Boronic Acid Compounds as
`Potential Pharmaceutical Agents
`
`WenqianYang,XingmingGao,BingheWang
`
`Department of Chemistry, North Carolina State University, Raleigh, North Carolina 27695-8204
`
`Published online 00 Month 2003 in Wiley InterScience (www.interscience.wiley.com).
`DOI 10.1002/med.10043
`
`!
`
`Abstract: Boronic acid compounds have been used, because of their unique structural features, for
`the development of potent enzyme inhibitors, boron neutron capture agents for cancer therapy, and
`as antibody mimics that recognize biologically important saccharides. Consequently, there has
`been a surge of interests in boronic acid compounds. This study reviews the recent development in
`this area during the last six years. ß 2003 Wiley Periodicals, Inc. Med Res Rev, 23 No. 3, 346–368, 2003
`
`Key words: boronic acid; enzyme inhibitor; boron neutron; capture agent; antibody mimics; drug
`delivery
`
`1. INTRODUCTION
`
`Recently, there is an increasing interest in boronic acid compounds. Such an interest stems from
`the tremendous importance of boronic acids in the synthesis of biologically active compounds and
`the use of boronic acid themselves as pharmaceutical agents. In the area of synthetic medicinal
`chemistry, boronic acids are important intermediates that have been widely used in Suzuki cross-
`coupling reactions,1 protection of diols,2 Diels-Alder reactions,3 asymmetric synthesis of amino
`acids,4 selective reduction of aldehydes,5 carboxylic acid activation,6,7 and as a template in organic
`synthesis.8 As potential pharmaceutical agents, boronic acids have been used for the development
`of enzyme inhibitors;9 boron neutron capture therapy (BNCT) agents,10 feedback controlled drug
`delivery polymers,11 saccharide sensors,12–14 and the antibody mimics for cell-surface polysacchar-
`ides.15,16 This review will focus on the recent development in the last six years in the development of
`enzyme inhibitors, BNCT agents, and polymers used for feedback controlled delivery insulin. There
`have been several reviews published in the area of saccharide sensors.13,14 Therefore, this part will not
`be duplicated here.
`
`Correspondence to: Professor BingheWang,Ph.D., Department of Chemistry, North Carolina State University, Raleigh,NC 27695-
`8204. E-mail: binghe _ wang@ncsu.edu
`Contract grant sponsor: National Institutes of Health; Contract Grant numbers: NO1-CO-27184, CA88343,
`DK55062.Contract grant sponsor: North Carolina Biotechnology Center; Contract Grant number: 2001ARG0016.
`
`Medicinal Research Reviews, Vol. 23, No. 3, 346 ^368, 2003
`ß 2003 Wiley Periodicals, Inc.
`
`CFAD v. Anacor, IPR2015-01776, CFAD EXHIBIT 1060 - Page 1 of 23
`
`

`
`BORONIC ACID COMPOUNDS
`
`*
`
`347
`
`2. GENERAL PROPERTIES OF BORONIC ACID COMPOUNDS AND
`THEIR IMPLICATIONS IN BIOLOGICAL APPLICATIONS
`
`The utility of boronic acid compounds as pharmaceutical agents is directly related to their unique
`electronic and physicochemical properties. Boron occupies a special place in the periodic table. It is in
`the same period as carbon, but has one less electron. Therefore, it has many similarities with carbon in
`terms of structural features, which make it very useful in the world of carbon in organic and medicinal
`chemistry. The fact that there are many boron-based reagents in organic synthesis reflects this
`structural similarity.17 In medicinal chemistry, the use of boronic acids as enzyme inhibitors to a large
`degree reflects the usefulness of boron as a carbon analog in the binding process, but not in terms
`of reactions, which is the essence of a good enzyme inhibitor. Boronic acids have been used for
`the development of enzyme inhibitors of peptidases/proteases, proteasomes, arginase, nitric oxide
`synthase (NOS), as well as transpeptidases.
`One unique property of boronic acid is that it is a strong Lewis acid because of the boron open
`shell. Most phenylboronic acids have a pKa in the range of 4.5–8.818 depending upon the phenyl
`substitution.19,20 This means that with the appropriate substitution, boronic acids would have the right
`property for ready conversion from a neutral and trigonal planar sp2 boron to an anionic tetrahedral sp3
`boron (Scheme 1) under physiological conditions. Realizing that the process of cleaving an amide
`bond also requires the conversion of an sp2 carbonyl carbon to a tetrahedral sp3 carbon, it is easy to
`understand that boronic acid compounds would make good transition state analogs for the inhibition
`of hydrolytic enzymes. This is indeed the case. Some 20 years ago, simple alkyl or arylboronic acids
`were recognized as serine protease inhibitors.21 –23 Since then, many boronic acid compounds with an
`appropriate peptide sequences have been designed and synthesized for the development of more
`potent and selective inhibitors.24 When compared with aldehyde-based inhibitors of hydrolytic
`enzymes, the ready conversion of boronic acids to their anionic sp3 form seems to make them better
`transition state analogs.25 Although not the emphasis of this review, it needs to be noted that Matteson
`et al. has established a general synthetic route to chiral a-aminoalkylboronic acid derivatives by
`stereoselective homologation of pinanediol boronic esters.25,26 This enabled the synthesis of many
`potent boronic acid-based enzyme inhibitors. Thereafter, several variations of the general route have
`been developed and used for the synthesis of different kinds of enzyme inhibitors.27 –31
`In addition to being developed as enzyme inhibitors, boron-based compounds (not limited to
`boronic acid compounds although that is the focus of this review) are also being studied for their
`utility as BNCT agents.10,32,33 Such applications are based on the unique property of boron-10,
`which emit a particles upon irradiation with neutron. Since a particles do not travel a long distance
`(a few millimeter), they are ideal for localized radiation therapy. Therefore, targeted delivery of high
`concentrations of boron agents can be used for BNCT of certain tumor.
`The third potential application of boronic acid compounds to be discussed is the development of
`feedback controlled delivery systems for insulin. For such an application, the ideal controlling signal
`is glucose concentration. Boronic acids have the unique properties of forming reversible complexes
`with diol-containing compounds such as sugars.19,34 Therefore, efforts have been made to prepare
`polymers that can respond to glucose concentration variation with permeability changes. Such perme-
`ability changes can in turn be used for controlling the release of insulin from the polymer encapsulation.
`Again, this review will focus on using boronic acid compounds for the development of enzyme
`inhibitors, BNCT agents, and feedback controlled delivery systems for insulin.
`
`Scheme1. Conversion from a neutral and trigonal planar sp2 boronto an anionic tetrahedral sp3 boron.
`
`CFAD v. Anacor, IPR2015-01776, CFAD EXHIBIT 1060 - Page 2 of 23
`
`

`
`348
`
`*
`
`YANG, GAO, AND WANG
`
`3. BORONIC ACID COMPOUNDS AS ENZYME INHIBITORS
`
`As discussed in the Introduction, the use of boronic acid compounds as enzyme inhibitors is mostly
`based on their easy conversion between the trigonal and tetrahedral forms (Scheme 1), which make
`them ideal transition state analogs in hydrolytic processes. Therefore, various boronic acid com-
`pounds have been widely studied for their inhibition of hydrolytic enzymes such as proteases.
`The following discussion is divided based on the target enzymes.
`
`A. Serine Protease Inhibitors
`
`Thrombin, as the final serine protease in the blood coagulation cascade, is a promising target for the
`development of an anticoagulant agent. Therefore, there is a great deal of interest in the development
`of thrombin inhibitors. Boronic acid compounds were found to exhibit potent inhibition activities.35
`Recently, through the examination of the X-ray crystal structure of boropeptide (1) bound to
`thrombin, it was found that the 3-phenylpropionyl chain attached to the proline residue forms a
`favorable edge-to-face interaction with the Trp-215 side chain located at the base of the S3 specificity
`pocket of thrombin36 (Figs. 1 and 2). To maximize this edge-to-face interaction, rigidified analogs of 1
`and 2 were designed. In such a design, a cyclohexane ring (3) or a pyrrolidine ring (4) was used to hold
`the phenylpropionyl moiety in an orientation favorable for the interaction with the Trp-215 residue as
`predicted by computer modeling studies based on the X-ray crystal structure. Both constrained
`analogs 3 and 4 showed a twofold increase in potency relative to their unconstrained counterparts
`1 and 2, respectively. In a related effort to maximize the edge-to-face interaction with the Trp-215 side
`chain, the P3 residue of 5, a previously discovered inhibitor, was replaced by benzoic acid-derived
`residues. This afforded an extremely potent thrombin inhibitor, compound 6, which is approximately
`threefold more potent than the lead compound (5).37
`
`Figure1
`
`CFAD v. Anacor, IPR2015-01776, CFAD EXHIBIT 1060 - Page 3 of 23
`
`

`
`BORONIC ACID COMPOUNDS
`
`*
`
`349
`
`Figure2. Thebindingofcompound1tothrombin.Three expectedinteractions are shown: (a) interactionoftheaminoside chainwith
`Asp-189 in the S1specificity pocket; (b) tetrahedral complex between the hydroxyl of Ser-195 and the boron of the inhibitor; (c) the
`edge-to-face interaction of the 3-phenylpropionyl (P3) residue withTrp-215 located at the base of the S3 specificity pocket.
`
`One concern with the use of thrombin inhibitors as anti-coagulants is the non-specific inhibition
`of other related enzymes. Earlier studies from DuPont Pharmaceuticals identified DuP-714 as a very
`potent thrombin inhibitor with a Ki of 0.07 nM. However, animal studies indicated that DuP-714
`caused side effect that appears to be related to the undesirable inhibition of complement factor I.
`To design inhibitors with minimal interaction with factor I, it was important to analyze the difference
`in the binding requirements between factor I and thrombin. However, the crystal structure of factor I
`was not available. Therefore, the crystal structure of factor Xa (fXa) was used, working on the
`assumption that the overall conformation of factor I is similar to that of fXa. Crystal structural
`analyses of the inhibitor–enzyme complexes with different inhibitors showed that there were very
`noticeable differences in the P2 pocket. Therefore, a series of b,b-dialkylphenethylglycine P2 analogs
`of DuP-714 were designed and synthesized. These compounds, such as 7 and 8, have greater
`selectivity for thrombin over factor I and improved safety profile.38
`There have also been efforts in designing selective thrombin inhibitors by varying the P1 position.
`For example, incorporation of m-cyano-substituted phenylalanine boronic acid analogues into R-
`(D)Phe-Pro-OH dipeptides produced several highly effective thrombin inhibitors such as H-(D)Phe-
`Pro-boroPhe(m-CN)-OH.39 The cyano group enhances binding by several orders of magnitude.
`Because of its structural and functional similarities with thrombin, trypsin was used as a surrogate
`in the crystal structural studies. The trypsin-H-(D)Phe-Pro-boroPhe(m-CN)-OH (Ki ¼ 0.48 nM)
`complex showed that the aromatic side chain was bound in the P1 binding site and that the cynao group
`acted as a H-bond acceptor for the amide proton of the Gly-219.
`
`Figure3
`
`CFAD v. Anacor, IPR2015-01776, CFAD EXHIBIT 1060 - Page 4 of 23
`
`

`
`350
`
`*
`
`YANG, GAO, AND WANG
`
`Figure4
`
`In a separate study, based on the known potent inhibition effect of hirudin on thrombin, a novel peptide
`boronate as thrombin inhibitor was designed and synthesized using solid phase chemistry and suitably
`protected aminoboronates.40 By conjugating a boronic acid moiety with a hirudin-based recognition
`moiety, [-D-PheProBoroBpgOPin]-CO(CH2)3COCly2Hir was synthesized and shown to have a very
`high affinity for the target enzyme (Ki ¼ 0.6 nM). It has a 10-fold higher potency relative to the
`corresponding non-hirudin-containing portion Z-D-PheProBoroBpgOPin (9) or the mixture of non-
`covalently linked units.
`Factor Xa (fXa) is another important protease in the coagulation cascade, which occupies
`the juncture of the intrinsic and extrinsic clotting pathways. The physiological role of fXa is the
`proteolytic cleavage of prothrombin to thrombin. Therefore, development of inhibitors against
`fXa should be an attractive method of thrombosis prevention. During the screening of a series of
`conformationally restricted boropeptide thrombin inhibitors, a borolysine compound (10) containing
`a 2-(2-cyanophenylthio)benzoyl in the P3 position was found to be a potent fXa inhibitor.41 It has a
`16-fold higher potency relative to the corresponding compound 11 without the nitrile moiety.
`The serine proteases subtilisin Carlsberg and a-chymotrypsin are commercially available
`enzymes for which high-resolution X-ray crystal structures are known. Therefore, they are good
`targets for probing the factors responsible for determining the structural and stereospecificity of
`enzymes toward unnatural substrates and inhibitors. Enantiomeric 1-acetomido boronic acid analogs
`of the L- and D-forms of alanine, phenylalanine, p-fluorophenyalanine, p-chlorophenylalanine, and
`1-naphthyanaline were synthesized and evaluated as inhibitors of the serine proteases subtilisin
`Carlsberg and a-chymotrypsin.42 All of the boronic acids examined are powerful competitive inhibi-
`tors of both enzymes. The L-enantiomers are generally more potent than the D-enantiomers. However,
`[1-acetamido-2-(1-naphthyl)ethyl]boronic acid showed a dramatic reversal of the normal stereo-
`selectivity preference with the D-enantiomer being 25-fold more potent than the L-enantiomer.
`Molecular modeling analyses of the possible binding modes of the inhibitors indicate that the
`stereoselectivity reversal is because of S1-pocket orientation differences between the naphthyl group
`and the aromatic chains of the phenylalanine analogs.
`To explore the possibility of forming a peptide boronate adduct in the serine protease active site
`that mimics the first tetrahedral intermediate in the peptide hydrolysis mechanism, peptidyl boronic
`acids (12), (13), and (14) were designed and synthesized.43 This design intended to take advantage of
`an intramolecular process hoping to overcome the inherent disadvantage of ternary adduct formation
`(Fig. 5a) by tethering P0 components to the peptidyl boronic acid (Fig. 5b). The complex boronates
`thus prepared are potent inhibitors of a-chymotrypsin. However, the affinity of 12 is neither time- nor
`pH-dependent, which would be expected for a covalent inhibitor, and it only shows a moderately
`increase in affinity compared to compounds 15, 16, and 17 that can not form a diester adduct. These
`results do not follow the predictions. The authors suggested that either boronate ester formation
`was not occurring, or that the energy derived from binding of the S0 binding fragment and boronate
`ester formation was not sufficient to offset the flexibility or binding characteristics of the linking
`group. They thus believed that a more detailed structural investigation was needed to confirm this
`interpretation.
`
`CFAD v. Anacor, IPR2015-01776, CFAD EXHIBIT 1060 - Page 5 of 23
`
`

`
`BORONIC ACID COMPOUNDS
`
`*
`
`351
`
`Figure5. Proposedbindingofserineproteasein (a) aternarycomplex withaboronicacidand (b) withanintramolecular tetrahedral
`boronic acid adduct.
`
`The bacterial b-lactamases catalyze the degradation of b-lactam antibiotics through an incredibly
`efficient hydrolysis of the lactam bond, which lead to antibiotic resistance to the b-lactam family of
`antibiotics. With bacterial resistance to b-lactam antibiotics continuing to increase, identification
`of new structural classes of b-lactamase inhibitors is of great clinical importance. TEM-1 is a
`representative member of the group 2b or class A b-lactamases that has achieved particular clinical
`notoriety.44 Based on the crystallographic structure of acyl-enzyme intermediate of TEM-1 bound
`to a substrate, penicillin G,45 (1R)-1-acetamido-2-(3-carboxyphenyl)ethane boronic acid (18) was
`designed to mimic the critical interactions observed in the penicillin G/TEM-1 complex and was
`found to be a potent TME-1 inhibitor.46 The structure of the b-lactamase TEM-1 has been solved in a
`complex with boronic acid (18) at 1.7 A˚ resolution, which suggested a novel transition state of the
`deacylation step in the b-lactamase-catalyzed reaction pathway. Using the same strategy, further
`structure-guided incorporation of additional side chain functionalities (20) or hydrogen bonding
`
`Figure6
`
`CFAD v. Anacor, IPR2015-01776, CFAD EXHIBIT 1060 - Page 6 of 23
`
`

`
`352
`
`*
`
`YANG, GAO, AND WANG
`
`groups (19) to maximize energetically favorable interactions of the boronate inhibitors and TEM-1
`b-lactamase was investigated.47 As designed, compounds 19 and 20 are highly effective deacylation
`transition state analogue inhibitors. The high-resolution crystallographic structures of these two
`inhibitors covalently bound to TEM-1 showed interesting and unanticipated changes in the active site
`area, including strong hydrogen bond formation, water displacement, and rearrangement of side
`chains, which provided new insights for the further design of this class of inhibitors.
`AmpC b-lactamase is a representative member of the class C b-lactamases that is among
`the most clinically important b-lactamase enzymes. Using the crystallographic structure of the
`m-aminophenylboronic acid–Escherichia coli AmpC b-lactamase complex, several types of boronic
`acid compounds were modeled into the AmpC binding site, and a total of 37 boronic acids were
`evaluated for b-lactamase inhibition. Among these inhibitors, benzo[b]thiophene-2-boronic acid is
`the most potent compound with Ki of 27 nM for AmpC.48 These boronic acid inhibitors were also
`found to potentiate the activity of b-lactam antibiotics. The X-ray crystallographic structure of
`benzo[b]thiophene-2-boronic acid in complex with AmpC was determined for probing the specific
`interactions between the enzyme and the inhibitor.49 The complex structure revealed several
`previously unknown interactions. The inhibitor was found to complement the conserved, R1-amide
`binding region of AmpC. Concerted interactions between one of the boronic acid oxygen atoms,
`Tyr-150, and an ordered water molecule implicate a mechanism for acid/base catalysis and a direction
`for hydrolytic attack in the enzyme catalyzed reaction. Further antimicrobial assays showed that
`benzo[b]thiophene-2-boronic acid significantly potentiated the activity of cephalosporin against
`AmpC-producing resistant bacteria.
`Hepatitis C virus (HCV) is the major cause of transfusion-associated hepatitis and community-
`acquired hepatitis worldwide.50 NS3 serine protease is one of the most intensively studied and best
`understood targets for antiviral therapy against HCV. Derived from previously identified potent
`hexapeptide aldehyde inhibitors of the HCV proteinase,51 a set of analogous boronic acid inhibitors
`were designed and synthesized. A high throughput method of synthesizing such boronic acid
`compounds was developed by using a resin bound diol as a boronic acid protecting group and
`immobilization linking point. Therefore, a hexapeptide boronic acid library was established by
`parallel synthesis using an Advanced Chemtech 496 synthesizer, from which compounds 21
`(Ki ¼ 80 nm) and 22 (Ki ¼ 80 nm) were found to be highly potent inhibitors of the HCV NS3
`protease.52
`
`Figure7
`
`CFAD v. Anacor, IPR2015-01776, CFAD EXHIBIT 1060 - Page 7 of 23
`
`

`
`BORONIC ACID COMPOUNDS
`
`*
`
`353
`
`Dipeptidyl peptidase IV (DPPIV) is a membrane-bound serine protease found on the surface
`of a variety of mammalian cells. Its inhibition could cause the suppression of the T-cell-mediated
`immune response both in vitro and in vivo. This enzyme cleaves a dipeptide from the amino
`terminus of polypeptides where the penultimate residue is proline. Based on dipeptides that possess a
`proline or proline mimic in the P1 position, different kinds of inhibitors of this enzyme have been
`developed.53 –55 Among them, several boronic acid dipeptides have been found to be exceptionally
`potent inhibitors. To reveal the structure-activity relationships associated with variations of the P2
`position of the dipeptide inhibitors, a series of proline boronic acid containing dipeptides (Fig. 8) were
`designed, synthesized, and assayed for their ability to inhibit DPPIV.56 It was found that inhibitory
`activity requires the (R)-stereoisomer of boroproline in the P1 position. A number of substituents
`tested, both polar and nonpolar amino acids, are tolerated in the P2 position except the unnatural
`amino acids, a,a-disubstituted amino acids and glycine.
`
`B. ProteasomeInhibitors
`
`The proteasome is a eukaryotic cytoplasmic protease complex that has several distinct catalytic sites.
`It plays a major role in cellular pathways for the breakdown and processing of proteins to peptides and
`amino acids.57 It is not surprising that defects of various components of this enzyme result in a range
`of human diseases including Angelman’s syndrome,58 cervical cancer,58 and Alzheimer’s disease59
`among others. As a result, these components provide attractive targets for therapeutic intervention.9
`The proteasome showing chymotrypsin-like activity was reported to be the first member of a newly
`identified class of threonine proteases. Recently, some selective and novel dipeptide aldehyde
`inhibitors of the chymotrypsin-like activity of the proteasome complex have been reported.60 Based
`on the dipeptide aldehyde inhibitors, a boropeptide compound 25 was designed and synthesized as an
`inhibitor of the chymotrypsin-like activity of proteasome with IC50 value of 8 nM.61
`Using the same strategy, a series of tri- and di-peptidyl boronic acid analogues were designed62
`based on the replacement of the corresponding aldehyde function of previously reported proteasome
`inhibitors.63 Bioassay of these compounds revealed that the incorporation of a boronic acid moiety
`in this series resulted in dramatically enhanced potency compared to the corresponding peptidyl
`aldehyde compounds. The enhancement in potency is presumed to be because of the formation of a
`stable tetrahedral boronic acid intermediate with the N-terminal threonine residue of the catalytically
`active proteasome b-subunits. In addition, compounds such as 2664 (PS-341) offer the advantages of
`low molecular weight and easy to synthesize. Furthermore, these compounds exhibited extremely
`high selectivity for the proteasome over common serine proteases. As shown in Table I, PS-341
`represents at least 500-fold selectivity for the 20S proteasome over 4 other proteases. This selectivity
`is because of the structural feature of PS-341. As a dipeptide, PS-341 does not fulfill the requirements
`of enzymes such as chymotrypsin and elastase for S3 and S4 subsite binding for optimal activity.
`The P1 position of this dipeptide boronic acid with a leucine residue does not match the preference of
`thrombin for basic residues at that position. All these results make compounds such as PS-341 very
`promising as new therapeutics for the treatment of cancer and inflammatory diseases. Many studies
`have been conducted aimed at developing PS-341 as a new agent in cancer therapy.65 –71 This
`compound is currently under Phase I clinical evaluation in advanced cancer patients.
`
`Figure8
`
`CFAD v. Anacor, IPR2015-01776, CFAD EXHIBIT 1060 - Page 8 of 23
`
`

`
`354
`
`*
`
`YANG, GAO, AND WANG
`
`Figure9
`
`To characterize the inhibition of peptidyl boronic acids against different forms of proteasomes
`and to explore their structure-activity relationships, a series of di- and tri-peptidyl boronic acids
`(such as (Bz)-Phe-boroLeu and (Cbz)-Leu-Leu-boroLeu pinacol ester) were designed. The peptidyl
`boronic acids were tested on the chymotrypsin-like activity of purified mammalian 20S and 26S
`proteasomes assayed with succinyl-Leu-Leu-Val-Tyr-amidomethylcoumarin as the substrate. The
`inhibition of 20S proteasomes is competitive but only slowly reversible. The Ki values for the best
`inhibitors are in the range 10–100 nM,72 but the compounds tested are much less effective on other
`proteasome activities measured with other substrates.
`
`C. Arginase Inhibitors
`
`Arginase plays a crucial role in the regulation of diverse metabolic pathways such as ureagenesis
`and nitric oxide (NO) biosynthesis. Recently, the synthesis and evaluation of nonreactive arginine
`analogues as possible enzyme inhibitors or receptor antagonists have attracted much attention. The
`X-ray crystal structure of rat liver arginase shows that the trimeric metaloenzyme contains a binuclear
`manganese cluster in the active site of each subunit required for maximal catalytic activity.73
`Using the information of X-ray crystal structure of the ternary arginase–ornithine–borate complex in
`which the manganese-bridging solvent molecule of the native enzyme is displaced by an oxygen of the
`tetrahedral borate anion, the first boronic acid analogue of arginine, 2(S)-amino-6-boronohexanoic
`acid (27), was designed, synthesized, and evaluated as an inhibitor for arginase.74 The inhibitory
`activity of compound 27 against Mn2 þ
`-arginase was evaluated using a radioactive assay to yield an
`IC50 of 0.8 mM. The crystal structure of the complex between arginase and 27 was also determined.75
`It was found that compound 27 binds as the tetrahedral boronate anion that mimics the intermediate of
`a metal-activated hydroxide mechanism. The tight binding and high specificity of compound 27 allow
`for the further study the physiological role of arginase in regulating the NO-dependent biological
`processes. Significant enhancement of nonadrenergic, noncholinergic nerve-mediated relaxation
`of penile corpus cavernosum smooth muscle was observed with compound 27 and these results
`suggested that arginase inhibition sustained L-arginine concentrations for NOS activity. Therefore,
`human penile arginase is a potential target for therapeutic intervention in the treatment of erectile
`dysfunction.
`
`Table I. Enzyme Inhibitory Profile of Compound 26 (PS-341)
`
`Enzyme
`
`20S proteasome
`Human leukocyte elastase
`Human cathepsin G
`Human chymotrypsin
`Thrombin
`
`Ki (nM)
`
`0.62
`2,300
`630
`320
`13,000
`
`CFAD v. Anacor, IPR2015-01776, CFAD EXHIBIT 1060 - Page 9 of 23
`
`

`
`BORONIC ACID COMPOUNDS
`
`*
`
`355
`
`Figure10
`
`Another similar boronic acid-based arginine analogue S-(2-boronoethyl)-L-cystein (28), in
`which a sulfur atom was introduced, was also designed and synthesized.76 Biological test showed it as
`a slow-binding competitive inhibitor of arginase with a Ki value of 0.4–0.6 mM. The X-ray crystal
`structure of the arginase-28 complex was also determined, and the structure of the complex revealed
`that the binding mode also mimics the tetrahedral intermediate in the arginine hydrolysis reaction as
`compound 27 does. Similarly, compound 28 also causes significant enhancement of NO-dependent
`smooth muscle relaxation in human penile corpus cavernosum tissue. Further biological studies77
`demonstrated that both compounds 27 and 28 are classical, competitive inhibitors of human type II
`arginase at pH 7.5 with Ki values of 0.25 and 0.31 mM, respectively. However, at pH 9.5, both were
`found to be slow-binding inhibitors of the enzyme with Ki values of 8.5 and 30 nM, respectively.
`It is expected that the pKa of the boronic moiety being8.5.19 Therefore, the compounds are expected
`in the tetrahedral form at pH 9.5 and trigonal form at pH 7.5. This could help to explain the enhanced
`potency at pH 9.5.
`
`D. OtherEnzymeInhibitors
`
`NO displays potent activities in the cardiovascular system as well as in the central and peripheral
`nervous systems. NO and its co-product L-citrulline are produced by the oxidation of L-arginine by
`NOS (Scheme 2). The selective modulation of NO biosynthesis offers the opportunity for therapeutic
`intervention of neurodegenerative diseases, among others. Based on the mechanism proposed for NO
`biosynthesis, two boronic acid analogues (29 and 30) of L-arginine were designed and synthesized as
`potential substrates or inhibitors of NOS.78 The abilities of the boro-L-arginine 29 and 30 to generate
`NO and to inhibit [3H]L-citrulline formation from [3H]L-arg were investigated in the presence of
`purified recombinant neuronal and inducible NOSs. The Na-acetyl derivative 30 did not lead to any
`significant NO formation and poorly inhibited L-citrulline formation (IC50 > 500 mM). However,
`the unprotected boro-L-arginine 29 selectively inhibited L-citrulline formation catalyzed by the
`inducible NOS (IC50 ¼ 50 mM) compared to the neuronal isoform (IC50 ¼ 300 mM). These results
`demonstrated the feasibility of using boronic acid compounds for the inhibition of this enzyme and the
`strict substrate specificity of NOSs.
`Cysteine proteases are involved in many disease processes. Developing effective inhibitors for
`these enzymes is of great pharmaceutical interest. The catalysis mechanism of cysteine protease is
`similar in a number of ways to that of serine protease. Both types of enzymes use a histidine-imidazole
`residue as a proton shuttle and both utilize nucleophilic catalysis. In the case of cysteine proteases, the
`cysteine thiol is the nucleophile, and for serine proteases the serine hydroxyl group is the nucleophile.
`
`Scheme2. Illustration of NO biosynthesis.
`
`CFAD v. Anacor, IPR2015-01776, CFAD EXHIBIT 1060 - Page 10 of 23
`
`

`
`356
`
`*
`
`YANG, GAO, AND WANG
`
`Figure11
`
`There were fewer examples of inhibition of cysteine proteases by boronic acid compounds, although
`boronic acids are often very potent transition state analogue inhibitors of serine proteases. Taking
`papain as the target protease, two peptidyl boronic acids (31 and 32) were prepared and assayed as
`potential transition state analogue inhibitors.79 The design of these peptidyl boronic acid substrates
`was derived from the substrate specificity and X-ray structures of papain. However, no inhibition
`could be detected at concentrations up to 10 mM. The reasons for the lack of inhibition were
`investigated with molecular modeling. Molecular mechanics and semi-empirical quantum mechanics
`calculations indicated that the absence of inhibition was because of boronic acid–cysteine protease
`tetrahedral complexes being thermodynamically less stable than their preceding non-covalent
`EI-complexes. In contrast, an analogous boronic acid–serine protease tetrahedral complex was
`calculated to be more stable than its precursor EI-complex because of the oxyanion hole stabilization
`of the tetrahedral intermediate. Therefore, such studies suggest that boronic acid compounds may not
`be good inhibitor candidates for cysteine proteases in general.
`g-Glutamyl transpeptidase (g-GT) is a membrane-associated enzyme that displays a crucial role
`in the metabolism of glutathione and is also a marker for neoplasia and cell transformation. This
`enzyme appears to function analogously to a serine protease. Based on the proposed transition state
`of g-GT, which was very similar to the proposed ternary complex of g-GT with serine and borate,
`L-2-amino-4-boronobutanoic acid (33) was designed and evaluated as a structural analog of the
`putative ternary complex. Compound 33 was found to be a potent inhibitor of the enzyme with a Ki
`value of 17 nM.80 The structural similarity of compound 33 to glutamate implicated that it might
`serve as a substrate for some glutamate-dependant enzymes or receptors. 13C-NMR studies demon-
`strated that transamination of pyruvate by compound 33 yielded alanine in the presence of glutamic
`pyruvic transaminase. Effects of compound 33 on the growth of cultured rat live cell line ARL-15C1
`and ARL-16T2 were investigated. At the high concentration of 1 mM of 33, there was a significant
`reduction in cell count for both cell lines. In contrast, growth inhibition of both cell lines by 10 mM
`compound 33 could be observed only in low cysteine media.
`
`4. BORONIC ACID COMPOUNDS AS BNCT AGENTS
`
`In addition to being used as enzyme inhibitors, boron compounds can also be used in BNCT,
`first proposed in 1936.32 BNCT was based on the unique feature of boron-10 that becomes lithium
`and emits a-particles upon irradiation with neutron. Because a-particles are very damaging and
`only travel a very short distance, they are ideal candidates for localized cancer radiation therapy.10
`Successful application of BNCT requires the development of boron compounds that specifically
`deliver substantial quantities of 10B to the target cells.81 –84
`
`CFAD v. Anacor, IPR2015-01776, CFAD EXHIBIT 1060 - Page 11 of 23
`
`

`
`BORONIC ACID COMPOUNDS
`
`*
`
`357
`
`Critical to the development of BNCT is the synthesis of boron-containing compounds that
`selectively target tumor cells. Numerous boron-containing compounds have been sy