In Silico Fragment-Based Design Identifies Subfamily B1 Metallo-β-lactamase Inhibitors

Article abstract of DOI:10.1021/acs.jmedchem.7b01728

Zinc ion-dependent β-lactamases (MBLs) catalyze the hydrolysis of almost all β-lactam antibiotics and resist the action of clinically available β-lactamase inhibitors. We report how application of in silico fragment-based molecular design employing thiol-

Full text of DOI:10.1021/acs.jmedchem.7b01728

Subscriber access provided by READING UNIV
Article
In silico fragment based design identifies
subfamily B1 metallo-#-lactamase inhibitors
Ricky Cain, Jürgen Brem, David Zollman, Michael A. McDonough, Rachel M.
Johnson, James Spencer, Anne Makena, Martine I. Abboud, Samuel Cahill,
Sook Y. Lee, Peter McHugh, Christopher J. Schofield, and Colin W. G. Fishwick
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.7b01728 • Publication Date (Web): 22 Dec 2017
Downloaded from http://pubs.acs.org on December 24, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 24
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
In silico fragment based design identifies subfamily
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
B1 metalloꢀβꢀlactamase inhibitors
Ricky Cain,[a] ‡ Jürgen Brem,[b] ‡ David Zollman,[b] Michael A. McDonough,[b] Rachel M.
Johnson,[a] James Spencer,[c] Anne Makena,[b] Martine I. Abboud,[b] Samuel Cahill, [b] Sook Y.
Lee,[b,d] Peter J. McHugh,[d] Christopher J. Schofield,[b]* and Colin W. G. Fishwick [a]*
[a] School of Chemistry, University of Leeds, Leeds, LS2 9JT, United Kingdom.
[b] Department of Chemistry, University of Oxford, 12 Mansfield Road, Oxford, OX1 3TA,
United Kingdom.
[c] School of Cellular and Molecular Medicine, University of Bristol, Biomedical Sciences
Building, Bristol, BS8 1TD, United Kingdom.
[d] Department of Oncology, Weatherall Institute of Molecular Medicine, University of Oxford,
John Radcliffe Hospital, Oxford, United Kingdom.
Eꢀmail: christopher.schofield@chem.ox.ac.uk and c.w.g.fishwick@leeds.ac.uk
ABSTRACT
Zinc ion dependent βꢀlactamases (MBLs) catalyze the hydrolysis of almost all βꢀlactam
antibiotics and resist the action of clinically available βꢀlactamase inhibitors. We report how
application of in silico fragmentꢀbased molecular design employing thiolꢀmediated metal
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 24
1
2
3
4
5
6
7
8
9
anchorage leads to potent MBL inhibitors. The new inhibitors manifest potent inhibition of
clinically important B1 subfamily MBLs, including the widespread NDMꢀ1, IMPꢀ1 and VIMꢀ2
enzymes; with lower potency, some of them also inhibit clinically relevant Class A and D serineꢀ
βꢀlactamases. The inhibitors show selectivity for bacterial MBL enzymes compared to human
MBL fold nucleases. Coꢀcrystallization of one inhibitor, which shows potentiation of
meropenem activity against MBLꢀexpressing Enterobacteriaceae, with VIMꢀ2 reveals an
unexpected binding mode, involving interactions with residues from conserved active site
bordering loops.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
INTRODUCTION
βꢀLactam containing inhibitors remain the single most important antibacterials.1, 2 βꢀLactams
inhibit transpeptidases (or penicillinꢀbinding proteins, PBPs) involved in cell wall biosynthesis
by reacting with a catalytically crucial nucleophilic serine reside. Bacteria have developed
multiple resistance mechanisms to βꢀlactams, most importantly by βꢀlactamase catalysed
hydrolysis. There are four classes of βꢀlactamases:3 classes A, C, and D are serineꢀβꢀlactamases
(SBLs), which are mechanistically related to the PBPs; class B are metalloꢀβꢀlactamases
(MBLs), employing (a) zinc ion(s) in catalysis.
Coꢀadministration of a βꢀlactam antibacterial with a class A βꢀlactamase inhibitor (clavulanic
acid,4 tazobactam,5 sulbactam) has successfully overcome some SBLꢀmediated resistance.
Avibactam,6 a non βꢀlactam broader spectrum inhibitor of class A, C, and some class D SBLs,
has recently been introduced into the clinic.7 As yet there are, however, no reports of clinically
useful class B MBL inhibitors (Figure 1a).8ꢀ11 MBLs are a concern because they catalyze
hydrolysis of almost all βꢀlactam antibacterials, including the carbapenems, which are commonly
used for treatment of severe and highly resistant infections. MBLs are divided into the B1, B2,
2
ACS Paragon Plus Environment

Page 3 of 24
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
and B3 subclasses based on metal site occupancy and sequence similarity.12ꢀ14 Subclass B1
MBLs, which employ two zinc ions (Znꢀ1/Znꢀ2) are the most clinically relevant and include the
New Delhi metalloꢀβꢀlactamase (NDMꢀ1), Verona integronꢀencoded MBL (VIM), and
Imipenemase (IMP) types of B1 MBLs.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Combating increasing resistance mediated by MBLs requires the identification of either novel
antibacterials or MBL resistant βꢀlactams (both likely substantial undertakings), or developing an
MBL inhibitor for coꢀadministration with an existing βꢀlactam. Here we report how the
application of in silico fragmentꢀbased15, 16 molecular design led to identification of potent
inhibitors of clinically relevant MBLs, some of which have (lower) activity against certain SBLs.
The in silico studies employed a structureꢀguided approach in which fragments were anchored to
the di zinc ion centre via a thiol and then optimised for binding via hydrophilic and electrostatic
interactions. Whilst such metal anchorage (or 'support ligand') based approach has been applied
in dynamic combinatorial based identification17 of ligands to MBLs, to our knowledge this is the
first application of an in silico fragmentꢀbased approach for production of specific MBL
inhibitors.
RESULTS AND DISCUSSION
De Novo Ligand Design
The de novo molecular design program SPROUT15, 16 generates potential ligands via docking
of fragments into targeted regions; subsequent userꢀdirected assembly using template spacers to
link these fragments together yields synthetically tractable scaffolds (Figure 1b). Analysis of an
crystal structure of NDMꢀ1 structure complexed with hydrolysed ampicillin (PDB code 3Q6X)14
using SPROUT16 identified active site regions for targeting with in silicoꢀgenerated fragments,
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 24
1
2
3
4
5
6
7
8
9
i.e. the regions proximal to: the Lys224 side chain, the Znꢀ2 metal ion, the nucleophilic
hydroxide / water that “bridges” the two zinc ions, and a conserved tryptophan (Trp87) that
makes hydrophobic interactions with the aromatic ampicillin C6 side chain18 (Figure 1c).
Importantly, while details differ between individual MBLs, these features are predicted to be
crucial for βꢀlactam binding by most, if not all, B1 MBLs.10 This design approach led to the
identification of a simple benzyl thiol 1 (Figure 1) as a potential NDMꢀ1 substrateꢀcompetitive
inhibitor. A detailed, stepꢀby step description of the in silico fragment based design novo design
is given in the SPROUT utilizing design sections (Supporting Information).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The in silico predicted interactions of 1 with NDMꢀ1 include interaction between its
carboxylate and the Lys224 side chain and the Znꢀ2 ion; anchoring of its thiol with both Znꢀ1
and Znꢀ2; and between the aryl ring of 1 with Trp87, via πꢀstacking interactions (Figure 2a).
Compound 1 was synthesized using Suzuki coupling (67%)19 of 5ꢀbromophthalide and phenyl
boronic acid to give a lactone which was ringꢀopened using Me3SiI to give the corresponding
iodide (62%). Iodide displacement using thiourea, followed by hydrolysis gave 1 (81%; see
Methods and Supporting Information for details).
Using a fluorogenic MBL assay20 1 was then evaluated for inhibition of recombinant forms of
NDMꢀ1, IMPꢀ1, and VIMꢀ2. Inhibition studies with 1 and subsequent compounds were
conducted in the presence of 0.01% Triton Xꢀ100 to reduce the potential of identifying protein
aggregating compounds21ꢀ23. The results (Table 1) reveal 1 to be an inhibitor with subꢀ
micromolar IC50 values for all three of the tested B1 MBLs, with IC50 values 232 – 234 nM;
these values compare favourably with the thiolꢀMBL inhibitors such as Lꢀcaptopril for which
IC50 values against B1 MBLs are in the micromolar range.10
4
ACS Paragon Plus Environment

Page 5 of 24
Journal of Medicinal Chemistry
1
2
3
4
Crystallographic studies
5
6
7
8
To investigate the binding of 1 with B1 MBLs, a structure of it complexed with VIMꢀ2 was
obtained by coꢀcrystallization (1.74 Å resolution; space group C21; see Supporting Information,
Figure S1 and Table S1). As predicted by the design studies, analysis of the crystal structure
confirms that the thiol of 1 acts as a metal binding ligand that displaces the proposed ‘hydrolytic’
hydroxide ion bridging the two metal ions (Figure 2b). The crystallographically observed
conformation of the biphenyl unit in VIMꢀ2 was not, however, as predicted for NDMꢀ1 (Figure 1
and Figure 2a); unexpectedly, the aryl carboxylate of 1 is positioned to interact with Trp87 of
VIMꢀ2 and, via a water molecule, with Asp119, rather than with Arg228 (analogous to Lys 224
in NDMꢀ1) (Figure 2b). In the VIMꢀ2 structure, 1 presents two binding faces, as also present in
the predicted βꢀlactam substrate, and observed hydrolysed βꢀlactam, binding modes.10 One face
is hydrophobic and interacts with the substantially hydrophobic L3 loop, via apparent πꢀstacking
of the aryl ring of 1 with Tyr67 and interactions with Trp87 (Figure 2b). The other binding face
of 1 involves electrostatic and hydrogen bonding interactions with residues from L10 loop,
apparent interaction between the Arg228 guanidino group and the phenyl ring of 1 (3.8 Å), and
potential interaction between one of the carboxylate oxygen’s of 1 with the side chain amide of
Asn233 (3.6 Å). The difference in the predicted NDMꢀ1 binding mode for 1 versus the
crystallographically observed binding mode of 1 with VIMꢀ2 may, in part, reflect the removal of
water molecules from the NDMꢀ1 structure prior to application of SPROUT to simplify the
calculations, as well as active site differences between the wo enzymes15, 16 (see Supporting
Information).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 24
1
2
3
4
Structure Activity Relationship Studies
5
6
7
8
Following the encouraging levels of inhibition observed with 1, we conducted structure
activity relationship (SAR) studies. Modification of the terminal phenyl ring of 1 (Table 1)
revealed no clear trend in potency when changing the electron density or steric bulk, via
incorporation of bromine (17) or fiveꢀmembered ring substituents (15 and 16), implying that the
πꢀstacking interaction with Tyr67 is not crucial for affinity. Similarly, positioning substituents in
the orthoꢀposition of the phenyl ring (4, 9 and 10), to investigate whether changes to the dihedral
angle between the aryl rings might affect affinity, did not yield substantial changes in potency.
The importance of the spatial arrangement of functional groups was investigated via 19
(Supporting Information, Figure S5) in which the methylene linking the thiol to the phenyl ring
is switched to link the carboxylic acid. The thiolactone 18, (Supporting Information, Figure S4)
was synthesised to test whether it undergoes hydrolysis to 1 (Supporting Information, Table S2).
Neither 18 nor 19 manifested significant activity against VIMꢀ2 (IC50 > 100 ꢁM) showing that
the arrangement of the functional groups is important and indicating that the thiolactone 18 does
not appear to be an efficient substrate for VIMꢀ2.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2ꢀ17 were evaluated for inhibition against NDMꢀ1 and IMPꢀ1 (Table 1 and Supporting
Information, Table S2); selected compounds (4ꢀ8) were screened against the model MBL BcII
and the ‘hybrid’ B1/B2 like MBL SPMꢀ1.24 The results reveal 4ꢀ8 to be active against BcII and
SPMꢀ1, with IC50 values 0.78 – 7.8 ꢁM (Supporting Information, Table S3). Moreover, except
for 11, all compounds manifested subꢀmicromolar IC50 values against IMPꢀ1 with several
compounds showing subꢀmicromolar activity against NDMꢀ1 (5, 6, 7, 8 and 13). Binding of
1
selected inhibitors to VIMꢀ2 and NDMꢀ1 was investigated by NMR:25, 26 The H CPMG NMR
19
results validated25 binding of 1 and 9 to VIMꢀ2 and F NMR26 was used to validate binding to
6
ACS Paragon Plus Environment

Page 7 of 24
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
NDMꢀ1 (Supporting Information, Figure S7). Consistent with the inhibition data, the NMR
analyses show that 1 and 9 bind with KD values < 1 µM to VIMꢀ2 and 9 out competes 11 for
binding to NDMꢀ1 (Supporting Information, Figure S8).
9
In the initial assessment of our inhibitors as hits we screened 4 and 9 against the SBL AmpC21; 4
and 9 showed activity against AmpC both in the presence and absence of Triton Xꢀ100
(Supplementary Information, Table S4). We then screened 1, 7, 8 and 9 against TEMꢀ1, AmpC,
and OXAꢀ10, which are representatives of Class A, C, and D SBLs, respectively (Supporting
Information, Table S4). Interestingly, 8 and 9 inhibited TEMꢀ1 and 9 inhibited OXAꢀ10 (IC50s
32.8, 5.5 and 37.3 µM, respectively, Supporting Information, Table S4). Although we have been
unable to obtain structures of these compounds complexed with SBLs, these results show some
of the inhibitors could interact with both SBLs and MBLs via related binding modes as shown in
Figure 2); indeed, our predicted binding mode of 1 to TEMꢀ1 and OXAꢀ10 is similar to that
observed crystallographically for a recently identified biphenyl carboxylic acid obtained through
a surface plasmon resonance (SPR) based screen for inhibitors of the OXAꢀ48 SBL27. However,
the unexpected crystallographically observed binding mode of 1 to VIMꢀ2 (Figure 2) means
caution should be exercised in predicting the SBL binding modes of our compounds.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Selectivity screening
Clinically relevant MBL inhibitors should show selectivity towards the bacterial MBLs over
human metalloꢀenzymes, including human MBLꢀfold enzymes.28 We tested 1, 5, 13 and 17
against two human MBL fold enzymes involved in DNA repair, DNA crossꢀlink repair enzymes
1A and B (DCLRE1A and DCLRE1B). DCLRE1A was only weakly inhibited by 1 (IC50 247 ±
154 µM) and DCLRE1B was inhibited by 1, 5 and 17 (IC50s 149 ± 27, 268 ±112 and 185 ± 50
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 24
1
2
3
4
5
6
7
8
µM, respectively), implying, at least partial, selectivity of the compounds towards the bacterial
MBLs. Compound 1 showed good aqueous solubility and underwent moderate metabolism in
human liver microsomes (see Supporting Information).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Antibacterial Activity
Compounds 1 and 11ꢀ17 were screened in minimal inhibitory concentration (MIC)
antimicrobial assays, in coꢀadministration studies with the clinically important carbapenem
meropenem, using MBL producing strains of E. coli and K. pneumoniae (species where NDMꢀ1
production is a growing clinical problem).29 Tests in the absence of inhibitors manifested
meropenem MICs against MBLꢀproducing bacteria of >128 ꢁg/ml, and that when tested alone 1
has no antibacterial activity within limits of detection (Table 2). Encouragingly, at 100 ꢁg/ml all
tested compounds, apart from 11, reduced the MIC of meropenem against NDMꢀ1ꢀproducing
strains. Compound 16 showed the strongest activity, reducing the meropenem MIC to 8 ꢁg/ml
against K. pneumoniae.
CONCLUSIONS
Overall, the results reveal how computer aided design can help generate broad spectrum MBL
inhibitors active against three types of clinically important B1 MBLs, as well as some SBLs. The
inhibitors can reduce MICs of meropenem against MBLꢀproducing bacteria, indicating they are
suitable for further development. Unlike some reported MBL inhibitors,30 our compounds bind
directly to the MBL active sites, i.e. do not function by removing the zinc ions, as shown by
NMR and crystallographic analyses.31 While utilizing its thiol to bind the MBL active site by
intercalating between the two B1 MBL subclass zinc ions has precedent,8 the
8
ACS Paragon Plus Environment

Page 9 of 24
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
crystallographically observed mode of binding 1 with VIMꢀ2 (Figure 1), has not been previously
observed. The difference between the crystallographically observed binding mode for 1 with
VIMꢀ2 and that predicted for NDMꢀ1 (Figure 2) indicates how differences in closely related
metalloꢀenzymes can apparently enable different binding modes of the same inhibitors, even
when they are anchored in the same manner, i.e. to the two active site zinc ions of the B1 MBLs.
It may also reflect of limitations in our current computational methodology, including with an
inability to deal with solvation.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The observation of activity for the designed compounds against several different types of
clinically relevant B1 MBLs is notable (Table 1) because inhibition of a spectrum of MBLs is a
likely prerequisite for clinical application of an MBL inhibitor. Although the compounds
described here are unlikely to be of direct clinical utility, the work suggests that computational
chemistry can have an important role in designing inhibitors that are active for use against both
current and future predicted antibacterial resistance mediated by βꢀlactamases. The need for an
ability to rapidly optimise against emerging resistance is highlighted by the identifications of the
multiple class B1 MBL variants already identified.12 Our approach may have utility in efficiently
identifying other inhibitor chemotypes involving alternative metalꢀcoordinating moieties, as well
as inhibitors not relying on metal coordination.
EXPERIMENTAL SECTION
The Supporting Information contains a complete general experimental section, including all
procedures and equipment used. Chemicals were from commonly used suppliers (Aldrich, Acros,
and Alfa Aesar) and used without purification. The purity of compounds submitted for screening
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 24
1
2
3
4
5
6
was ≥95% as determined by UV analysis of liquid chromatography (HPLC) chromatograms at
254 nM.
7
8
9
General procedure for the Suzuki cross coupling reaction
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The boronic acid (1.1 equiv.) was added to a stirred solution of 5ꢀbromophthalide (1.0 equiv.),
potassium carbonate (1.0 equiv.) and tetrakis(triphenylphosphine) palladium (0) (0.05 equiv.) in
THF (4 mL) and water (2 mL). The reaction mixture was heated at reflux for 12 h, then filtered
through Celite®, and concentrated in vacuo. The residue was diluted with water (20 mL), then
extracted using dichloromethane (3 × 20 mL). The combined organic extracts were dried
(MgSO4) and concentrated in vacuo to give a paleꢀyellow solid, which was purified using flash
column chromatography to afford the coupled product.
General procedure for lactone opening with iodotrimethylsilane forming the iodides
The benzofuranone (1.0 equiv.) was dissolved in dichloromethane (4.5 mL); TMSI (1.5 equiv.)
was then added and the reaction was refluxed under nitrogen for 3 h. After this time, the mixture
was cooled to room temperature and quenched with water (3 mL). The precipitate was isolated
by filtration washing with water (2 × 10 mL) to give the desired iodide product as an offꢀwhite
solid, which was used without further purification.
General procedure for introduction of the thiol32
A mixture of thiourea (1.1 equiv.) in THF (3 mL) was heated to reflux. To this mixture was
added the iodide product of the previous procedure (1.0 equiv.). The reaction mixture was heated
at reflux for 16 h under nitrogen. The reaction mixture was then cooled to room temperature and
concentrated in vacuo. The residue was reꢀsuspended in 2M aqueous NaOH (5 mL) and the
solution refluxed for 3 h. The reaction mixture then acidified with 2M aqueous HCl to ~pH2.
The colorless precipitate was isolated by filtration.
10
ACS Paragon Plus Environment

Page 11 of 24
Journal of Medicinal Chemistry
1
2
3
4
Representative synthesis: for compound 1
5
6
Synthesis of 5ꢀphenylꢀ1,3ꢀdihydoꢀ2ꢀbenzofuranone
7
8
9
Phenyl boronic acid (0.32 g, 2.58 mmol, 1.1 equiv.) was added to a stirred solution of 5ꢀ
bromophthalide (0.50 g, 2.35 mmol, 1.0 equiv.), potassium carbonate (0.33 g, 2.35 mmol, 1.0
equiv.) and tetrakis(triphenylphosphine) palladium (0) (0.14 g, 0.12 mmol, 0.05 equiv.) in THF
(4 mL) and water (2 mL). The reaction mixture was heated to reflux for 12 h, then filtered
through Celite®, and concentrated in vacuo. The residue was diluted with water (20 mL) and
extracted using dichloromethane (3 × 20 mL). The combined organic extracts were dried
(MgSO4), then concentrated in vacuo to give a paleꢀyellow solid, which was purified by flash
column chromatography to afford the desired product as a cream colored solid, which was used
without further purification. (0.33 g, 1.58 mmol, 67%) m.p.: 124.4ꢀ125.8 °C; Rf: 0.5 (2:1 petrolꢀ
EtOAc); δH (300 MHz, CDCl3): 7.98 (1H, d, J 7.6, 3ꢀH), 7.75 (1H, d, J 7.6, 4ꢀH), 7.68 (1H, s, 6ꢀ
H), 7.62 (2H, d, J 7.1, 2’ꢀH and 6’ꢀH), 7.47 (2H, app. t, J 7.1, 3’ꢀH and 5’ꢀH) 7.45 (1H, t, J 7.1,
4’ꢀH), 5.38 (2H, s, CH2O); δC (75 MHz, CDCl3): 147.5 (C=O), 147.3 (C1), 139.6 (C3), 129.1
(C3’’ and C5’), 128.6 (C2), 128.5 (C5),127.5 (C2’ and C6’) 126.1 (C4’), 124.5 (C1’), 120.6
(C6), 69.9 (CH2O); νmax/ cmꢀ1 (solid): 3473, 3070, 2924, 1747, 1615, 1452; HPLC: Tr= 3.00
(100% rel. area); m/z (ES): (Found: MNa+, 233.0573. C14H10O2 requires MNa, 223.0570).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Synthesis of 2ꢀiodomethylꢀ4ꢀphenylꢀbenzoic acid
5ꢀPhenylꢀ1,3ꢀdihydoꢀ2ꢀbenzofuranone (0.20 g, 0.95 mmol, 1.0 equiv.) was dissolved in
dichloromethane (4.5 mL). TMSI (0.20 mL, 1.43 mmol, 1.5 equiv.) was added and the reaction
was refluxed under nitrogen for 3 h. The mixture was then cooled to room temperature and
quenched with water (3 mL). The precipitate was isolated by filtration, washing with water (2 ×
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 24
1
2
3
4
5
6
7
8
9
10 mL) to give the desired product as a pale brown solid, which was used without further
purification. (0.20 g, 0.59 mmol, 62%) m.p.: 81.3ꢀ82.4 °C; Rf: 0.75 (CHCl3); δH (300MHz,
DMSOꢀd6): 7.94 (1H, d, J 8.3, 6ꢀH), 7.88 (1h, d, J 1.8, 3ꢀH), 7.73 (2H, d, J 7.3, 6’ꢀH and 2’ꢀH),
7.68 (1H, dd, J 8.3 and 1.8, 5ꢀH), 7.51 (2H, app. t, J 7.3, 5’ꢀH and 3’ꢀH), 7.44 (1H, t, J 7.3, 4’ꢀH),
5.09 (2H, s, CH2); νmax/ cmꢀ1 (solid): 2806, 2630, 1750, 1603, 1257, 909; m/z (ES): (Found:
[M+2NaꢀH]+, 382.9516. C14H11IO2 requires [M+2NaꢀH], 382.9515).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Synthesis of 4ꢀphenylꢀ2ꢀsulfonylmethylbenzoic acid (1)
A mixture of thiourea (0.16 g, 2.11 mmol, 1.1 equiv.) and 2ꢀIodomethylꢀ4ꢀphenylꢀbenzoic acid
(0.65 g, 1.92 mmol, 1.0 equiv.) in THF was heated at reflux for 16 h under an atmosphere of
nitrogen, then cooled to room temperature. The mixture was then concentrated in vacuo, then reꢀ
suspended in 2M aqueous NaOH (5 mL) and refluxed for 3 h. The resultant mixture then
acidified with 2M aqueous HCl to ~pH2. The pale yellow microcrystals were isolated by
filtration. C14H10Cl2O2S requires [M-H], 310.9706). (0.25 g, 1.01 mmol, 53%). m.p.: 168.2ꢀ
169.9 °C; Rf: 0.20 (19:1 DCM−MeOH); δH (300 MHz, DMSOꢀd6): 7.95 (1H, d, J 8.2, 6ꢀH),
7.77 (1H, d, J 2.0, 3ꢀH), 7.73 (2H, d, J 7.4, 6’ꢀH and 2’ꢀH), 7.65 (1H, dd, J 8.2 and 2.0, 5ꢀH),
7.52 (2H, app. t, J 7.4, 5’ꢀH and 3’ꢀH), 7.44 (1H, t, J 7.4, 4’ꢀH), 4.13 (2H, d, J 8.5, CH2), 2.87
(1H, t, J 8.5, SH); δC (75 MHz, DMSOꢀd6): 168.0 (C=O), 144.3 (C4), 143.6 (C1’), 138.7 (C2),
131.5 (C6), 129.0 (C5’ and C3’), 128.9 (C4’), 128.2 (C5), 127.8 (C1), 126.9 (C6’ and C2’),
124.9 (C3), 26.4 (CH2); νmax/ cmꢀ1 (solid): 2976, 1681, 1350, 1278, 756; HPLC: Tr= 2.88
(95% rel. area); m/z (ES): (Found: [MꢀH]ꢀ, 243.0477 C14H12O2S requires [MꢀH], 243.0485).
12
ACS Paragon Plus Environment

Page 13 of 24
Journal of Medicinal Chemistry
1
2
3
4
ASSOCIATED CONTENT
5
6
7
Supporting Information.
8
9
The following files are available free of charge.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Experimental procedures, characterization of intermediates and target compounds, description of
protein production and purification, biological assays and determination of IC50 values. (PDF)
Accession Codes
PDB ID code: VIMꢀ2 with bound 1 is 5K48. Authors will release the atomic coordinates and
experimental data upon article publication.
AUTHOR INFORMATION
Corresponding Authors
*Christopher
J.
Schofield
Phone:
+44
(0)
1865
275625
Eꢀmail:
christopher.schofield@chem.ox.ac.uk and Colin W. G. Fishwick Phone: +44 (0) 1133 436510 Eꢀ
mail: c.w.g.fishwick@leeds.ac.uk
Author Contributions
‡R.C. and J.B. contributed equally to this work.
Notes
The authors declare no competing financial interests.
Acknowledgements
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 24
1
2
3
4
5
6
7
8
9
We thank the Medical Research Council (MRC)/Canadian Grant G1100135 (RC, JS, CWGF,
JB and CJS), the Wellcome Trust, and the Biotechnology and Biological Research Council for
funding (BBSRC). This work was supported by Cancer Research UK (CRꢀUK) grant number
C5255/A18085, through the Cancer Research UK Oxford Centre (JB).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ABBREVIATIONS USED
MBL, metalloꢀβꢀlactamase; SBL, serineꢀβꢀlactamase; TMSI, iodotrimethylsilane.
14
ACS Paragon Plus Environment

Page 15 of 24
Journal of Medicinal Chemistry
1
2
3
4
5
6
References
7
8
9
1.
2.
Hamad, B. The antibiotics market. Nat. Rev. Drug Discov. 2010, 9, 675ꢀ676.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Molstad, S.; Lundborg, C. S.; Karlsson, A. K.; Cars, O. Antibiotic prescription rates vary
markedly between 13 European countries. Scand. J. Infect. Dis. 2002, 34, 366ꢀ371.
3.
4.
Ambler, R. P. The structure of βꢀlactamases. Phil. Trans. R. Soc. B 1980, 289, 321ꢀ331.
Arulanantham, H.; Kershaw, N. J.; Hewitson, K. S.; Hughes, C. E.; Thirkettle, J. E.;
Schofield, C. J. ORF17 from the clavulanic acid biosynthesis gene cluster catalyzes the ATPꢀ
dependent formation of Nꢀglycylꢀclavaminic acid. J. Biol. Chem. 2006, 281, 279ꢀ287.
5.
interactions and resistance. Pharmacol. Ther. 1999, 83, 141ꢀ151.
6. Li, H.; Estabrook, M.; Jacoby, G. A.; Nichols, W. W.; Testa, R. T.; Bush, K. In vitro
Yang, Y.; Rasmussen, B. A.; Shlaes, D. M. Class A βꢀlactamasesꢀenzymeꢀinhibitor
susceptibility of characterized βꢀlactamaseꢀproducing strains tested with avibactam
combinations. Antimicrob. Agents. Chemother. 2015, 59, 1789ꢀ1793.
7.
Stone, G. G.; Bradford, P. A.; Newell, P.; Wardman, A. In vitro activity of ceftazidimeꢀ
avibactam against isolates in a phase 3 openꢀlabel clinical trial for complicated intraꢀabdominal
and urinary tract infections caused by ceftazidimeꢀnonsusceptible gramꢀnegative pathogens.
Antimicrob. Agents. Chemother. 2017, 61, e01820ꢀ16.
8.
lactamases. Sci. Pharm. 2013, 81, 309ꢀ327.
9. Cahill, S. T.; Cain, R.; Wang, D. Y.; Lohans, C. T.; Wareham, D. W.; Oswin, H. P.;
Faridoon; UI Islam, N. An update on the status of potent inhibitors of metalloꢀβꢀ
Mohammed, J.; Spencer, J.; Fishwick, C. W.; McDonough, M. A.; Schofield, C. J.; Brem, J.
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 24
1
2
3
4
Cyclic boronates inhibit all classes of βꢀlactamases. Antimicrob. Agents. Chemother. 2017, 61,
5
6
e02260ꢀ16.
7
8
10.
Brem, J.; van Berkel, S. S.; Zollman, D.; Lee, S. Y.; Gileadi, O.; McHugh, P. J.; Walsh,
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
T. R.; McDonough, M. A.; Schofield, C. J. Structural basis of metalloꢀβꢀlactamase inhibition by
captopril stereoisomers. Antimicrob. Agents. Chemother. 2015, 60, 142ꢀ150
11.
Brem, J.; van Berkel, S. S.; Aik, W.; Rydzik, A. M.; Avison, M. B.; Pettinati, I.; Umland,
K. D.; Kawamura, A.; Spencer, J.; Claridge, T. D.; McDonough, M. A.; Schofield, C. J.
Rhodanine hydrolysis leads to potent thioenolate mediated metalloꢀβꢀlactamase inhibition. Nat.
Chem. 2014, 6, 1084ꢀ1090.
12.
βꢀlactams? Lancet Infect. Dis. 2011, 11, 381ꢀ393.
13. Bebrone, C.; Delbruck, H.; Kupper, M. B.; Schlomer, P.; Willmann, C.; Frere, J. M.;
Cornaglia, G.; Giamarellou, H.; Rossolini, G. M. Metalloꢀβꢀlactamases: a last frontier for
Fischer, R.; Galleni, M.; Hoffmann, K. M. The structure of the dizinc subclass B2 metalloꢀβꢀ
lactamase CphA reveals that the second inhibitory zinc ion binds in the histidine site.
Antimicrob. Agents. Chemother. 2009, 53, 4464ꢀ4471.
14.
mechanism. FASEB J. 2011, 25, 2574ꢀ2582.
15. Gillet, V.; Johnson, A. P.; Mata, P.; Sike, S.; Williams, P. SPROUT: a program for
structure generation. J. Comput. Aided. Mol. Des. 1993, 7, 127ꢀ153.
16. Gillet, V. J.; Newell, W.; Mata, P.; Myatt, G.; Sike, S.; Zsoldos, Z.; Johnson, A. P.
Zhang, H.; Hao, Q. Crystal structure of NDMꢀ1 reveals a common βꢀlactam hydrolysis
SPROUT: recent developments in the de novo design of molecules. J. Chem. Inf. Comput. Sci.
1994, 34, 207ꢀ217.
16
ACS Paragon Plus Environment

Page 17 of 24
Journal of Medicinal Chemistry
1
2
3
4
17.
Demetriades, M.; Leung, I. K.; Chowdhury, R.; Chan, M. C.; McDonough, M. A.; Yeoh,
5
6
7
8
K. K.; Tian, Y. M.; Claridge, T. D.; Ratcliffe, P. J.; Woon, E. C.; Schofield, C. J. Dynamic
combinatorial chemistry employing boronic acids/boronate esters leads to potent oxygenase
inhibitors. Angew. Chem. Int. Ed. Engl. 2012, 51, 6672ꢀ6675.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
18.
Brem, J.; Cain, R.; Cahill, S.; McDonough, M. A.; Clifton, I. J.; JimenezꢀCastellanos, J.
C.; Avison, M. B.; Spencer, J.; Fishwick, C. W.; Schofield, C. J. Structural basis of metalloꢀβꢀ
lactamase, serineꢀβꢀlactamase and penicillinꢀbinding protein inhibition by cyclic boronates. Nat.
Commun. 2016, 7, 12406.
19.
Miyaura, N.; Yamada, K.; Suzuki, A. A new stereospecific crossꢀcoupling by the
palladiumꢀcatalyzed reaction of 1ꢀalkenylboranes with 1ꢀalkenyl or 1ꢀalkynyl halides.
Tetrahedron Lett. 1979, 20, 3437ꢀ3440.
20.
van Berkel, S. S.; Brem, J.; Rydzik, A. M.; Salimraj, R.; Cain, R.; Verma, A.; Owens, R.
J.; Fishwick, C. W.; Spencer, J.; Schofield, C. J. Assay platform for clinically relevant metalloꢀβꢀ
lactamases. J. Med. Chem. 2013, 56, 6945ꢀ6953.
21.
Irwin, J. J.; Duan, D.; Torosyan, H.; Doak, A. K.; Ziebart, K. T.; Sterling, T.; Tumanian,
G.; Shoichet, B. K. An aggregation advisor for ligand discovery. J. Med. Chem. 2015, 58, 7076ꢀ
7087.
22.
Med. Chem. 2006, 49, 7274ꢀ7277.
23. Feng, B. Y.; Simeonov, A.; Jadhav, A.; Babaoglu, K.; Inglese, J.; Shoichet, B. K.; Austin,
Shoichet, B. K. Interpreting steep doseꢀresponse curves in early inhibitor discovery. J.
C. P. A highꢀthroughput screen for aggregationꢀbased inhibition in a large compound library. J.
Med. Chem. 2007, 50, 2385ꢀ2390.
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 24
1
2
3
4
24.
Brem, J.; Struwe, W. B.; Rydzik, A. M.; Tarhonskaya, H.; Pfeffer, I.; Flashman, E.; van
5
6
7
8
Berkel, S. S.; Spencer, J.; Claridge, T. D.; McDonough, M. A.; Benesch, J. L.; Schofield, C. J.
Studying the activeꢀsite loop movement of the Sao Paolo metalloꢀβꢀlactamaseꢀ1. Chem. Sci.
2015, 6, 956ꢀ963.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
25.
cascade for fragmentꢀbased drug discovery. Nat. Protoc. 2013, 8, 2309ꢀ2324.
26. Gee, C. T.; Arntson, K. E.; Urick, A. K.; Mishra, N. K.; Hawk, L. M.; Wisniewski, A. J.;
Mashalidis, E. H.; Sledz, P.; Lang, S.; Abell, C. A threeꢀstage biophysical screening
Pomerantz, W. C. Proteinꢀobserved 19FꢀNMR for fragment screening, affinity quantification and
druggability assessment. Nat. Protoc. 2016, 11, 1414ꢀ1427.
27.
Lund, B. A.; Christopeit, T.; Guttormsen, Y.; Bayer, A.; Leiros, H. K. Screening and
design of inhibitor scaffolds for the antibiotic resistance oxacillinaseꢀ48 (OXAꢀ48) through
surface plasmon resonance screening. J. Med. Chem. 2016, 59, 5542ꢀ5554.
28.
Pettinati, I.; Brem, J.; McDonough, M. A.; Schofield, C. J. Crystal structure of human
persulfide dioxygenase: structural basis of ethylmalonic encephalopathy. Hum. Mol. Genet.
2015, 24, 2458ꢀ2469.
29.
Delhi metalloꢀβꢀlactamase (NDM): a threat to public health. BMC Microbiol. 2017, 17, 101.
30. McGeary, R. P.; Tan, D. T.; Schenk, G. Progress toward inhibitors of metalloꢀβꢀ
lactamases. Future Med. Chem. 2017, 9, 673ꢀ691.
31. Li, G. B.; Abboud, M. I.; Brem, J.; Someya, H.; Lohans, C. T.; Yang, S. Y.; Spencer, J.;
Khan, A. U.; Maryam, L.; Zarrilli, R. Structure, genetics and worldwide spread of New
Wareham, D. W.; McDonough, M. A.; Schofield, C. J. NMRꢀfiltered virtual screening leads to
nonꢀmetal chelating metalloꢀβꢀlactamase inhibitors. Chem. Sci. 2017, 8, 928ꢀ937.
18
ACS Paragon Plus Environment

Page 19 of 24
Journal of Medicinal Chemistry
1
2
3
4
32.
Cossar, B. C.; Fournier, J. O.; Fields, D. L.; Reynolds, D. D. Preparation of Thiols. J.
5
6
Org. Chem. 1962, 27, 93ꢀ95.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 24
1
2
3
4
5
6
7
8
9
Figure 1. Design of broad spectrum MBL inhibitors. A) Known MBL binders (inhibitors) and
compound 1.8ꢀ11 B) Key elements used in the SPROUT computational design process leading to
the identification of putative broad spectrum MBL inhibitors. C) Outline mechanism for SBLs
and MBLs as shown for a penicillin substrate.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
20
ACS Paragon Plus Environment

Page 21 of 24
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
Figure 2. Predicted potential binding modes for the designed NDMꢀ1 inhibitors and the
crystallographically observed binding mode for VIMꢀ2. (A), (C), and (D) shows views from
modelling of 1 into the active sites of NDMꢀ1, TEMꢀ1, and OXAꢀ10, respectively (see
Supporting Information for details); (B) shows view from a crystal structure of 2 coꢀcrystalized
with VIMꢀ2 (PDB ID: 5K48).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 24
1
2
3
4
5
6
Table 1. IC50 values of de novo designed thiols against clinically relevant B1 MBLs. Assays
7
8
9
were conducted as reported20.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
No.
X
cLogPa
IC50 NDMꢀ1
(µM)
IC50 VIMꢀ2
(ꢁM)
IC50 IMPꢀ1
(ꢁM)
C6H5
2ꢀClC6H4
3ꢀClC6H4
4ꢀClC6H4
3ꢀFC6H4
3.76
4.37
4.37
4.37
3.90
3.90
4.97
4.64
5.52
4.75
4.72
3.96
2.82
3.54
2.88
5.59 ± 0.05
2.85 ± 0.03
0.31 ± 0.05
0.42 ± 0.02
0.59 ± 0.02
0.71 ± 0.01
1.16 ± 0.05
8.82 ± 0.05
43.95 ± 0.06
2.42 ±0.03
0.57 ± 0.02
7.59 ± 0.03
10.21 ± 0.04
6.78 ± 0.09
> 100
0.23 ± 0.04
0.18 ± 0.09
0.07 ± 0.05
0.10 ± 0.03
0.20 ± 0.06
0.23 ± 0.03
0.04 ± 0.03
0.15 ± 0.08
0.45 ± 0.07
0.07 ± 0.03
0.10 ± 0.03
0.18 ± 0.16
0.36 ± 0.02
0.17 ± 0.12
0.17 ± 0.01
0.23 ± 0.04
0.30 ± 0.06
0.14 ± 0.05
0.06 ± 0.05
0.39 ± 0.08
0.30 ± 0.09
0.37 ± 0.06
0.48 ± 0.04
1.72 ± 0.04
0.22 ± 0.03
0.05 ± 0.03
0.32 ± 0.03
0.39 ± 0.03
0.18 ± 0.04
0.80 ± 0.03
1
4
5
6
7
4ꢀFC6H4
8
2,3ꢀClC6H3
2ꢀCF3C6H4
3,5ꢀCF3C6H3
2ꢀNaphthyl
4ꢀEtC6H4
3ꢀEtOꢀC6H4
2ꢀFuryl
9
10
11
12
13
14
15
16
17
3ꢀThienyl
Br
acLogP values were calculated using MarvinSketch 17.28 (ChemAxon).
b The IC50 values are shown as the mean ± SD from minimum three separate experiments.
22
ACS Paragon Plus Environment

Page 23 of 24
Journal of Medicinal Chemistry
1
2
3
4
5
Table 2. Determination of Meropenem MICs against bacterial strains when coꢀadministered with
6
7
the designed βꢀlactamase inhibitors. Assays were conducted as reported11.
8
9
Compound
Number
X
Meropenem MIC (ꢁg/mL)
NDMꢀ1 producing K.
Meropenem MIC (ꢁg/mL)
NDMꢀ1 producing E. coli
(MG1655)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
pneumoniae (ATCC 5055)
> 128
32
> 128
16
Meropenem
C6H5
Br
1
64
32
17
11
12
13
14
15
16
3,5ꢀCF3C6H3
2ꢀNaphthyl
4ꢀEtꢀC6H4
3ꢀEtOꢀtC6H4
2ꢀFuryl
>128
32
>128
32
64
32
16
64
64
32
3ꢀThienyl
8
16
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 24
1
2
3
4
5
Table of Contents Graphic
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
24
ACS Paragon Plus Environment