. (30) Mehats, C.; Andersen, C. B.; Filopanti, M.; Jin, S. L. C.; Conti, M.
(9) Li, G.; Ning, X.; Li, G.-B. Recent advances in metal binding Cyclic nucleotide phosphodiesterases and their role in endocrine cell
pharmacophores targeting metalloenzymes. Prog. Pharm. Sci. 2020, 44 signaling. Trends Endocrinol. Metab. 2002, 13, 29−35.
(9), 667−680. (31) Jin, S. L. C.; Ding, S.-L.; Lin, S.-C. Phosphodiesterase 4 and its
(10) Dick, B. L.; Cohen, S. M. Metal-binding isosteres as new scaffolds inhibitors in inflammatory diseases. Chang Gung Med. J. 2012, 35, 197−
for metalloenzyme inhibitors. Inorg. Chem. 2018, 57, 9538−9543. 210.
(11) Prosser, K. E.; Kohlbrand, A. J.; Seo, H.; Kalaj, M.; Cohen, S. M. (32) Wittmann, M.; Helliwell, P. S. Phosphodiesterase 4 inhibition in
19F-tagged metal binding pharmacophores for NMR screening of the treatment of psoriasis, psoriatic arthritis and other chronic
metalloenzymes. Chem. Commun. 2021, 57, 4934−4937. inflammatory diseases. Dermatol. Ther. 2013, 3, 1−15.
(12) Credille, C. V.; Dick, B. L.; Morrison, C. N.; Stokes, R. W.; (33) Zebda, R.; Paller, A. S. Phosphodiesterase 4 inhibitors. J. Am.
Adamek, R. N.; Wu, N. C.; Wilson, I. A.; Cohen, S. M. Structure-activity Acad. Dermatol. 2018, 78, S43−S52.
(34) Xu, R. X.; Rocque, W. J.; Lambert, M. H.; Vanderwall, D. E.;
relationships in metal-binding pharmacophores for influenza endonu-
Luther, M. A.; Nolte, R. T. Crystal structures of the catalytic domain of
clease. J. Med. Chem. 2018, 61, 10206−10217. phosphodiesterase 4B complexed with AMP, 8-Br-AMP, and Rolipram.
(13) Natesh, R.; Schwager, S. L. U.; Evans, H. R.; Sturrock, E. D.;
J. Mol. Biol. 2004, 337, 355−365.
Acharya, K. R. Structural details on the binding of antihypertensive (35) Akama, T.; Baker, S. J.; Zhang, Y.-K.; Hernandez, V.; Zhou, H.;
drugs captopril and enalaprilat to human testicular angiotensin I- Sanders, V.; Freund, Y.; Kimura, R.; Maples, K. R.; Plattner, J. J.
converting enzyme. Biochemistry 2004, 43, 8718−8724. Discovery and structure-activity study of a novel benzoxaborole anti-
(14) Sippel, K. H.; Robbins, A. H.; Domsic, J.; Genis, C.; Agbandje- inflammatory agent (AN2728) for the potential topical treatment of
McKenna, M.; McKenna, R. High-resolution structure of human psoriasis and atopic dermatitis. Bioorg. Med. Chem. Lett. 2009, 19,
carbonic anhydrase II complexed with acetazolamide reveals insights 2129−2132.
into inhibitor drug design. Acta Crystallogr., Sect. F: Struct. Biol. Cryst. (36) Zhang, Y.-K.; Plattner, J. J.; Akama, T.; Baker, S. J.; Hernandez, V.
Commun. 2009, 65, 992−995. S.; Sanders, V.; Freund, Y.; Kimura, R.; Bu, W.; Hold, K. M.; Lu, X.-S.

17724 https://doi.org/10.1021/acs.jmedchem.1c01691
J. Med. Chem. 2021, 64, 17706−17727
Journal of Medicinal Chemistry pubs.acs.org/jmc Perspective

Design and synthesis of boron-containing PDE4 inhibitors using soft- (54) Yao, T.; Cohen, R. E. A cryptic protease couples deubiquitination
drug strategy for potential dermatologic anti-inflammatory application. and degradation by the proteasome. Nature 2002, 419, 403−407.
Bioorg. Med. Chem. Lett. 2010, 20, 2270−2274. (55) Borcić, O.; Straus, B. Separation of arginase isoenzymes from
(37) Chu, Z.; Xu, Q.; Zhu, Q.; Ma, X.; Mo, J.; Lin, G.; Zhao, Y.; Gu, Y.; human tissues by agar gel electrophoresis. J. Clin. Chem. Clin. Biochem.
Bian, L.; Shao, L.; Guo, J.; Ye, W.; Li, J.; He, G.; Xu, Y. Design, synthesis 1976, 14, 533−535.
and biological evaluation of novel benzoxaborole derivatives as potent (56) Morris, S. M. Regulation of enzymes of the urea cycle and
PDE4 inhibitors for topical treatment of atopic dermatitis. Eur. J. Med. arginine metabolism. Annu. Rev. Nutr. 2002, 22, 87−105.
Chem. 2021, 213, 113171. (57) Morris, S. M.; Bhamidipati, D.; Kepka-Lenhart, D. Human type II
(38) Strohbach, J. W.; Akama, T.; Blakemore, D. C.; Jacobs, R. T.; arginase: sequence analysis and tissue−specific expression. Gene 1997,
Jones, P.; Limburg, D. C.; Oderinde, M. S.; Perry, M. A.; Plattner, J. J.; 193, 157−161.
Torella, R. F.; Zhou, Y.; Yeoh, T. Y. Boron Containing PDE4 Inhibitors. (58) Jenkinson, C. P.; Grody, W. W.; Cederbaum, S. D. Comparative
U.S. Patent Application 20200108083 A1, 2020. properties of arginases. Comp. Biochem. Physiol., Part B: Biochem. Mol.
(39) Tokumura, A.; Majima, E.; Kariya, Y.; Tominaga, K.; Kogure, K.; Biol. 1996, 114, 107−132.
Yasuda, K.; Fukuzawa, K. Identification of human plasma lysophos- (59) Caldwell, R. B.; Toque, H. A.; Narayanan, S. P.; Caldwell, R. W.
pholipase D, a lysophosphatidic acid−producing enzyme, as autotaxin, Arginase: an old enzyme with new tricks. Trends Pharmacol. Sci. 2015,
a multifunctional phosphodiesterase. J. Biol. Chem. 2002, 277, 39436− 36, 395−405.
39442. (60) Baniyash, M. TCR zeta-chain downregulation: curtailing an
(40) Umezu-Goto, M.; Kishi, Y.; Taira, A.; Hama, K.; Dohmae, N.; excessive inflammatory immune response. Nat. Rev. Immunol. 2004, 4,
Takio, K.; Yamori, T.; Mills, G. B.; Inoue, K.; Aoki, J.; Arai, H. Autotaxin 675−687.
has lysophospholipase D activity leading to tumor cell growth and (61) Munder, M. Arginase: an emerging key player in the mammalian
motility by lysophosphatidic acid production. J. Cell Biol. 2002, 158, immune system. Br. J. Pharmacol. 2009, 158, 638−651.
227−233. (62) Shishova, E. Y.; Di Costanzo, L.; Emig, F. A.; Ash, D. E.;
(41) Nishimasu, H.; Okudaira, S.; Hama, K.; Mihara, E.; Dohmae, N.; Christianson, D. W. Probing the specificity determinants of amino acid
Inoue, A.; Ishitani, R.; Takagi, J.; Aoki, J.; Nureki, O. Crystal structure of recognition by arginase. Biochemistry 2009, 48, 121−131.
autotaxin and insight into GPCR activation by lipid mediators. Nat. (63) Cox, J. D.; Cama, E.; Colleluori, D. M.; Pethe, S.; Boucher, J. L.;
Struct. Mol. Biol. 2011, 18, 205−212. Mansuy, D.; Ash, D. E.; Christianson, D. W. Mechanistic and metabolic
(42) Hausmann, J.; Keune, W.-J.; Hipgrave Ederveen, A. L.; van Zeijl, inferences from the binding of substrate analogues and products to
L.; Joosten, R. P.; Perrakis, A. Structural snapshots of the catalytic cycle arginase. Biochemistry 2001, 40, 2689−2701.
of the phosphodiesterase Autotaxin. J. Struct. Biol. 2016, 195, 199−206. (64) Borek, B.; Gajda, T.; Golebiowski, A.; Blaszczyk, R. Boronic acid-
(43) Castagna, D.; Budd, D. C.; Macdonald, S. J.; Jamieson, C.; based arginase inhibitors in cancer immunotherapy. Bioorg. Med. Chem.
Watson, A. J. Development of autotaxin inhibitors: an overview of the 2020, 28, 115658.
patent and primary literature. J. Med. Chem. 2016, 59, 5604−5621. (65) Pudlo, M.; Demougeot, C.; Girard-Thernier, C. Arginase
(44) Matralis, A. N.; Afantitis, A.; Aidinis, V. Development and inhibitors: a rational approach over one century. Med. Res. Rev. 2017,
therapeutic potential of autotaxin small molecule inhibitors: from 37, 475−513.
bench to advanced clinical trials. Med. Res. Rev. 2019, 39, 976−1013. (66) Moretto, J.; Pudlo, M.; Demougeot, C. Human-based evidence
(45) Nikolaou, A.; Kokotou, M. G.; Limnios, D.; Psarra, A.; Kokotos, for the therapeutic potential of arginase inhibitors in cardiovascular
G. Autotaxin inhibitors: a patent review (2012−2016). Expert Opin. diseases. Drug Discovery Today 2021, 26, 138−147.
Ther. Pat. 2017, 27, 815−829. (67) Baggio, R.; Elbaum, D.; Kanyo, Z. F.; Carroll, P. J.; Cavalli, R. C.;
(46) Albers, H. M.; van Meeteren, L. A.; Egan, D. A.; van Tilburg, E. Ash, D. E.; Christianson, D. W. Inhibition of Mn2+2-arginase by borate
W.; Moolenaar, W. H.; Ovaa, H. Discovery and optimization of boronic leads to the design of a transition state analogue inhibitor, 2(S)-amino-
acid based inhibitors of autotaxin. J. Med. Chem. 2010, 53, 4958−4967. 6-boronohexanoic Acid. J. Am. Chem. Soc. 1997, 119, 8107−8108.
(47) Albers, H. M. H. G.; Dong, A.; van Meeteren, L. A.; Egan, D. A.; (68) Van Zandt, M. C.; Whitehouse, D. L.; Golebiowski, A.; Ji, M. K.;
Sunkara, M.; van Tilburg, E. W.; Schuurman, K.; van Tellingen, O.; Zhang, M.; Beckett, R. P.; Jagdmann, G. E.; Ryder, T. R.; Sheeler, R.;
Morris, A. J.; Smyth, S. S.; Moolenaar, W. H.; Ovaa, H. Boronic acid- Andreoli, M.; Conway, B.; Mahboubi, K.; D’Angelo, G.; Mitschler, A.;
based inhibitor of autotaxin reveals rapid turnover of LPA in the Cousido-Siah, A.; Ruiz, F. X.; Howard, E. I.; Podjarny, A. D.; Schroeter,
circulation. Proc. Natl. Acad. Sci. U. S. A. 2010, 107, 7257−7262. H. Discovery of (R)-2-amino-6-borono-2-(2-(piperidin-1-yl)ethyl)-
(48) Groll, M.; Berkers, C. R.; Ploegh, H. L.; Ovaa, H. Crystal hexanoic acid and congeners as highly potent inhibitors of human
Structure of the boronic acid−based proteasome inhibitor Bortezomib arginases I and II for treatment of myocardial reperfusion injury. J. Med.
in complex with the yeast 20S proteasome. Structure 2006, 14, 451− Chem. 2013, 56, 2568−2580.
456. (69) Di Costanzo, L.; Sabio, G.; Mora, A.; Rodriguez, P. C.; Ochoa, A.
(49) Hausmann, J.; Kamtekar, S.; Christodoulou, E.; Day, J. E.; Wu, C.; Centeno, F.; Christianson, D. W. Crystal structure of human
T.; Fulkerson, Z.; Albers, H. M. H. G.; van Meeteren, L. A.; Houben, A. arginase I at 1.29-Å resolution and exploration of inhibition in the
J. S.; van Zeijl, L.; Jansen, S.; Andries, M.; Hall, T.; Pegg, L. E.; Benson, immune response. Proc. Natl. Acad. Sci. U. S. A. 2005, 102, 13058−
T. E.; Kasiem, M.; Harlos, K.; Kooi, C. W. V.; Smyth, S. S.; Ovaa, H.; 13063.
Bollen, M.; Morris, A. J.; Moolenaar, W. H.; Perrakis, A. Structural basis (70) Shin, H.; Cama, E.; Christianson, D. W. Design of amino acid
of substrate discrimination and integrin binding by autotaxin. Nat. aldehydes as transition-state analogue inhibitors of arginase. J. Am.
Struct. Mol. Biol. 2011, 18, 198−204. Chem. Soc. 2004, 126, 10278−10284.
(50) Kawaguchi, M.; Okabe, T.; Okudaira, S.; Nishimasu, H.; Ishitani, (71) Busnel, O.; Carreaux, F.; Carboni, B.; Pethe, S.; Goff, S. V.-L.;
R.; Kojima, H.; Nureki, O.; Aoki, J.; Nagano, T. Screening and X-ray Mansuy, D.; Boucher, J.-L. Synthesis and evaluation of new ω-borono-
crystal structure-based optimization of autotaxin (ENPP2) inhibitors, α-amino acids as rat liver arginase inhibitors. Bioorg. Med. Chem. 2005,
using a newly developed fluorescence probe. ACS Chem. Biol. 2013, 8, 13, 2373−2379.
1713−1721. (72) Collet, S.; Carreaux, F.; Boucher, J.-L.; Pethe, S.; Lepoivre, M.;
(51) Kraljić, K.; Jelić, D.; Ž iher, D.; Cvrtila, A.; Dragojević, S.; Danion-Bougot, R.; Danion, D. Synthesis and evaluation of ω-borono-
Sinković, V.; Mesić, M. Benzoxaboroles−novel autotaxin inhibitors. α-amino acids as active-site probes of arginase and nitric oxide
Molecules 2019, 24, 3419. synthases. J. Chem. Soc., Perkin Trans. 1. 2000, 177−182.
(52) Albers, H. M. H. G.; Ovaa, H. Chemical evolution of autotaxin (73) Ilies, M.; Di Costanzo, L.; Dowling, D. P.; Thorn, K. J.;
inhibitors. Chem. Rev. 2012, 112, 2593−2603. Christianson, D. W. Binding of α,α-disubstituted amino acids to
(53) Kossel, A.; Dakin, H. D. Ü ber die Arginase. Hoppe-Seyler's Z. arginase suggests new avenues for inhibitor design. J. Med. Chem. 2011,
Physiol. Chem. 1904, 41, 321−331. 54, 5432−5443.

17725 https://doi.org/10.1021/acs.jmedchem.1c01691
J. Med. Chem. 2021, 64, 17706−17727
Journal of Medicinal Chemistry pubs.acs.org/jmc Perspective

(74) Golebiowski, A.; Whitehouse, D.; Beckett, R. P.; Van Zandt, M.; (91) Larcher, A.; Nocentini, A.; Supuran, C. T.; Winum, J. Y.; van der
Ji, M. K.; Ryder, T. R.; Jagdmann, E.; Andreoli, M.; Lee, Y.; Sheeler, R.; Lee, A.; Vasseur, J. J.; Laurencin, D.; Smietana, M. Bis-benzoxaboroles:
Conway, B.; Olczak, J.; Mazur, M.; Czestkowski, W.; Piotrowska, W.; design, synthesis, and biological evaluation as carbonic anhydrase
Cousido-Siah, A.; Ruiz, F. X.; Mitschler, A.; Podjarny, A.; Schroeter, H. inhibitors. ACS Med. Chem. Lett. 2019, 10, 1205−1210.
Synthesis of quaternary α-amino acid-based arginase inhibitors via the (92) Innocenti, A.; Winum, J.-Y.; Hall, R.; Mü hlschlegel, F.;
Ugi reaction. Bioorg. Med. Chem. Lett. 2013, 23, 4837−4841. Scozzafava, A.; Supuran, C. Carbonic anhydrase inhibitors. Inhibition
(75) Blaszczyk, R.; Brzezinska, J.; Dymek, B.; Stanczak, P. S.; of the fungal β-carbonic anhydrases from Candida albicans and
Mazurkiewicz, M.; Olczak, J.; Nowicka, J.; Dzwonek, K.; Zagozdzon, A.; Cryptococcus neoformans with boronic acids. Bioorg. Med. Chem. Lett.
Golab, J.; Golebiowski, A. Discovery and pharmacokinetics of 2009, 19, 2642−2645.
sulfamides and guanidines as potent human arginase 1 inhibitors. (93) Supuran, C. T. Bortezomib inhibits bacterial and fungal β-
ACS Med. Chem. Lett. 2020, 11, 433−438. carbonic anhydrases. Bioorg. Med. Chem. 2016, 24, 4406−4409.
(76) Van Zandt, M. C.; Jagdmann, G. E.; Whitehouse, D. L.; Ji, M.; (94) Nocentini, A.; Cadoni, R.; del Prete, S.; Capasso, C.; Dumy, P.;
Savoy, J.; Potapova, O.; Cousido-Siah, A.; Mitschler, A.; Howard, E. I.; Gratteri, P.; Supuran, C. T.; Winum, J.-Y. Benzoxaboroles as efficient
Pyle, A. M.; Podjarny, A. D. Discovery of N-substituted 3-amino-4-(3- inhibitors of the β-Carbonic anhydrases from pathogenic fungi: activity
boronopropyl)pyrrolidine-3-carboxylic acids as highly potent third- and modeling study. ACS Med. Chem. Lett. 2017, 8, 1194−1198.
generation inhibitors of human arginase I and II. J. Med. Chem. 2019, (95) Bonardi, A.; Nocentini, A.; Cadoni, R.; Del Prete, S.; Dumy, P.;
62, 8164−8177. Capasso, C.; Gratteri, P.; Supuran, C. T.; Winum, J. Y. Benzoxaboroles:
(77) Mitcheltree, M. J.; Li, D.; Achab, A.; Beard, A.; Chakravarthy, K.; new potent inhibitors of the carbonic anhydrases of the pathogenic
Cheng, M.; Cho, H.; Eangoor, P.; Fan, P.; Gathiaka, S.; Kim, H. Y.; bacterium Vibrio cholerae. ACS Med. Chem. Lett. 2020, 11, 2277−2284.
Lesburg, C. A.; Lyons, T. W.; Martinot, T. A.; Miller, J. R.; McMinn, S.; (96) Bush, K.; Jacoby, G. A. Updated functional classification of beta-
O’Neil, J.; Palani, A.; Palte, R. L.; Saurí, J.; Sloman, D. L.; Zhang, H.; lactamases. Antimicrob. Agents Chemother. 2010, 54, 969−976.
Cumming, J. N.; Fischer, C. Discovery and optimization of rationally (97) King, A. M.; Reid-Yu, S. A.; Wang, W.; King, D. T.; De Pascale,
designed bicyclic inhibitors of human arginase to enhance cancer G.; Strynadka, N. C.; Walsh, T. R.; Coombes, B. K.; Wright, G. D.
immunotherapy. ACS Med. Chem. Lett. 2020, 11, 582−588. Aspergillomarasmine A overcomes metallo-β-lactamase antibiotic
(78) Sjogren, E. B.; Li, J.; Van, Z. M.; Whitehouse, D. Compositions resistance. Nature 2014, 510, 503−506.
and Method for Inhibiting Arginase Activity. World Patent Application- (98) Torelli, N. J.; Akhtar, A.; DeFrees, K.; Jaishankar, P.; Pemberton,
O. A.; Zhang, X.; Johnson, C.; Renslo, A. R.; Chen, Y. Active-site
WO 2017/075363 A1.
druggability of carbapenemases and broad-spectrum inhibitor discov-
(79) Buemi, M. R.; Di Fiore, A.; De Luca, L.; Angeli, A.; Mancuso, F.;
ery. ACS Infect. Dis. 2019, 5, 1013−1021.
Ferro, S.; Monti, S. M.; Buonanno, M.; Russo, E.; De Sarro, G.; De
(99) Queenan, A. M.; Bush, K. Carbapenemases: the versatile beta-
Simone, G.; Supuran, C. T.; Gitto, R. Exploring structural properties of
lactamases. Clin. Microbiol. Rev. 2007, 20, 440−458.
potent human carbonic anhydrase inhibitors bearing a 4-(cyclo-
(100) Bush, K.; Bradford, P. A. Interplay between β-lactamases and
alkylamino-1-carbonyl)benzenesulfonamide moiety. Eur. J. Med. Chem.
new β-lactamase inhibitors. Nat. Rev. Microbiol. 2019, 17, 295−306.
2019, 163, 443−452. (101) Abboud, M. I.; Damblon, C.; Brem, J.; Smargiasso, N.; Mercuri,
(80) Supuran, C. T. Carbonic anhydrases: novel therapeutic P.; Gilbert, B.; Rydzik, A. M.; Claridge, T. D. W.; Schofield, C. J.; Frère,
applications for inhibitors and activators. Nat. Rev. Drug Discovery J.-M. Interaction of avibactam with class B metallo-β-lactamases.
2008, 7, 168−181. Antimicrob. Agents Chemother. 2016, 60, 5655−5662.
(81) Ferry, J. G. How to make a living by exhaling methane. Annu. Rev. (102) Yan, Y.-H.; Li, G.; Li, G.-B. Principles and current strategies
Microbiol. 2010, 64, 453−473. targeting metallo-β-lactamase mediated antibacterial resistance. Med.
(82) Jensen, E. L.; Clement, R.; Kosta, A.; Maberly, S. C.; Gontero, B. Res. Rev. 2020, 40, 1558−1592.
A new widespread subclass of carbonic anhydrase in marine (103) Krajnc, A.; Lang, P. A.; Panduwawala, T. D.; Brem, J.; Schofield,
phytoplankton. ISME J. 2019, 13, 2094−2106. C. J. Will morphing boron-based inhibitors beat the β-lactamases? Curr.
(83) Buabeng, E. R.; Henary, M. Developments of small molecules as Opin. Chem. Biol. 2019, 50, 101−110.
inhibitors for carbonic anhydrase isoforms. Bioorg. Med. Chem. 2021, (104) Ness, S.; Martin, R.; Kindler, A. M.; Paetzel, M.; Gold, M.;
39, 116140. Jensen, S. E.; Jones, J. B.; Strynadka, N. C. J. Structure-based design
(84) Kumar, S.; Rulhania, S.; Jaswal, S.; Monga, V. Recent advances in guides the improved efficacy of deacylation transition state analogue
the medicinal chemistry of carbonic anhydrase inhibitors. Eur. J. Med. inhibitors of TEM-1 β-lactamase. Biochemistry 2000, 39, 5312−5321.
Chem. 2021, 209, 112923. (105) Brem, J.; Cain, R.; Cahill, S.; McDonough, M. A.; Clifton, I. J.;
(85) Supuran, C. T. Advances in structure-based drug discovery of Jiménez-Castellanos, J.-C.; Avison, M. B.; Spencer, J.; Fishwick, C. W.
carbonic anhydrase inhibitors. Expert Opin. Drug Discovery 2017, 12, G.; Schofield, C. J. Structural basis of metallo-β-lactamase, serine-β-
61−88. lactamase and penicillin-binding protein inhibition by cyclic boronates.
(86) Eriksson, A. E.; Jones, T. A.; Liljas, A. Refined structure of human Nat. Commun. 2016, 7, 12406.
carbonic anhydrase II at 2.0 Å resolution. Proteins: Struct., Funct., Genet. (106) Liu, B.; Trout, R. E. L.; Chu, G. H.; McGarry, D.; Jackson, R.
1988, 4, 274−282. W.; Hamrick, J. C.; Daigle, D. M.; Cusick, S. M.; Pozzi, C.; De Luca, F.;
(87) Sjöblom, B.; Polentarutti, M.; Djinovic-Carugo, K. Structural Benvenuti, M.; Mangani, S.; Docquier, J. D.; Weiss, W. J.; Pevear, D. C.;
study of X-ray induced activation of carbonic anhydrase. Proc. Natl. Xerri, L.; Burns, C. J. Discovery of Taniborbactam (VNRX-5133): a
Acad. Sci. U. S. A. 2009, 106, 10609−10613. broad-spectrum serine- and metallo-β-lactamase inhibitor for carbape-
(88) Winum, J.-Y.; Innocenti, A.; Scozzafava, A.; Montero, J.-L.; nem-resistant bacterial infections. J. Med. Chem. 2020, 63, 2789−2801.
Supuran, C. T. Carbonic anhydrase inhibitors. Inhibition of the human (107) Krajnc, A.; Brem, J.; Hinchliffe, P.; Calvopiña, K.; Panduwawala,
cytosolic isoforms I and II and transmembrane, tumor-associated T. D.; Lang, P. A.; Kamps, J.; Tyrrell, J. M.; Widlake, E.; Saward, B. G.;
isoforms IX and XII with boronic acids. Bioorg. Med. Chem. 2009, 17, Walsh, T. R.; Spencer, J.; Schofield, C. J. Bicyclic boronate VNRX-5133
3649−3652. inhibits metallo- and serine-β-lactamases. J. Med. Chem. 2019, 62,
(89) Supuran, C. T. Bortezomib inhibits mammalian carbonic 8544−8556.
anhydrases. Bioorg. Med. Chem. 2017, 25, 5064−5067. (108) Koike, S.; Takai, K.; Mizushima, S.; Kanai, T. Heterocyclic
(90) Alterio, V.; Cadoni, R.; Esposito, D.; Vullo, D.; Fiore, A. D.; derivatives having excellent β-lactamase-inhibiting activity as prophylactic
Monti, S. M.; Caporale, A.; Ruvo, M.; Sechi, M.; Dumy, P.; Supuran, C. or therapeutic agent for bacterial infectious diseases. World Patent
T.; Simone, G. D.; Winum, J.-Y. Benzoxaborole as a new chemotype for Application WO 2018/199291 A1, 2018.
carbonic anhydrase inhibition. Chem. Commun. 2016, 52, 11983− (109) Hecker, S. J.; Reddy, K. R.; Lomovskaya, O.; Griffith, D. C.;
11986. Rubio-Aparicio, D.; Nelson, K.; Tsivkovski, R.; Sun, D.; Sabet, M.;

17726 https://doi.org/10.1021/acs.jmedchem.1c01691
J. Med. Chem. 2021, 64, 17706−17727
Journal of Medicinal Chemistry pubs.acs.org/jmc Perspective

Tarazi, Z.; Parkinson, J.; Totrov, M.; Boyer, S. H.; Glinka, T. W.;
Pemberton, O. A.; Chen, Y.; Dudley, M. N. Discovery of cyclic boronic
acid QPX7728, an ultrabroad-spectrum inhibitor of serine and metallo-
β-lactamases. J. Med. Chem. 2020, 63, 7491−7507.
(110) Yan, Y.-H.; Li, Z.-F.; Ning, X.-L.; Deng, J.; Yu, J.-L.; Luo, Y.;
Wang, Z.; Li, G.; Li, G.-B.; Xiao, Y.-C. Discovery of 3-aryl substituted
benzoxaboroles as broad-spectrum inhibitors of serine- and metallo-β-
lactamases. Bioorg. Med. Chem. Lett. 2021, 41, 127956.
(111) Xiao, Y.-C.; Chen, X.-P.; Deng, J.; Yan, Y.-H.; Zhu, K.-R.; Li, G.;
Yu, J.-L.; Brem, J.; Chen, F.; Schofield, C. J.; Li, G.-B. Design and
enantioselective synthesis of 3-(α-acrylic acid) benzoxaboroles to
combat carbapenemase resistance. Chem. Commun. 2021, 57, 7709−
7712.
(112) Kiener, P. A.; Waley, S. G. Reversible inhibitors of penicillinases.
Biochem. J. 1978, 169, 197−204.
(113) Cendron, L.; Quotadamo, A.; Maso, L.; Bellio, P.; Montanari,
M.; Celenza, G.; Venturelli, A.; Costi, M. P.; Tondi, D. X-ray
crystallography deciphers the activity of broad-spectrum boronic acid
β-lactamase inhibitors. ACS Med. Chem. Lett. 2019, 10, 650−655.
(114) Wang, Y. L.; Liu, S.; Yu, Z. J.; Lei, Y.; Huang, M. Y.; Yan, Y. H.;
Ma, Q.; Zheng, Y.; Deng, H.; Sun, Y.; Wu, C.; Yu, Y.; Chen, Q.; Wang,
Z.; Wu, Y.; Li, G. B. Structure-based development of (1-(3′-
mercaptopropanamido)methyl)boronic acid derived broad-spectrum,
dual-action inhibitors of metallo- and serine-β-lactamases. J. Med.
Chem. 2019, 62, 7160−7184.
(115) Farley, A. J. M.; Ermolovich, Y.; Calvopiña, K.; Rabe, P.;
Panduwawala, T.; Brem, J.; Björkling, F.; Schofield, C. J. Structural basis
of metallo-β-lactamase inhibition by N-sulfamoylpyrrole-2-carboxy-
lates. ACS Infect. Dis. 2021, 7, 1809−1817.
(116) Shirley, M. Ixazomib: first global approval. Drugs 2016, 76,
405−411.
(117) Salvini, M.; Troia, R.; Giudice, D.; Pautasso, C.; Boccadoro, M.;
Larocca, A. Pharmacokinetic drug evaluation of ixazomib citrate for the
treatment of multiple myeloma. Expert Opin. Drug Metab. Toxicol.
2018, 14, 91−99.
(118) Dai, Q.; Yan, Y.; Ning, X.; Li, G.; Yu, J.; Deng, J.; Yang, L.; Li, G.-
B. AncPhore: A versatile tool for anchor pharmacophore steered drug
discovery with applications in discovery of new inhibitors targeting
metallo-β-lactamases and indoleamine/tryptophan 2,3-dioxygenases.
Acta Pharm. Sin. B 2021, 11, 1931−1946.
(119) Yu, J.-L.; Dai, Q.-Q.; Li, G.-B. Deep learning in target prediction
and drug repositioning: recent advances and challenges. Drug Discovery
Today 2021, DOI: 10.1016/j.drudis.2021.10.010.
(120) Santucci, M.; Spyrakis, F.; Cross, S.; Quotadamo, A.; Farina, D.;
Tondi, D.; De Luca, F.; Docquier, J.-D.; Prieto, A. I.; Ibacache, C.;
Blázquez, J.; Venturelli, A.; Cruciani, G.; Costi, M. P. Computational
and biological profile of boronic acids for the detection of bacterial
serine- and metallo-β-lactamases. Sci. Rep. 2017, 7, 17716.

17727 https://doi.org/10.1021/acs.jmedchem.1c01691
J. Med. Chem. 2021, 64, 17706−17727

You might also like