Identification and Structure-Activity Relationships of Novel Compounds that Potentiate the Activities of Antibiotics in Escherichia coli

Article abstract of DOI:10.1021/acs.jmedchem.7b00453

In Gram-negative bacteria, efflux pumps are able to prevent effective cellular concentrations from being achieved for a number of antibiotics. Small molecule adjuvants that act as efflux pump inhibitors (EPIs) have the potential to reinvigorate existing antibiotics that are currently ineffective due to efflux mechanisms. Through a combination of rigorous experimental screening and in silico virtual screening, we recently identified novel classes of EPIs that interact with the membrane fusion protein AcrA, a critical component of the AcrAB-TolC efflux pump in Escherichia coli. Herein, we present initial optimization efforts and structure-activity relationships around one of those previously described hits, NSC 60339 (1). From these efforts we identified two compounds, SLUPP-225 (17h) and SLUPP-417 (17o), which demonstrate favorable properties as potential EPIs in E. coli cells including the ability to penetrate the outer membrane, improved inhibition of efflux relative to 1, and potentiation of the activity of novobiocin and erythromycin.

Full text of DOI:10.1021/acs.jmedchem.7b00453

Subscriber access provided by University of Newcastle, Australia
Article
Identification and Structure-Activity Relationships of Novel Compounds
that Potentiate the Activities of Antibiotics in Escherichia coli
Keith M. Haynes, Narges Abdali, Varsha Jhawar, Helen I. Zgurskaya, Jerry M. Parks, Adam
T. Green, Jerome Yves Baudry, Valentin V. Rybenkov, Jeremy C. Smith, and John K Walker
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.7b00453 • Publication Date (Web): 26 Jun 2017
Downloaded from http://pubs.acs.org on June 27, 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 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Identification and Structure-Activity Relationships of Novel Compounds that Potentiate the
Activities of Antibiotics in Escherichia coli
1
2
2
2
3,4
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
Keith M. Haynes, Narges Abdali, Varsha Jhawar, Helen I. Zgurskaya, Jerry M. Parks, Adam
3
,4
3,5
2
3,5
1,6*
T. Green, Jerome Baudry, Valentin V. Rybenkov, Jeremy C. Smith and John K. Walker
1
Department of Pharmacological & Physiological Science, Saint Louis University School of
Medicine, St Louis, MO, 63104, USA
2
Department of Chemistry and Biochemistry, University of Oklahoma, Norman, OK 73019, USA
UT/ORNL Center for Molecular Biophysics, Biosciences Division, Oak Ridge National
3
Laboratory, Oak Ridge, TN 37831, USA
4
Graduate School of Genome Science and Technology, University of Tennessee, Knoxville, TN,
3
7996, USA
5
Department of Biochemistry and Cellular and Molecular Biology, University of Tennessee,
Knoxville, TN, 37996, USA
6
Department of Chemistry, Saint Louis University, St. Louis, MO, 63104, USA
*
Corresponding author. Email: walkerjk@slu.edu
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 60
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ABSTRACT: In Gramꢀnegative bacteria, efflux pumps are able to prevent effective cellular
concentrations from being achieved for a number of antibiotics. Small molecule adjuvants that act
as efflux pump inhibitors (EPIs) have the potential to reinvigorate existing antibiotics that are
currently ineffective due to efflux mechanisms. Through a combination of rigorous experimental
screening and in silico virtual screening we recently identified novel classes of EPIs that interact
with the membrane fusion protein AcrA, a critical component of the AcrABꢀTolC efflux pump in
E. coli. Herein, we present initial optimization efforts and structureꢀactivity relationships around
one of those previously described hits, NSC 60339 (1). From these efforts we identified two
compounds, SLUPPꢀ225 (17h) and SLUPPꢀ417 (17o), which demonstrate favorable properties as
potential EPIs in E. coli cells including the ability to penetrate the outer membrane, improved
inhibition of efflux relative to 1 and potentiation of the activity of novobiocin and erythromycin.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment

Page 3 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
INTRODUCTION
The number of bacterial pathogens exhibiting resistance to at least one or more classes of
antibiotics is on the rise and poses an everꢀgrowing health concern. Resistance to almost all major
antibacterial drug classes has been observed in both Gramꢀnegative and Gramꢀpositive species.
The situation is particularly acute with the significant increase in difficult to treat nosocomial
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
1
,2
infections resulting from Gramꢀnegative bacteria. In Gramꢀnegative bacteria one of the primary
obstacles to overcome for any antibacterial agent is the synergistic efforts of the
3
lipopolysaccharide (LPS) rich outer membrane (OM). The OM is effective at limiting permeation
of drugs into the bacterial cell and efflux pumps located in the cytoplasmic membrane which
extrude compounds that do cross the OM back out of the cell before they can reach their potential
4
,5
site of action. Bacterial strains that develop mutations in the OM porins or overexpress efflux
6
pumps tend to exhibit greater drug resistance. As a result, strategies to develop adjuvants that
inhibit the action of bacterial efflux pumps (EPI) have the ability to potentiate (i.e., improve
antibacterial activity relative to no adjuvant being present) the effectiveness of existing
7
ꢀ9
antibacterial agents.
10,11
In E. coli the major efflux pump contributing to resistance is the AcrABꢀTolC tripartite pump.
This class of efflux pumps consists of three separate proteins that assemble to form the active
1
1,12
efflux pump complex.
The AcrB transporter, which belongs to the Resistance Nodulation Cell
Division (RND) superfamily of proteins, is responsible for the binding of substrates for efflux. The
porin TolC is the channel that spans the OM and allows for passage of substrates from the cell into
the extracellular milieu. These two proteins are unable to fully engage on their own but require the
membrane fusion protein (MFP) AcrA to bridge the gap between them and form a complete
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 60
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
11ꢀ15
channel that excludes the periplasm.
The final composition of the active pump complex is
1
6
believed to be a 3:6:3 ratio of AcrB:AcrA:TolC. The AcrB substrateꢀbinding protein has been
8
,17ꢀ19
studied extensively as a potential therapeutic target for the development of new EPIs.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
20
However, until our recent disclosure, there were no reports, to the best of our knowledge, of
specifically targeting the MFP AcrA as a strategy for the development of EPIs.
We were interested in determining whether assembly of the AcrABꢀTolC efflux pump system, and
thus its function, could be disrupted via binding of a small molecule to AcrA. Toward that end, the
Diversity Set V from the DTP program of the NCI/NIH was screened experimentally and virtual
docking was used to identify EPIs that potentially work by targeting AcrA. Experimental screening
involved initial evaluation of test compounds in an E. coli cell line (WTꢀPore) engineered to
2
1
express large pores (~ 2.4 nm) in the OM. This approach allowed us to identify compounds that
are potential EPIs and might have been overlooked otherwise due to poor OM penetration.
Specifically, compounds were tested in the presence of novobiocin at a concentration (16 ꢀg/mL)
that is ¼ of its minimum inhibitory concentration (MIC) in that strain. Test compounds showing
MICs ~100 ꢀM or better were further evaluated to determine if they were acting as EPIs (vide
2
0, 22
infra). The other key step involved using surface plasmon resonance
(SPR) to determine
whether compounds could bind to AcrA. Compounds that potentiated novobiocin and were also
found to bind to AcrA in the SPR experiments were prioritized as potential hits.
23
From the screening efforts, NSC 60339 (1), (Figure 1), a polybasic terephthalic acid derivative
2
4,25
studied previously
as a potential cancer chemotherapeutic agent, stood out as one of the more
attractive candidates for medicinal chemistry followꢀup. This compound’s profile was consistent
with an EPI potentially working via binding to AcrA. Testing of 1 in WTꢀPore cells, in the
ACS Paragon Plus Environment

Page 5 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
presence of novobiocin, resulted in an observed MIC of 25 ꢀM while in the absence of novobiocin
it had only weak antibacterial activity (MIC = 200 ꢀM). In SPR experiments with AcrA protein
and 50 ꢀM of compound 1 a strong binding signal was observed. Binding of 1 to AcrA was further
established through in vivo proteolysis experiments with AcrA. It was found that the cleavage
products of AcrA differed when exposed to 1 compared to control experiments done in the absence
of 1, suggesting 1 binds to AcrA and induces structural changes to AcrA as a result of binding.
Furthermore, 1 was shown to dramatically reduce the rate of efflux of a known AcrABꢀTolC
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
26
substrate, bisbenzamide dye Hꢀ33342 (HT), in E. coli cells. Taken together, these findings
suggest 1 is an EPI that functions at least in part through binding of AcrA. Lastly, in terms of
synthetic feasibility, 1 is a relatively simple terephthalic acid analog derived from 2ꢀ
chloroterephthalic acid (2a) and aniline 3 (Figure 2). As a result, we envisioned rapid entry into
new analogs to explore critical structureꢀactivity relationships (SAR) around this compound.
Herein we report the synthesis, pharmacological evaluation, and SAR of several new analogs of 1,
some of which demonstrated an equivalent or better profile for potentiation, efflux inhibition
and/or cell penetration. These studies are expected to pave the way for identifying more potent
efflux pump inhibitors in the future.
RESULTS AND DISCUSSION
Design. To aid in the selection of novel compounds, we used a previously generated model of
20
AcrA (Figure 3A) to examine relevant possible poses of 1 at the hinge site, which is located at
the interface of the α-helical hairpin and the lipoyl domains. The highest ranked pose of 1 was
used to suggest possible modifications that may improve binding affinities. The calculated pKa
values for the dihydroimidazoline groups are > 9, suggesting that 1 is dicationic at pH 7. The two
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 60
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
dihydroimidazolinium groups were calculated to interact with the backbone carbonyls of Ile 65
and Ala 172 (Figure 3B). In addition, the orientation of the chloro group, which interacts with the
side chain of Val 176, also suggests that modification of this group may alter binding and activity.
Our initial set of analogs was designed to test some of these observations. However, we stress that
although the binding pose at the hinge site provided guidance for modifying the central and
dihydroimidazoline rings, the location of the binding site for 1 has not been determined
definitively.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Chemistry. General routes for the preparation of new analogs are outlined in Schemes 1-4.
The simple terephthalic acid derivatives 4a-g were prepared following one of the two
straightforward procedures shown in Scheme 1. Starting with the desired terephthalic acids 2a or
2b and combining with various anilines using a standard peptide coupling approach mediated by
Oꢀ(benzotriazolꢀ1ꢀyl)ꢀN,N,N’N’,ꢀtetramethyluronium tetrafluoroborate (TBTU) typically led to the
desired products in good overall yields. Alternatively, reaction of the appropriate terephthalic acid
chlorides with the desired anilines in the presence of an amine base such as N,Nꢀ
diisopropylethylamine also gave the corresponding bisꢀamide analogs 4a-g in good yield.
However, though perhaps not unexpectedly, neither of these protocols could be extended to
coupling reactions involving the dihydroimidazolineꢀsubstituted aniline, 3. For example, when
trying to resynthesize 1, coupling of acid 2a with aniline 3 using TBTU led exclusively to amide 5.
The same outcome was also realized when using the diacid chloride of 2a to couple with 3. None
of the desired product from amide formation with the aniline nitrogen was observed (Scheme 1).
Interestingly, when amide 5 was profiled in biological testing it was found to potentiate
novobiocin in the WTꢀPore cells (MPC = 50 ꢀM).
4
ACS Paragon Plus Environment

Page 7 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
To prepare analogs 9a-f from Table 1, as well as to resynthesize 1, we employed either of the two
general procedures shown in Scheme 2. In one procedure, the starting diacids 2a-f were converted
to the corresponding diacid chloride intermediates of type 6 which could subsequently be
converted to the nitrileꢀcontaining amides 8. The resulting nitriles, 8, were then converted to the
dihydroimidazoline moiety with ethylenediamine and sodium sulfide following the procedure of Ji
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
2
7
et al to give the desired products 9. The nitriles, 8, could also be prepared via a standard amide
formation protocol using TBTU and either 3ꢀamino (7a) or 4ꢀaminobenzonitrile (7b). It was also
possible to form the desired dihydroimidazoline compounds via a direct coupling of aniline 3 with
the diacid chlorides, 6, in varying yields when the reaction was conducted in glacial acetic acid.
The aniline was dissolved in glacial acetic acid first, followed by subsequent addition of the acid
28
chloride to this mixture. In some cases the solubility of the monoꢀaddition product in glacial
acetic acid was poor and the monoꢀaddition product precipitated out of solution before the desired
bisꢀaddition product could form. In these cases, the reaction was run in refluxing toluene with a
stoichiometric amount of acetic acid.
Analogs 11 and 13, in which the central phenyl has been replaced with a saturated linker were
prepared from transꢀ1,4ꢀcyclohexanedicarboxylic acid (10) and succinic acid (12) respectively, as
outlined in Scheme 3. Formation of the nitrile amides followed by conversion of the nitriles to the
dihydroimidazoline group les to analogs 11 and 13. The 1,3ꢀdiamide analog, 16, was also prepared
in a similar fashion starting from isophthalic acid (15). The known reversed amide compound 14
29
was prepared as described previously by Dong et al.
The monoꢀamide derivatives 17a-q shown in Table 3 were prepared as outlined in Scheme 4. In
general, aniline 3 was coupled directly to a suitable acid chloride using the glacial acetic acid
procedure. The acid chloride, 18, needed to prepare analog 17c, was made via coupling of
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 60
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
terephthalic acid (2b) with benzyl amine (1.3 eq.) mediated by TBTU. This gave rise to a mixture
of the desired monoꢀbenzylamide and the dibenzyl amide. This mixture was easily separated and
the resulting acid was converted to 18 with oxalyl chloride and catalytic DMF. The isoindoline
intermediate 21 was prepared in 4 total steps from the commercially available ester 19. Exposure
of 19 to benzyl bromide in refluxing acetonitrile with cesium carbonate led to efficient
introduction of the benzyl group to give 20. Saponification with lithium hydroxide and subsequent
acid chloride formation with oxalyl chloride gave the desired acid chloride precursor, 21.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Cell Assays. To identify compounds that block efflux and potentiate the antibacterial activity
of novobiocin the testing scheme outlined in Figure 4 was followed for all compounds.
Compounds were initially evaluated for their ability to prevent bacterial cell growth in
combination with novobiocin (16 ꢀg/mL, 0.25 MIC) in E. coli WTꢀPore cells. In this way,
inhibition of cell growth should be due either to the test compound alone or through working in
combination with novobiocin in some manner. In addition, the presence of the large nonꢀspecific
pore in the outer membrane of WTꢀPore cells eliminates the permeability barrier for the
2
1
compounds and possible nonꢀspecific effects on the outer membrane. Compounds were also
evaluated for the ability to inhibit bacterial growth both in the wildꢀtype E. coli cells (WT) and in
the WTꢀPore cells in the absence of novobiocin to determine whether they had any intrinsic
antibacterial properties. Compounds having an MIC ≤100 ꢀM in combination with novobiocin
were then evaluated in a checkerboard assay to determine the minimum potentiation concentration
2
0
(
MPC) that results in a fourfold improvement (MPC ) in the MIC of novobiocin. Compounds
4
with an MPC Nov ≤ 50 ꢀM were examined for their affinity as substrates for the AcrABꢀTolC
4
efflux pump and also their ability to permeate the OM barrier. To accomplish this, compounds
were tested in two additional cellꢀbased bacterial growth inhibition assays. One assay tested the
ACS Paragon Plus Environment

Page 9 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
compound’s ability to inhibit growth in E. coli cells engineered to lack the AcrABꢀTolC efflux
pump (ꢁTolC). The other looked at antibacterial activity in E. coli cells in which the AcrABꢀTolC
(ꢁTolCꢀPore) pump was knocked out and with large pores in the OM. These assays help to define:
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
(
i) if potentiation is likely due to blocking the AcrABꢀTolC pump or by other mechanisms; (ii)
whether the compounds are efflux substrates of the pump; and (iii) how modifications affect the
compound’s ability to penetrate the OM. Compounds having an MPC Nov ≤ 50 ꢀM were also
4
profiled in the SPR and efflux inhibition experiments as described below. Finally, to test the
general applicability of this approach the most promising compounds were tested for their ability
to potentiate the macrolide antibiotic erythromycin. As an antibiotic, erythromycin works through
a different mechanism of action from novobiocin but it is a substrate for the AcrABꢀTolC pump
30
and is therefore only weakly effective against E. coli. In this way, we can further substantiate the
compounds are working generally as EPIs and are not specific just for novobiocin. Similar to
experiments with novobiocin, test compounds were administered in the presence of a subꢀlethal
dose (5 ꢀg/mL, 0.25 MIC) of erythromycin and compounds that potentiated the antibacterial
activity of erythromycin were run in the full checkerboard assay format to determine the minimum
potentiation concentration.
SPR and HT Assays. In addition to growth inhibition assays, compounds were evaluated for
their ability to bind to the purified AcrA protein and to inhibit or slow down the rate of drug efflux
in the cell. To determine if new compounds could bind to AcrA, SPR sensorgrams were collected
with 50 or 100 ꢀM analyte injected over the purified AcrA protein immobilized onto a CM5
sensorchip, using a Biacore instrument. To evaluate the ability of the compound to inhibit drug
efflux, the fluorescent HT dye was used. When HT intercalates into membranes or intracellular
DNA it gives a fluorescent signal. HT has also been shown to be a substrate for the AcrABꢀTolC
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 60
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
26
efflux pump in E. coli that is rapidly effluxed out of the cell. As a result, fluorescence from HT
accumulation increases very slowly in these cells with active efflux. In the presence of an EPI,
however, HT can accumulate in the cell, as indicated by an increase in fluorescence. Therefore,
compounds that behave as EPIs, such as 1, should accelerate the rate of fluorescence change.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Structure-Activity Relationships. Compound analogs of 1 were prepared to evaluate the roles
of the halo substituent (R ) in the central ring and the dihydroimidazoline moiety in the benzamide
1
groups (R ) within the hit template (Table 1) using the predicted binding pose at the hinge site as a
2
starting point. The docking calculations suggest that the dihydroimidazoline groups could
potentially engage in ionic hydrogen bonding interactions with the backbone carbonyl groups of
Ile 65, Ala 172, or both. To test this hypothesis, we prepared a small set of analogs (4aꢀg) in which
the dihydroimidazolines were replaced with groups that are capable of either accepting or donating
hydrogen bonds. The cyano analogs, 4a and 4b, capable of accepting hydrogen bonds, suffered a
significant loss in potentiation. Similarly, the hydroxymethyl compound, 4c, capable of both
accepting and donating hydrogen bonds, also failed to show any significant potentiation activity.
The same was also true of the negatively charged sulfonic acid derivative, 4d. However, the
aminomethyl analog, 4e, which has estimated pK values of ~ 9.0 and 9.6, showed only a minor
a
reduction in potentiation activity (MPC = 50 ꢀM) compared to 1 (See Supplemental Material,
4
Figures S1 and Table S1) for additional information and docking poses for these compounds). We
also prepared the closely related bioisosteres 4ꢀmethyl,1,2,4ꢀtriazole, 4f, and imidazole, 4g, which
failed to exhibit any significant potentiation of novobiocin. These findings suggest that a basic
nitrogen capable of acting as a hydrogen bond donor is required in this position. When the
dihydroimidazoline was shifted from the para to the meta position, 9a, we observed only a twofold
drop (50 ꢀM) in the MPC value. On the basis of the observed MPC values, replacement of
4
4
ACS Paragon Plus Environment

Page 11 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
chlorine with a nitro, 9e, or methyl group, 9f, had minimal impact on novobiocin potentiation
values (MPC = 25 ꢀM). The bromo, 9d, and desꢀchloro analog, 9b, were also similar to 1,
4
exhibiting only a twofold loss in potentiation. Analog 9c, where R = F, was the only one to show
1
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
a significant change in potentiation values with a fourfold loss being observed. Within the
symmetrical starting scaffold only minor variations in potentiation values were observed for the
majority of analogs containing the dihydroimidazoline group. In contrast, significant losses in
potentiation were observed when that group was replaced, strongly highlighting that the
dihydroimidazoline is critical for activity but the substituent R does not have a dramatic effect on
1
activity. These findings are generally consistent with the model depicting binding in the hinge site
where modifications of the dihydroimidazoline ring would be expected to disrupt favorable
interactions and reduce the affinity for AcrA.
We also prepared a small number of analogs with changes to the central phenyl ring (Table 2).
Preliminary docking experiments suggested that the analog with a saturated transꢀcyclohexyl ring,
11, could bind to AcrA (See Supplemental Material, Figure S3). Experimental results, however,
showed no potentiation of novobiocin, suggesting that steric hindrance or the lack of aromaticity in
the cyclohexyl ring may have a more substantial effect than predicted. Similarly, when the central
ring was replaced with a less rigid acyclic core derived from succinic acid, 13, no potentiation was
observed. The reversed amide analog, 14, which lacks the chlorine substituent in the central ring,
potentiated novobiocin better (MPC = 12.5 ꢀM) than 1. Altering the substitution pattern about the
4
central ring, in this case switching from a 1,4ꢀdisubstituted amide to a 1,3ꢀdisubsituted amide, 16,
also led to significant reduction of potentiation activity relative to 1.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 60
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
In light of the result with the meta-isomer analog 9a having similar activity to 1 we speculated that
perhaps only one of the dihydroimidazoline groups is critical for activity. Additionally, docking
calculations of 1 identified poses in the AcrA site in which only one of the dihydroimidazoline
groups was engaged with the protein (See Supplemental Figure S2). With that in mind, we turned
our focus to preparing nonꢀsymmetrical analogs derived from addition of aniline 3 to various
monoꢀacid chlorides (Scheme 4). We prepared a number of compounds with simple R groups and
several of these are shown in Table 3. Amides derived from simple substituted benzoyl chlorides
such as 17a-17c failed to show any substantial potentiation. The phenylacetamide analog 17d,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
however, did show modest potentiation (MPC = 100 ꢀM), suggesting that was possible to achieve
4
activity with simpler nonꢀsymmetrical analogs. The dihydroisoindoline 17e and the phenethyl
derivative 17f were also inactive with regard to potentiation. However, the more rigid cinnamoyl
derivative, 17g, did show hints of activity but was somewhat obscured by poor solubility.
Replacement of hydrogen with chlorine in the phenyl ring of the cinnamoyl group improved
solubility and both the orthoꢀanalog 17h and paraꢀanalog 17l showed a marked improvement in
potentiation, around 25 ꢀM, while the metaꢀchloro, 17k, was only slightly worse than these two.
Compounds 17h and 17l are the first simplified analogs with potentiation at least as good as 1.
Replacement of the 2ꢀchloro with 2ꢀbromo, 17i, led to a small reduction in potentiation but was
still active. A similar result was observed with the 2ꢀtrifluoromethyl group, 17j. As mentioned
above, the 4ꢀchloro analog was active and modifications at this position led to a greater effect on
activity. The bromo analog, 17m, showed a twofold loss in activity relative to 17l while the
methoxy analog, 17n, suffered a fourfold loss in activity. However, addition of a bulkier alkyl
group such as isopropyl, 17o, or tꢀbutyl, 17p, led to analogs with improved potentiation. With the
activity shown by the cinnamoyl derivatives, we decided to try a simple naphthyl derivative, 17q,
ACS Paragon Plus Environment

Page 13 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
which would be expected to mimic the conformation of the cinnamoyl derivatives. Indeed, 17q
proved to be active with an MPC value of 25 ꢀM.
4
Discussion. We identified a number of compounds that potentiated the activity of novobiocin
in E. coli with values equivalent or better than the original compound, 1. Several of these were
profiled in additional cellꢀbased assays to provide a better understanding of their overall properties
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
(Table 4). Additionally, we assessed their ability to bind to AcrA, inhibit efflux (as measured by
HT uptake) and potentiate erythromycin. From this evaluation, compounds that potentiated
novobiocin fell into three broad categories: (i) those that do not appear to act primarily through an
EPI mechanism, as evidenced by possessing antibacterial activity, weak or no binding to AcrA and
little to no effect on efflux, (ii) those that do appear to act primarily as an EPI, as evidenced by
only weak antibacterial activity, binding to AcrA and a reduced rate of efflux, or (iii) those that
possess some characteristics of both (i) and (ii).
One interesting compound that did not fall into the above categories was the cyclohexyl derivative
11, which did not potentiate novobiocin but did possess antibacterial activity with an MIC = 50
ꢀM in WT cells but decreased only twofold in WTꢀpore cells. This compound showed additional
twofold reductions in MIC when tested in the ꢁTolC and ꢁTolCꢀPore cells, further suggesting it is
a weak substrate of the efflux pump and readily penetrates the OM. In agreement with these
findings, SPR experiments did not show evidence of binding to AcrA, and it only modestly
enhanced uptake of HT. The mechanism of action for this antibacterial activity has not been
determined. The reversed amide compound, 14, is an example that falls into the first category
described above. It did potentiate novobiocin but is also potent as an antibacterial agent in the WT
cells with an MIC = 25 ꢀM. It showed good cell penetration, as its MIC was relatively unaffected
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 60
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
in the ꢁTolC or ꢁTolCꢀWTꢀPore cell lines. Unlike 11, it did bind weakly to AcrA (Figure 5) but
did not improve HT uptake in the cells. These two compounds had interesting antibacterial
properties with good penetration of the OM but do not appear to be acting as EPIs. The metaꢀ
substituted dihydroimidazoline compound, 9a, a close analog to 1, profiled in a similar manner to
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
. It showed only weak antibacterial activity by itself with an MIC = 200 ꢀM in WT cells. The
activity dropped twofold in the WTꢀPore cells and was further reduced to 50 ꢀM in the ꢁTolC
cells, representing a fourfold change in activity when the efflux pump was deleted. These results
suggest that 9a acts an EPI but is also a substrate of the AcrABꢀTolC efflux pump. We were
unable to detect binding to AcrA because of nonꢀspecific binding of the compound to the chip
surface. It did however enhance the rate of HT uptake, but at a rate less than that of 1 (See Figure
6
).
The orthoꢀhalo substituted cinnamoyl derivatives, 17h and 17i, were also profiled. Neither
compound had appreciable antibacterial activity (MICs > 200 ꢀM) in the WT or WTꢀPore cells.
They both showed a twofold decrease in MICs in the ꢁTolC cells, suggesting they have some
affinity as substrates for the efflux pump. Both were found to bind to AcrA (Figure 5) with greater
affinity than 1 and both yielded a significant enhancement of HT uptake relative to 1 (Figure 6).
The analog with the chloro substituent moved to the 4ꢀ position of the cinnamoyl group, 17l,
behaved similarly to that of 17h but was slightly more potent as an antibiotic, with an MIC = 200
ꢀM in WT cells. When tested in the ꢁTolCꢀPore cells, a fourfold reduction in the MIC was
observed, similar to what was observed with 17h and 17i. Replacing the halo substituent in 17l
with bulkier alkyl groups promoted some significant changes in the properties of these compounds.
For instance, the 4ꢀisopropyl analog 17o had an MIC = 100 ꢀM in WT cells while the larger tꢀ
butyl compound, 17p, was even more potent with an MIC fourfold lower at 25 ꢀM. The target
ACS Paragon Plus Environment

Page 15 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
responsible for the greater antibacterial activity of 17p remains unclear at present. The MICs of
these compounds in the ꢁTolC cells were only twofold lower than in the WT cells, suggesting both
17o and 17p are less prone to efflux by the TolC efflux pump (Table 5) relative to the other
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
analogs tested. The isopropyl analog 17o showed one of the strongest binding signals to AcrA we
observed in addition to showing a strong enhancement of HT uptake. The t-butyl analog 17p also
showed binding to AcrA with a 2.6ꢀfold rate enhancement of the HT uptake rate. Finally, we
evaluated the naphthyl compound, 17q, which showed relatively little activity in the WT cells with
only a slight improvement in the WTꢀPore cells. However, in the ꢁTolCꢀPore cells it had an MIC
=
12.5 ꢀM, representing a 16ꢀfold increase in antibacterial activity when the pump was removed.
This compound is clearly a strong substrate for the AcrABꢀTolC efflux pump. It also displayed
strong binding to AcrA but with only a slight enhancement in rate (1.6ꢀfold improvement) for HT
uptake in cells.
Due to the cationic nature of these compounds, the possibility exists that they might disrupt the
proton motive force that drives the AcrABꢀTolC efflux pump. To address this point, we evaluated
1 and the analogs 17h and 17o for their effects on the transmembrane potential in E. coli cells
(Figure 7a). From the data, it appears that our compounds had no effect on the membrane
potential, providing further evidence they work primarily through binding of AcrA and disrupting
the efflux pump assembly. Additionally, because many of these compounds contain a cinnamoyl
moiety, which has the potential to act as a Michael acceptor, we performed a basic cytotoxicity
assay in HEK293 cells with 17h and 17o (Figure 7b). The cytotoxicity of the two compounds
differed significantly, with IC ’s of 10.5 ꢀM and >100 ꢀM, respectively. From this data, it does
5
0
not appear that cytotoxicity contributes to the potentiation activity of 17o but we cannot rule out
that there it doesn’t make a minor contribution to the potentiation seen with 17h. Overall, it does
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 60
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
not appear that the cinnamoyl group poses a major liability with these compounds. The difference
in observed cytotoxicity IC ’s is likely due to the nature of the substitution in the cinnamoyl ring.
50
From a chemical properties standpoint, these new analogs of 1 are characterized primarily by a
reduction in molecular weight (MW) (~120ꢀ150 Da), increased rigidity, and an increased
amphiphilic moment. The latter two properties, along with low globularity and the presence of an
unhindered basic nitrogen, were recently shown to be favorable for antibiotic accumulation in
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
1
Gramꢀnegative bacteria. These compounds fall nicely within the favorable property range for
crossing the OM of Gramꢀnegative bacteria (See Supplemental Material, Table 2S), which is
consistent with our own experimental data. Additionally, with the reduced MW we have the
flexibility to optimize these molecules for binding to AcrA, antibacterial targets, or introduce
additional modifications to the cinnamoyl group to mitigate potential risk while remaining in the
optimal chemical space.
From the data collected a few general SAR trends can be observed. Substitution about the phenyl
ring in the cinnamoyl moiety greatly influences the profile of these compounds. Analogs with
substituents in the 2ꢀposition (i.e. 17h and 17i) tend to have lower antibacterial activity, some
susceptibility to be substrates for efflux, and also inhibit efflux to a greater extent than other
analogs. These compounds also are very effective at crossing the OM as measured by the ratio of
activities in WT vs. WTꢀPore cells (Table 5). Conversely, substitution in the 4ꢀposition (i.e. 17l,
17o and 17p) led to compounds with greater antibacterial activity and less susceptibility to efflux,
but also were less effective at inhibiting efflux. Like the other cinnamoyl derivatives, these
compounds are also effective at crossing the OM. The more rigid naphthyl compound 17q is
interesting because it shows the greatest affinity for AcrA in the SPR experiments, but it is also the
most susceptible to efflux by the AcrABꢀTolC efflux pump. That it does not show greater
ACS Paragon Plus Environment

Page 17 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
potentiation is likely attributed to its greater susceptibility to efflux. It is also one of the few
analogs that are unable to cross the OM readily, with a ratio similar to the original lead, 1. Finally,
all of the compounds are able to potentiate the activity of a second antibiotic, erythromycin (See
Table 5). Erythromycin is a macrolide antibiotic that is ineffective against E. coli because it is a
substrate for AcrABꢀTolC. Erythromycin potentiation values for these compounds are consistent
with those of novobiocin, further suggesting these compounds are working as EPIs. Taken
together, our findings suggest that there are subtle factors that influence whether these compounds
act as EPIs, are substrates for efflux or bind to an antibacterial target. In many cases, the
compounds exhibit more than one of these characteristics. Of the compounds tested, analogs 17h
and 17o appear to be the most promising for future development. Both readily cross the OM, bind
to AcrA, and inhibit efflux in E. coli by potentiating the activity of two separate antibiotics.
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
Conclusion. In this report, we have identified several novel potentiators of novobiocin and
erythromycin in E. coli. These compounds were designed starting from the recently identified
inhibitor 1 with the objective of inhibiting the AcrABꢀTolC efflux pump by interacting with the
MFP constituent AcrA of the AcrABꢀTolC efflux pump. In many cases, the new compounds were
shown to bind AcrA and also to inhibit efflux to some extent in assays measuring the uptake of the
fluorescent probe HT into cells. Two compounds of particular interest are 17h and 17o, both of
which show good potentiation of novobiocin and erythromycin, bind to AcrA, and inhibit efflux at
a rate significantly higher than the original hit. They differ slightly in that the chloroꢀsubstituted
analog 17h has minimal antibacterial activity on its own while 17o has a good MIC in wildꢀtype
cells. Another important aspect of this class of compounds is that they show the potential to cross
the outer membrane of Gramꢀnegative bacteria. In summary, we have identified compounds with
diverse characteristics that inhibit efflux and potentiate the antibacterial properties of novobiocin
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 60
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
and erythromycin in E. coli. Some of these compounds are significantly more potent than our
original lead, 1.
EXPERIMENTAL SECTION
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
General Methods. Compounds 1 and 4d (NSC 55156) were obtained from the National
Cancer Institute. Novobiocin, erythromycin and Hoechst 33342 were obtained from Sigma-Aldrich
Chemie GMBH (Steinham, Germany). All reagents and solvents for the synthesis of analogs 4a-g,
5
, 9a-f and 17a-g were purchased from Sigma-Aldrich Chemie GMBH (Steinham, Germany), Alfa
Aesar GMBH & Co KG (Karlsruhe, Germany), Acros Organics (Geel, Belgium), TCI America
Portland, United States) or Enamine Ltd. (Monmouth, United States) and used as is without
(
further purification. The purities of the final compounds were characterized by highꢀperformance
liquid chromatography (HPLC) using a gradient elution program (Ascentis Express Peptide C18
column, acetonitrile/water 5/95  95/5, 5 min, 0.05% trifluoracetic acid) and UV detection (254
nM). The purities of all compounds final compounds were 95% or greater. 1HꢀNMR and 13Cꢀ
NMR were obtained on a Bruker 400 Mhz instrument and all chemical shifts are referenced to
residual solvent peaks.
PAINS. All of the biologically tested compounds were evaluated for structural attributes
32
consistent with classification as panꢀassay interference compounds (PAINS). The results for each
compound were negative.
1
4
2
-chloro-N ,N -bis(4-(4,5-dihydro-1H-imidazol-2-yl)phenyl)terephthalamide, (1). 2ꢀ
chloroterephthalic acid (50 mg, 0.25 mmol) was converted to the corresponding acid chloride via
treatment with 2.0 M oxalyl chloride solution (0.312 mL, 0.625 mmol) in dichloromethane. A
catalytic quantity of dimethyformamide was added and the reaction was complete in 4 hours on an
ACS Paragon Plus Environment

Page 19 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
ice bath. 2ꢀchloroterephthaloyl dichloride was dried to a yellow solid and dissolved in 1.5 mL
anhydrous toluene. In a second vial, sodium acetate (150 mg, 1.8 mmol) was dissolved in 2.5 mL
glacial acetic acid. To this vial, 4ꢀ(4,5ꢀdihydroꢀ1Hꢀimidazolꢀ2ꢀyl)aniline (88 mg, 0.55 mmol) was
added and the mixture sonicated until a homogeneous solution was achieved. The acetic acid
solution was added dropwise to the acid chloride, purged with argon, and allowed to react
overnight at room temperature. Product precipitated and was collected by filtration. Crude product
was dissolved in a minimal amount of water before being made basic with 4.0 M sodium
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
o
hydroxide solution. Solution was then heated at 50 C until product had completely precipitated as
freeꢀbase. Precipitate was collected and triturated with methanol to yield product as a white solid.
1
(
18 mg, 15% yield). H NMR (400 MHz, DMSOꢀd ) δ 10.84 (s, 1H), 10.61 (s, 1H), 8.17 (s, 1H),
6
1
3
8.05 (d, J = 8.0 Hz, 1H), 7.82 (m, 9H), 3.63 (s, 8H); C NMR (400 MHz, DMSOꢀd ) δ 164.59,
6
163.70, 139.21, 137.00, 130.15, 129.15, 128.78, 128.34, 128.29, 128.23, 128.17, 126.81, 119.79,
+
+
1
19.05, 48.21; HRMS: m/z calcd. for C H ClN O [M+H] : 487.1651; found [M+H] : 487.1691,
26 23
6
2
[
M+2H]: 244.0863.
1
4
2
-chloro-N ,N -bis(4-cyanophenyl)terephthalamide (4a). A solution of 2ꢀchloroterephthalic
o
acid (50 mg, 0.25 mmol) in 2.0 mL of dichloromethane was cooled to 0 C using an ice bath. To
this solution was added sequentially a catalytic amount (~ 1ꢀ2 drops) of DMF followed by a 2.0 M
solution of oxalyl chloride (1.68 mL, 3.3 mmol) in dichloromethane. The resulting mixture was
stirred overnight allowing the temperature to gradually warm to rt. The resulting mixture was
concentrated down to dryness on a rotoevaporator and then dried under vacuum for 2h. The
o
resulting solid was then dissolved in 3.0 mL of THF and cooled to 0 C in an ice bath. To this
mixture was added triethyl amine (0.087 mL, 1.2 mmol) and the reaction was stirred for 5 min. To
this mixture was added 4ꢀaminobenzonitrile (74 mg, 0.62 mmol) as a solution in 1.0 mL of THF.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 60
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
o
The resulting mixture was stirred for 1h at 0 C and then allowed to warm to rt and stir for 2h. The
reaction was quenched with saturated solution of sodium bicarbonate, which led to the formation
of a white precipitate. The precipitate was isolated by filtration and rinsed several times with
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
deionized water. The resulting solid was collected and dried under vacuum to give a white solid
1
(69 mg, 69%). H NMR (400 MHz, DMSOꢀd ) δ 10.96 (s, 2H), 8.18 (d, J = 1.6 Hz, 1H), 8.06 (dd,
6
J = 8.0, 1.6 Hz, 1H), 7.99 (d, J = 8.8 Hz, 2H), 7.91 (d, J = 8.8 Hz, 2H), 7.824 (m, 5H); HRMS: m/z
+
+
calcd. for C H ClN O [M+H] : 401.0807; found [M+H] : 401.0780.
2
2
13
4
2
1
4
2-chloro-N ,N -bis(4-cyanophenyl)terephthalamide (4b). Compound 4b was prepared
according to the procedure outlined above for example 4a using 3ꢀaminobenzonitrile. White solid,
1
yield: 85%; H NMR (400 MHz, DMSOꢀd ) δ 10.72 (s, 2H), 8.28 (m, 2H), 8.13 (s, 4H), 8.07 (m,
6
+
+
2
H), 7.61 (m, 4H); HRMS: m/z calcd. for C H N O [M+H] : 367.1197; found [M+H] :
22 14 4 2
367.1199.
1
4
2-chloro-N ,N -bis(4-(hydroxymethyl)phenyl)terephthalamide (4c). To a 25 mL round bottom
flask with stir bar, 4ꢀaminobenzyl alcohol (0.500 g, 4.05 mmol) and imidazole (0.303 g, 4.46
mmol) were added. Dichloromethane (10.0 mL) was added followed by the addition of tertꢀ
butydimethylsilyl chloride (0.610 g, 4.05 mmol). The reaction was allowed to stir overnight at
room temperature and stopped with addition of ethyl acetate, which induces precipitation of byꢀ
products that can be filtered off before extraction with water. The organic layer was dried with
sodium sulfate and concentrated down to a thick oil before being purified on a 25 g silica flash
column with 40% ethyl acetate in hexane. 4ꢀ(((tertꢀbutyldimethylsilyl)oxy)methyl)aniline obtained
as colorless oil after drying on high vacuum overnight. (0.944 g, 98% yield). 1H NMR (400
MHz, CDCl ) δ 8.98 (s, 1H), 8.83 (s, 1H), 7.71 (t, J = 7.2 Hz, 4H), 7.63 (d, J = 2.0 Hz, 1H), 7.52
3
ACS Paragon Plus Environment

Page 21 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(
dd, J = 8.0, 1.6 Hz, 1H), 7.42 (d, J = 8.0 Hz, 1H), 7.32 (dd, J = 8.4, 4.8 Hz, 4H), 4.73 (d, J = 2.8
Hz, 4H), 0.95 (d, J = 0.8 Hz, 18H), 0.109 (d, J = 1.2 Hz, 12H).
To a reaction vial with magnetic stir bar and Teflon cap, 2ꢀchloroterephthalic acid (100 mg, 0.50
mmol) and TBTU (320 mg, 1.00 mmol) were added and the vial was purged with argon. 5.0 ml
DMF was added through the cap followed by diisopropylethylamine (0.44 mL, 2.50 mmol). The
reaction was allowed to stir for 15 minutes at room temperature before the addition of 4ꢀ(((tertꢀ
butyldimethylsilyl)oxy)methyl)aniline (296 mg, 1.25 mmol). The reaction was run overnight
before being quenched with water and extracted into ethyl acetate. The organic layer was treated
with sodium sulfate and dried to give an oil. The product was run through a 25 g silica flash
column with 2% methanol in dichloromethane and dried to a white solid (256 mg, 80% yield).
Cleavage of the TBDMS group was achieved by dissolving the compound in THF (5.0 ml),
followed by addition of 0.1 M tetrabutylammonium fluoride (0.98 mL) and stirring overnight. The
reaction mixture was dried to a solid and run through a 24 g silica flash column with 2% methanol
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
1
in dichloromethane. The product was dried to a white solid (37 mg, 22% yield). H NMR (400
MHz, DMSOꢀd ) δ 10.58 (s, 1H), 10.40 (s, 1H), 8.14 (d, J = 1.6 Hz, 1H), 8.03 (dd, J = 8.0, 1.6 Hz,
6
1H), 7.74 (m, 3H), 7.67 (d, J = 8.8 Hz, 2H), 7.31 (dd, J = 8.8, 4 Hz, 4H) 5.15 (dt, J = 5.6, 1.6 Hz,
1
3
2
H), 4.48 (dd, J = 5.6, 3.2 Hz, 4H); C NMR (400 MHz, DMSOꢀd ) δ 164.22, 163.26, 139.38,
6
1
38.27, 138.21, 137.33, 137.08, 130.11, 129.01, 128.59, 126.96, 126.81, 126.61, 120.27, 119.41,
+
+
62.59, 62.56; HRMS: m/z calcd. for C H ClN O [M+H] : 411.1113; found [M+H] : 411.1090.
2
2
19
2
4
1
4
N ,N -bis(4-(aminomethyl)phenyl)terephthalamide (4e). To a reaction vial with magnetic stir
bar and teflon cap, terephthaloyl chloride (40 mg, 0.20 mmol) was added along with 2.0 mL
anhydrous dichloromethane. 4ꢀ((NꢀBoc)aminomethyl)aniline (93 mg, 0.42 mmol) was added along
with triethylamine (0.06 mL, 0.42 mmol). The vial was purged with argon and allowed to react
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 60
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
overnight at room temperature. Methanol was added to the reaction mixture and it was filtered to
collect the Bocꢀprotected product as a white solid. The protecting group was removed with the
addition of TFA and stirred for 4 hours before drying to a solid. The crude product was sonicated
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
in 4M NaOH and filtered to collect the product as a white solid (21 mg, 28 % yield); H NMR
(400 MHz, DMSOꢀd ) δ 10.50 (s, 2H), 8.11 (s, 8H), 7.83 (d, J = 8.4 Hz, 4H), 7.45 (d, J = 8.4 Hz,
6
1
3
4
H), 4.02 (s, 4H); C NMR (400 MHz, DMSOꢀd ) δ 164.88, 139.19, 137.32, 129.46, 129.36,
6
+
+
1
27.85, 120.53, 42.00; HRMS: m/z calcd. for C H N O [M+H] : 375.1823; found [M+H] :
22 22 4 2
375.1822, [M+2H]: 171.0680.
1
4
2
-chloro-N ,N -bis(4-(4-methyl-4H-1,2,4-triazol-3-yl)phenyl)terephthalamide (4f).
Compound 4f was prepared from 2ꢀchloroterephthalic acid and 4ꢀ(4ꢀmethylꢀ4Hꢀ1,2,4ꢀtriazolꢀ3ꢀ
1
yl)aniline as described above for the synthesis of compound 4a. White solid, yield: 57%; H NMR
(
400 MHz, DMSOꢀd ) δ 10.89 (s, 1H), 10.66 (s, 1H), 8.56 (s, 2H), 8.20 (d, J = 0.8 Hz, 1H), 8.08
6
(dd, J = 7.6, 1.6 Hz, 1H), 7.99 (d, J = 8.8 Hz, 2H), 7.91 (d, J = 8.8 Hz, 2H), 7.83 (d, J = 7.6 Hz,
+
1
H), 7.78 (m, 4H), 3.76 (d, J = 3.6 Hz, 6H); HRMS: m/z calcd. for C H ClN O [M+H] :
2
6
21
8
2
+
5
13.1556; found [M+H] : 513.1557, [M+2H]: 257.0821.
1
4
N ,N -bis(4-(1H-imidazol-2-yl)phenyl)-2-chloroterephthalamide (4g). Compound 4g was
prepared from 2ꢀchloroterephthalic acid and 4ꢀ(1Hꢀimidazolꢀ2ꢀyl)aniline as described above for
1
the synthesis of compound 4a. White solid, yield: 24%; H NMR (400 MHz, DMSOꢀd ) δ 12.48
6
(s, 2H), 10.75 (s, 1H), 10.54 (s, 1H), 8.18 (d, J = 1.6 Hz, 1H), 8.06 (dd, J = 7.6, 1.6 Hz, 1H), 7.93
13
(
m, 4H), 7.88 (d, J = 9.2 Hz, 2H), 7.80 (m, 3H), 7.13 (s, 4H); C NMR (400 MHz, DMSOꢀd ) δ
6
1
64.38, 163.45, 139.32, 138.64, 138.58, 137.09, 130.18, 129.14, 128.70, 126.72, 126.41, 125.45,
25.30, 120.50, 119.72;
1
ACS Paragon Plus Environment

Page 23 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(
2-chloro-1,4-phenylene)bis((2-(4-aminophenyl)-4,5-dihydro-1H-imidazol-1-yl)methanone)
5). 2ꢀchloroterephthalic acid (102 mg, 0.51 mmol) was added to a reaction vial with a stir bar.
(
TBTU was added (327 mg, 1.02 mmol) before the addition of 10 mL anhydrous DMF.
Diisopropylethylamine (0.45 mL, 2.55 mmol) was added and the reaction stirred for 10 minutes
before the addition of 4ꢀ(4,5ꢀdihydroꢀ1Hꢀimidazolꢀ2ꢀyl)aniline (3) (210 mg, 1.29 mmol). The vial
was purged with argon and allowed to react overnight at room temperature before extracting the
product into dichloromethane from water. The organic layer was dried with sodium sulfate and
dried to a solid on a rotavap. It was then purified on an ISCO 12 g silica column (4% methanol in
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
1
DCM). The product was obtained as a yellow solid. (51 mg, 21% yield). H NMR (400 MHz,
DMSOꢀd ) δ 7.43 (s, 3H), 7.15 (s, 4H), 6.39 (s, 4H), 5.47 (m, 4H), 3.81 (s, 8H).
6
1
4
N ,N -bis(3-(4,5-dihydro-1H-imidazol-2-yl)phenyl)terephthalamide (9a). Compound 9a was
1
4
prepared from bisꢀnitrile intermediate N ,N ꢀbis(3ꢀ(4,5ꢀdihydroꢀ1Hꢀimidazolꢀ2ꢀ
yl)phenyl)terephthalamide, 4b, as described below for the preparation of compound 9b. White
1
solid; yield: 30%; H NMR (400 MHz, DMSOꢀd ) δ 10.85 (s, 2H), 10.60 (s, 4H), 8.54 (m, 2H),
6
13
8
.17 (s, 4H), 7.98 (m, 2H), 7.68 (m, 4H), 4.03 (s, 8H); C NMR (400 MHz, DMSOꢀd ) δ 165.32,
6
165.07, 139.65, 137.01, 129.73, 128.01, 126.58, 124.07, 122.76, 120.36, 44.43; HRMS: m/z calcd.
+
+
for C H N O [M+H] : 453.2041; found [M+H] : 453.2043, [M+2H]: 227.1057.
2
6
24
6
2
1
4
N ,N -bis(4-(4,5-dihydro-1H-imidazol-2-yl)phenyl)terephthalamide dihydrochloride (9b). To
a reaction vial with magnetic stir bar and Teflon cap, terephthaloyl dichloride (70 mg, 0.34 mmol)
and 4ꢀaminobenzonitrile (100 mg, 0.86 mmol) were added and vial was purged with argon. 3.0 ml
THF was added along with trimethylamine (0.12 mL, 0.86 mmol). The reaction was allowed to stir
at room temperature overnight. The reaction was quenched with the addition of saturated sodium
bicarbonate solution and filtered to collect crude product, which was then purified with trituration
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 60
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
4
in methanol to collect N ,N ꢀbis(4ꢀcyanophenyl)terephthalamide intermediate as a white solid.
(
107 mg, 86% yield). 1H NMR (400 MHz, DMSOꢀd ) δ 10.92 (s, 2H), 8.21 (s, 4H), 8.02 (d, J =
6
9
.2 Hz, 4H), 7.85 (d, J = 8.8 Hz, 4H).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
4
N ,N ꢀbis(4ꢀcyanophenyl)terephthalamide (40 mg, 0.11 mmol) was added to a 2 dram vial with stir
bar. Ethylenediamine (0.768 mL, 11.5 mmol) was added along with 1.0 ml dimethylacetamide
before the addition of sodium hydrosulfide hydrate (88 mg, 1.57 mmol). The solution was capped
o
and carefully heated to 115 C. The reaction was set on a timer and allowed to react 12 hours with
constant stirring. Once cooled, the reaction was quenched with water to precipitate product and
filtered to collect product as the free base, which was not soluble in dimethylsulfoxide. It was then
o
converted to the HCl salt by adding 8 eq. HCl as a 4.0 M solution in dioxane and heating at 90 C
for one hour. The product was collected as a white solid (19 mg, 39% yield) and converted to the
1
triflate salt with trifluoroacetic acid (1 eq.) H NMR (400 MHz, DMSOꢀd ) δ 10.95 (s, 2H), 10.47
6
1
3
(
s, 4H), 8.17 (s, 4H), 8.10 (d, J = 9.2 Hz, 4H), 8.01 (d, J = 9.2 Hz, 4H), 4.00 (s, 8H); C NMR
(
400 MHz, DMSOꢀd ) δ 165.47, 164.42, 144.55, 137.24, 129.53, 128.07, 120.05, 116.80, 44.32;
6
+
+
HRMS: m/z calcd. for C H N O [M+H] : 453.2041; found [M+H] : 453.2038, [M+2H]:
2
6
24
6
2
227.1061.
1
4
N ,N -bis(4-(4,5-dihydro-1H-imidazol-2-yl)phenyl)-2-fluoroterephthalamide (9c). Compound
9
c was prepared from 2ꢀfluoroterephthalic acid and 4ꢀ(4,5ꢀdihydroꢀ1Hꢀimidazolꢀ2ꢀyl)aniline as
1
outlined above for compound 1. White solid, converted to TFA Salt, yield: 18%; H NMR (400
MHz, DMSOꢀd ) δ 11.15 (s, 1H), 10.99 (s, 1H), 10.52 (s, 1H), 8.11 (d, J = 9.2 Hz, 2H), 8.02 (m,
6
+
8
H), 7.91 (t, J = 8.0 Hz, 1H), 4.01 (s, 8H); HRMS: m/z calcd. for C H FN O [M+H] : 471.1947;
26 23
6
2
+
found [M+H] : 471.1945, [M+2H]: 236.1010.
ACS Paragon Plus Environment

Page 25 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
4
N ,N -bis(4-(4,5-dihydro-1H-imidazol-2-yl)phenyl)-2-bromoterephthalamide (9d). Compound
9
d was prepared from 2ꢀbromoterephthalic acid and 4ꢀ(4,5ꢀdihydroꢀ1Hꢀimidazolꢀ2ꢀyl)aniline as
1
outlined above for compound 1. Tan solid, yield: 17%; H NMR (400 MHz, DMSOꢀd ) δ 10.79 (s,
6
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
1
H), 10.58 (s, 1H), 8.31 (d, J = 1.6 Hz, 1H) 8.08 (dd, J = 4.8, 2.0 Hz, 1H), 7.84 (d, J = 4.4 Hz,
1
3
4H), 7.81 (s, 2H), 7.76 (m, 3H), 3.61 (s, 8H); C NMR (400 MHz, DMSOꢀd ) δ 165.43, 163.49,
6
1
63.22, 163.17, 141.45, 140.69, 140.58, 136.98, 131.73, 128.94, 127.94, 127.77, 127.23, 125.68,
+
1
25.18, 119.71, 119.05, 118.95, 49.36; HRMS: m/z calcd. for C H BrN O [M+H] : 531.1146;
2
6
23
6
2
+
found [M+H] : 531.1120.
1
4
N ,N -bis(4-(4,5-dihydro-1H-imidazol-2-yl)phenyl)-2-nitroterephthalamide (9e). Compound
9
e was prepared from 2ꢀnitroterephthalic acid and 4ꢀ(4,5ꢀdihydroꢀ1Hꢀimidazolꢀ2ꢀyl)aniline as
1
outlined above for compound 1. White solid, yield: 9%; H NMR (400 MHz, DMSOꢀd ) δ 10.95
6
(s, 1H), 10.77 (s, 1H), 8.73 (d, J = 1.6 Hz, 1H), 8.45 (dd, J = 4.8, 1.6 Hz, 1H), 8.00 (d, J = 8.0 Hz,
1
3
1
1
1
H), 7.84 (m, 6H), 7.72 (d, J = 8.8 Hz, 2H), 3.61 (s, 8H); C NMR (400 MHz, DMSOꢀd ) δ
6
63.72, 163.22, 163.18, 162.88, 146.28, 140.47, 140.46, 136.81, 134.80, 133.26, 129.81,
28.01,127.83, 126.04, 125.92, 123.58, 119.83, 119.01, 49.52; HRMS: m/z calcd. for C H N O
4
2
6
23
7
+
+
[M+H] : 498.1892; found [M+H] : 498.1896, [M+2H]: 249.5984.
1
4
N ,N -bis(4-(4,5-dihydro-1H-imidazol-2-yl)phenyl)-2-methylterephthalamide (9f). Compound
9
f was prepared from 2ꢀmethyterephthalic acid and 4ꢀ(4,5ꢀdihydroꢀ1Hꢀimidazolꢀ2ꢀyl)aniline as
1
outlined above for compound 9b. White solid, yield: 44%; H NMR (400 MHz, DMSOꢀd ) δ
6
1
8
1
1
1.00 (s, 1H), 10.88 (s, 1H), 8.10 (d, J = 9.2, 2H), 8.02 (m, 7H), 7.96 (d, J = 8.4, 1H), 7.69 (d, J =
1
3
.0 Hz, 1H), 3.99 (s, 8H), 2.49 (s, 3H); C NMR (400 MHz, DMSOꢀd ) δ 167.37, 165.05, 140.92,
6
40.88, 139.73, 135.61, 135.57, 129.79, 127.82, 127.71, 127.39, 125.14, 119.59, 118.98, 49.35,
+
+
9.36; HRMS: m/z calcd. for C H N O [M+H] : 467.2197; found [M+H] : 467.2183.
2
7
26
6
2
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 60
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
4
(
1r,4r)-N ,N -bis(4-(4,5-dihydro-1H-imidazol-2-yl)phenyl)cyclohexane-1,4-dicarboxamide
1
4
(
11). Compound 11 was prepared from the bisꢀnitrile, (1r,4r)ꢀN ,N ꢀbis(4ꢀ
cyanophenyl)cyclohexaneꢀ1,4ꢀdicarboxamide, following the procedure described above for
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
compound 9b. White solid, yield: 23%; H NMR (400 MHz, DMSOꢀd ) δ 10.52 (s, 2H), 7.93 (d, J
6
= 9.6 Hz, 4H), 7.87 (d, J = 8.8 Hz, 4H), 3.98 (s, 8H), 3.04 (s, 2H), 1.96 (d, J = 6.8 Hz, 4H), 1.49 (t,
1
3
J = 10.0 Hz, 4H); C NMR (400 MHz, DMSOꢀd ) δ 174.93, 164.38, 144.96, 129.66, 118.73,
6
+
1
15.88, 44.29, 44.14, 28.07; HRMS: m/z calcd. for C H N O [M+H] : 459.2510; found
2
6
30
6
2
+
[M+H] : 459.2512, [M+2H]: 230.1294.
1
4
N ,N -bis(4-(4,5-dihydro-1H-imidazol-2-yl)phenyl)succinamide (13). Succinic acid (35 mg, 0.3
mmol) was converted to the corresponding acid chloride via treatment with 2.0 M oxalyl chloride
solution in dichloromethane. A catalytic quantity of DMF was added and the reaction was
complete in 4 hours on an ice bath. The solvent was removed, leaving succinyl chloride as a
yellow solid. 4ꢀ(4,5ꢀdihydroꢀ1Hꢀimidazolꢀ2ꢀyl)aniline (97 mg, 0.6 mmol) was dissolved in 1.5 mL
glacial acetic acid before being added to the acid chloride. The solution was sonicated to mix
before being allowed to react overnight at room temperature. The solvent was removed and crude
product was purified on a 50 g C18 reversedꢀphase column (acetonitrile/water). White solid, 2
1
TFA salt, yield: 9%; H NMR (400 MHz, DMSOꢀd ) δ 10.57 (s, 2H), 10.34 (s, 4H), 7.89 (d, J =
6
1
3
9
.2 Hz, 4H), 7.82 (d, J = 8.8 Hz, 4H), 3.98 (s, 8H), 2.75 (s, 4H); C NMR (400 MHz, DMSOꢀd )
6
δ 171.37, 164.37, 144.76, 129.69, 118.56, 115.83, 44.26, 30.95; HRMS: m/z calcd. for C H N O
2
2
2
24
6
+
+
[
M+H] : 405.2041; found [M+H] : 405.2041, [M+2H]: 203.1056.
1
3
N ,N -bis(4-(4,5-dihydro-1H-imidazol-2-yl)phenyl)isophthalamide (16). Compound 16 was
1
3
prepared from the bisꢀnitrile, N ,N ꢀbis(4ꢀcyanophenyl)isophthalamide, following the procedure
1
described above for compound 9a. White solid, 2 HCl salt, yield: 45%; H NMR (400 MHz,
ACS Paragon Plus Environment

Page 27 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
DMSOꢀd ) δ 11.20 (s, 2H). 10.44 (s, 4H), 8.84 (s, 1H), 8.21 (m, 6H), 8.01 (d, J = 8.8 Hz, 4H), 7.76
6
1
3
(
t, J = 7.6 Hz, 1H), 4.00 (s, 8H); C NMR (400 MHz, DMSOꢀd ) δ 165.48, 164.29, 144.79,
6
+
1
34.14, 131.66, 129.49, 119.81, 116.59, 44.22; HRMS: m/z calcd. for C H N O [M+H] :
2
6
24
6
2
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
+
4
53.2041; found [M+H] : 453.2037, [M+2H]: 227.1056.
4-chloro-N-(4-(4,5-dihydro-1H-imidazol-2-yl)phenyl)benzamide (17a). 4ꢀchlorobenzoic
acid (47 mg, 0.3 mmol) was converted to the corresponding acid chloride via treatment with 2.0
M oxalyl chloride solution (0.18 mL, 0.36 mmol) in dichloromethane. A catalytic quantity
of DMF was added and the reaction was complete in 4 hours on an ice bath. The solvent was
removed, leaving 4ꢀchlorobenzoyl chloride as a yellow solid. 4ꢀ(4,5ꢀdihydroꢀ1Hꢀimidazolꢀ2ꢀ
yl)aniline (48 mg, 0.3 mmol) was dissolved in 1.5 mL glacial acetic acid before being added to the
acid chloride. The solution was sonicated to mix before being allowed to react overnight at room
temperature. The solvent was removed and crude product was purified on a 50 g C18 reversedꢀ
1
phase column (acetonitrile/water). The product was obtained as a TFA salt, (45 mg, 36% yield). H
NMR (400 MHz, DMSOꢀd ) δ 10.76 (s, 1H), 10.36 (s, 2H), 8.04 (d, J = 8.8 Hz, 2H), 8.00 (d, J =
6
13
8
.4 Hz, 2H), 7.94 (d, J = 9.2 Hz, 2H), 7.66 (d, J = 8.8 Hz, 2H), 4.00 (s, 4H); C NMR (400 MHz,
DMSOꢀd ) δ 165.11, 164.39, 144.57, 136.95, 129.81, 129.47, 128.59, 119.92, 116.63, 44.28;
6
+
+
HRMS: m/z calcd. for C H ClN O [M+H] : 300.0905; found [M+H] : 300.0904.
1
6
14
3
Methyl 4-((4-(4,5-dihydro-1H-imidazol-2-yl)phenyl)carbamoyl)benzoate (17b). Compound
17b was prepared from 4ꢀ(methoxycarbonyl)benzoic acid and 4ꢀ(4,5ꢀdihydroꢀ1Hꢀimidazolꢀ2ꢀ
yl)aniline (3) following the procedure described above for compound 17a. Tan solid, yield: 32%;
1
H NMR (400 MHz, DMSOꢀd ) δ 10.63 (s, 1H), 8.10 (dd, J = 12.8, 8.8 Hz, 4H), 7.85 (dd, J =
6
1
3
1
6.5, 9.2 Hz, 4H), 3.90 (s, 3H), 3.65 (s, 4H); C NMR (400 MHz, DMSOꢀd ) δ 165.65, 165.02,
6
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 60
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
63.47, 141.62, 138.75, 132.23, 129.23, 128.19, 128.15, 119.70, 52.47; HRMS: m/z calcd. for
+
+
C H N O [M+H] : 324.1350; found [M+H] : 324.1352.
18 17
3
3
1
4
N -benzyl-N -(4-(4,5-dihydro-1H-imidazol-2-yl)phenyl)terephthalamide (17c). Compound
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
17c was prepared from acid chloride 18, described below, and 4ꢀ(4,5ꢀdihydroꢀ1Hꢀimidazolꢀ2ꢀ
yl)aniline (3) following the procedure described above for compound 17a. White solid, yield 25%;
1
H NMR (400 MHz, DMSOꢀd ) δ 10.53 (s, 1H), 9.24 (s, 1H), 8.05 (s, 4H), 7.84 (s, 4H), 7.31 (m,
6
1
3
5H), 4.52 (s, 2H), 3.60 (s, 4H); C NMR (400 MHz, DMSOꢀd ) δ 165.48, 165.04, 163.15, 140.66,
6
139.50, 139.48, 137.05, 136.95, 128.33, 127.78, 127.61, 127.32, 127.27, 126.82, 126.06, 119.59,
+
+
42.72; HRMS: m/z calcd. for C H N O [M+H] : 399.1823; found [M+H] : 399.1828.
24 22 4 2
2
-(4-chlorophenyl)-N-(4-(4,5-dihydro-1H-imidazol-2-yl)phenyl)acetamide (17d). Compound
7d was prepared from 2ꢀ(4ꢀchlorophenyl)acetic acid and 4ꢀ(4,5ꢀdihydroꢀ1Hꢀimidazolꢀ2ꢀyl)aniline
1
(3) following the procedure described above for compound 17a. Tan solid, TFA salt, yield: 11%;
1
H NMR (400 MHz, DMSOꢀd ) δ 10.70 (s, 1H), 10.32 (s, 2H), 7.88 (d, J = 9.2 Hz, 2H) 7.83 (d, J
6
1
3
=
8.8 Hz, 2H), 7.40 (d, J = 8.4 Hz, 2H), 7.35 (d, J = 8.8 Hz, 2H), 3.98 (s, 4H), 3.73 (s, 2H); C
NMR (400 MHz, DMSOꢀd ) δ 169.77, 164.34, 144.61, 134.41, 131.45, 131.17, 129.68, 128.28,
6
+
118.81, 116.18, 44.27, 42.41; HRMS: m/z calcd. for C H ClN O [M+H] : 314.1062; found
1
7
16
3
+
[
M+H] : 314.1059.
2-benzyl-N-(4-(4,5-dihydro-1H-imidazol-2-yl)phenyl)-1-oxoisoindoline-5-carboxamide (17e).
Ester 20 (75 mg, 0.48 mmol) was dissolved in 5 mL of a 9ꢀ1 mixture of THFꢀH O and stirred
2
vigorously. To this mixture was added LiOH (25 mg, 1 mmol) and the reaction was stirred at rt
overnight. The reaction was judged complete and the solvent was removed to near dryness by
passing a gentle stream of air over the vial. The resulting residue was neutralized to ~ pH 5ꢀ6 with
ACS Paragon Plus Environment

Page 29 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
N HCl and the product was extracted with dichloromethane (4X). The organics, were combined,
dried and concentrated in vacuo. The resulting acid was converted to the acid chloride without
further purification.
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 acid (55 mg, 0.23 mmol) was added to a vial fitted with a Teflon screwcap and slurried in 1.5
o
mL of dichloromethane. The reaction was then cooled to 0 C and a catalytic amount of DMF
(0.01 mL) was added followed by addition of 2.0 M solution of oxalyl chloride (0.125 mL, 0.5
o
mmol) in dichloromethane. The resulting solution was stirred at 0 C for 10 min then allowed to
warm to rt and stir for 1h to give a dark yellow homogeneous solution. The solvent was then
removed under vacuum to give a yellowꢀbrown residue and further dried under vacuum for 2h. In
a separate vial, 4ꢀ(4,5ꢀdihydroꢀ1Hꢀimidazolꢀ2ꢀyl)aniline (3) (37 mg, 0.23 mmol) was dissolved in
1
mL of glacial acetic acid. The resulting homogeneous mixture was stirred for 5 min at rt and then
added all at once to the acid chloride residue from above. A thick yellow precipitate forms and the
reaction was allowed to stir overnight. The resulting reaction mixture was filtered through a fritted
funnel with washing from diethyl ether and then dried. The solid is then slurried in water and a
saturated sodium bicarbonate solution was added until the pH ~ 10. The solution was then filtered
and washed sequentially with water and diethyl ether and further dried under vacuum to give a tan
1
solid (15 mg, 16%). This was converted to the TFA salt as described above, H NMR (400 MHz,
DMSOꢀd ) δ 10.90 (s, 1H), 10.41 (s, 2H), 8.13 (s, 1H), 8.09 (d, J = 7.6 Hz, 1H), 8.05 (d, J = 9.2
6
Hz, 2H), 7.95 (d, J = 8.8 Hz, 2H), 7.91 (d, J = 8.0 Hz, 1H), 7.37 (m, 2H), 7.30 (m, 3H), 4.78 (s,
13
2
1
4
H), 4.48 (s, 2H), 4.00 (s, 4H); C NMR (400 MHz, DMSOꢀd ) δ 166.53, 165.88, 164.42, 141.92,
6
37.28, 137.13, 129.53, 128.75, 127.75, 127.68, 127.45, 123.32, 123.05, 119.95, 116.72, 49.40,
+
+
5.53, 44.30; HRMS: m/z calcd. for C H N O [M+H] : 411.1823; found [M+H] : 411.1801.
2
5
22
4
2
ACS Paragon Plus Environment