Antibacterial drug leads: DNA and enzyme multitargeting

Article abstract of DOI:10.1021/jm501449u

We report the results of an investigation of the activity of a series of amidine and bisamidine compounds against Staphylococcus aureus and Escherichia coli. The most active compounds bound to an AT-rich DNA dodecamer (CGCGAATTCGCG)₂ and using DSC were found to increase the melting transition by up to 24 °C. Several compounds also inhibited undecaprenyl diphosphate synthase (UPPS) with IC₅₀ values of 100-500 nM, and we found good correlations (R² = 0.89, S. aureus; R² = 0.79, E. coli) between experimental and predicted cell growth inhibition by using DNA δT_m and UPPS IC₅₀ experimental results together with one computed descriptor. We also solved the structures of three bisamidines binding to DNA as well as three UPPS structures. Overall, the results are of general interest in the context of the development of resistance-resistant antibiotics that involve multitargeting.

Full text of DOI:10.1021/jm501449u

Subscriber access provided by SELCUK UNIV
Article
Antibacterial Drug Leads: DNA and Enzyme Multi-Targeting
Wei Zhu, Yang Wang, Kai Li, Jian Gao, Chun-Hsiang Huang, Chun-
Chi Chen, Tzu-Ping Ko, Yonghui Zhang, Rey-Ting Guo, and Eric Oldfield
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/jm501449u • Publication Date (Web): 09 Jan 2015
Downloaded from http://pubs.acs.org on January 10, 2015
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 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Antibacterial Drug Leads: DNA and Enzyme Multiꢀ
Targeting
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
Wei Zhu1, Yang Wang1, Kai Li1, Jian Gao2, Chun-Hsiang Huang2, Chun-Chi Chen2, Tzu-Ping
Ko3, Yonghui Zhang1, Rey-Ting Guo2 and Eric Oldfield1,*
1 Department of Chemistry, University of Illinois at UrbanaꢀChampaign, 600 South Mathews
Avenue, Urbana, Illinois 61801, United States
2 Industrial Enzymes National Engineering Laboratory, Tianjin Institute of Industrial
Biotechnology, Chinese Academy of Sciences, Tianjin 300308, China
3 Institute of Biological Chemistry, Academia Sinica, Taipei 11529, Taiwan
KEYWORDS
Drug discovery | Protein structure | DNA minor groove binder | Undecaprenyl diphosphate
synthase
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 47
1
2
3
4
Abstract
5
6
7
8
9
We report the results of an investigation of the activity of a series of amidine and bisamidine
compounds against Staphylococcus aureus and Escherichia coli. The most active compounds
bound to an ATꢀrich DNA dodecamer (CGCGAATTCGCG)2, and using DSC were found to
increase the melting transition by up to 24 ˚C. Several compounds also inhibited undecaprenyl
diphosphate synthase (UPPS) with IC50 values of 100ꢀ500 nM and we found good correlations
(R2 = 0.89, S. aureus; R2=0.78, E. coli)) between experimental and predicted cell growth
inhibition by using DNA ∆Tm and UPPS IC50 experimental results together with 1 computed
descriptor. We also solved the structures of three bisamidines binding to DNA as well as three
UPPS structures. Overall, the results are of general interest in the context of the development of
resistanceꢀresistant antibiotics that involve multiꢀtargeting.
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
ACS Paragon Plus Environment
2

Page 3 of 47
Journal of Medicinal Chemistry
1
2
3
4
Introduction
5
6
7
8
There is currently a dearth of new antibiotics being introduced with both the United
States Centers for Disease Control and Prevention1 as well as the World Health Organization,2
and others,3 warning of the seriousness of the development of widespread antibiotic resistance.
The problem is a multiꢀfaceted one with many factors such as use of antibiotics in feedstocks,
unnecessary prescriptions, efflux pumps, and drug modifications all contributing to the problem.
In addition, of course, random mutations in a target can occur, rendering a drug ineffective. One
approach to ameliorate the latter problem is the use of combination therapies and these are the
norm for malaria, HIV/AIDS, tuberculosis, as well as cancer therapeutics. While the
development of just one new drug is a daunting prospect, the development of two new drugs
(with new targets/mechanisms of action) will be even more challenging, hence the interest is new
leads that act by multipleꢀtargeting: one drug (drug lead) but two (or more) targets—an approach
that can improve efficacy as well as decrease the likelihood of resistance developing. In our lab,
we recently used computational screening to try and find novel inhibitors of two antiꢀbacterial
drug targets, farnesyl diphosphate synthase (FPPS)4,5 and undecaprenyl diphosphate synthase
(UPPS),6 that are involved in bacterial cell wall biosynthesis (and are not the targets of any
currently FDAꢀapproved antiꢀinfective drugs). Inhibiting either of these targets would be of
interest since it could lead to “resistance reversal” in e.g. meticillin or vancomycin resistant
bacteria (in a combination therapy), with UPPS inhibition being of particular interest since
humans do not possess a UPPS gene. We found from our in silico and in vitro work5,6 that the
bisamidine 1 (BPHꢀ1358) was an inhibitor of both FPPS (IC50 ~2 ꢁM) as well as UPPS (IC50
~100 nM) and was active against S. aureus in vitro (MIC ~250 ng/mL) and in vivo (20/20 mice
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
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 47
1
2
3
4
survived in an i.p. infection model with a MRSA strain) and in addition, 1 (Scheme 1) acted
5
6
synergistically with meticillin (with a fractional inhibitory concentration index, FICI = 0.25).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
Scheme 1. Structures of bisamidines with antiꢀinfective activity.
1 is, as noted above, a bisamidine (or a diamidine) and somewhat related compounds
(derived from synthalin, 2) have been developed as antiꢀinfectives since ~1937.7 For example,
pentamidine (3) has been used against trypanosomatid infections and is still used against
Pneumocystis jirovecii pneumonia in HIV/AIDS patients. It is, however, a rather toxic
compoundꢀdespite being on the World Health Organization's List of Essential Medicines. In later
work (beginning in 1960), the A. Wander Company8 developed large numbers of again generally
related compounds, bisamidines such as 4, primarily as antiꢀleukemia drug leads, but some were
also found to have activity against Trypanosoma brucei, T. congolense, the apicomplexan
Babesia rodaini, as well as the bacterium Mycobacterium tuberculosis.8 Other bisamidines were
later developed and found to have promising activity against several other bacteria9 and 5 and
10
related compounds (see
for a historical perspective) have been in clinical trials for human
African trypanosomiasis. In many cases, the mechanism of action (or target) is not completely
clear and, interestingly, it has been found with for example 6 and 7, that DNA synthesis as well
as cell wall biosynthesis are targeted (based on macromolecular labeling studies).11ꢀ13 In other
cases, mitochondrial function is disrupted.14 One possible inference from these results is that
bisamidines may often have more than one target, with soꢀcalled “multiꢀtarget” inhibitors being
ACS Paragon Plus Environment
4

Page 5 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
of interest because they could be less prone to the development of spontaneous resistance due to
target mutations, with multiꢀtarget inhibition15 being thought to be one of the main reasons that
some antibiotics have been relatively successful in monoꢀtherapy.16
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 bisamidines 6, 7 and several related compounds have also been shown to bind to
bacterial DNA (as observed by fluorescence – the DNAꢀbound ligands are highly fluorescent)
and, more specifically, ATꢀrich DNA11,12,17 in which hydrogenꢀbonding between the ligand and
DNA bases can occur.18 What is interesting here is that, in addition to these bisamidines binding
to bacterial DNA, DNAꢀbinding is also the proposed mechanism of action of several of the
bisamidines that have been developed as antiꢀtrypanosomatid drug leads (that reached clinical
trials) where it is kinetoplast DNA (kꢀDNA) that is targeted.17 DNA minor groove binders have
also been developed to target a broad range of fungal and bacterial pathogens including
Clostridium difficile,19 so the possibility exists that it might be possible to develop compounds
that target both DNA as well as a pathogen enzyme target (e.g. UPPS or FPPS), improving
selectivity.
In this work, we synthesized a range of amidines and bisamidines and tested them, as
well as other known bisamidines, for their binding to (or inhibition of) E. coli UPPS, S. aureus
UPPS, and an ATꢀrich DNAꢀduplex (CGCGAATTCGCG)2 and correlated these results with
their effects on S. aureus and E. coli cell growth. In some cases we found both DNA minor
groove binding as well as UPPS inhibition, leading to predictive models of cell growth
inhibition. We also solved three Xꢀray structures of some of the leads bound to the DNA
dodecamer duplex, in addition to determining three UPPS Xꢀray structures.
Results and Discussion
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 47
1
2
3
4
5
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
Figure 1. Structures of compounds of interest, discussed in the Text. Compounds are rank
ordered (top left to bottom right) by their effects on the foldedꢀunfolded transition (∆Tm) of a
DNA dodecamer duplex (CGCGAATTCGCG)2, as determined by differential scanning
calorimetry, the most potent binder being at the top left of the Figure.
We investigated the compounds shown in Figure 1 for their effects on enzyme (E. coli
UPPS, S. aureus UPPS) inhibition, E. coli and S. aureus cell growth inhibition, and on the
foldedꢀunfolded transition of the ATꢀrich DNA dodecamer duplex, (CGCGAATTCGCG)2.
Compounds 6ꢀ8 have known antiꢀbacterial activity and were discovered or developed by
Microbiotix (Worcester, MA) from a DTP/NCI (Developmental Therapeutics Program/National
Cancer Institute) screening library (6 = MBXꢀ1162; 7 = MBXꢀ1066 = NSCꢀ317881; 8 = MBXꢀ
ACS Paragon Plus Environment
6

Page 7 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1090 = NSCꢀ317880); 9 is the antiꢀbacterial netropsin; 1 is the bisamidine (NSC 50460) reported
previously5,6 to be a potent UPPS, FPPS inhibitor active against S. aureus in vitro and in vivo.
Compounds 10ꢀ13, 16-17 are all analogs of 1 or 10 containing various sideꢀchain or backbone
modifications (see Experimental Section for synthesis and characterization); 14 is a Wander A.G.
compound (NSCꢀ60340); 3 is pentamidine; 15 is an amidine based on the synthetic antimicrobial
1820 (which contains the “cationicꢀhydrophobicꢀcationic” motif found in the bisamidines).
Compound 19 is a very potent UPPS inhibitor (IC50 ~400 nM), but is not expected to bind to
DNA (since it is highly anionic).
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
Enzyme Inhibition. We first tested all 16 compounds against UPPS from E. coli and S.
aureus. Results are shown in Table 1. The compounds are rankꢀordered by their ∆Tm values, the
strongest DNAꢀbinder being at the top of the Table. As can be seen in Table 1, 1 in the most
potent inhibitor of both E. coli UPPS (EcUPPS) as well as S. aureus UPPS (SaUPPS) with an
IC50 = 110 nM. The tetraphosphonate 19 is also a potent UPPS inhibitor with an IC50 ~400 nM
against both enzymes. The bisindole 6 likewise has potent activity against both enzymes and 7
(the same structure as 6 except for the replacement of the 6ꢀmembered bisamidine ring by a 5ꢀ
membered ring) has good activity against EcUPPS (IC50 = 360 nM) but less (IC50 = 1.7 ꢀM)
against SaUPPS. Several other compounds (13, 18) have low or subꢀmicromolar activity against
SaUPPS, but are less active against EcUPPS. Clearly, the overall most active compounds are 1,
6, 7 and 19. Compounds 1, 6, 7 are bisamidines while 19 is a tetraphosphonate. With the analogs
of 1, replacement of the 5ꢀmembered ring by a 6 membered ring reduced UPPS inhibition
activity by 50 fold (Table 1) and replacement of the amide by a thioamide (1 → 11) reduced
activity by a similar amount. Other sideꢀchain modifications all greatly reduced activity.
ACS Paragon Plus Environment
7

Journal of Medicinal Chemistry
Page 8 of 47
1
2
3
4
Table 1. Enzyme inhibition, cell growth inhibition and differential scanning calorimetry results.
5
6
7
8
9
EcUPPS SaUPPS E. coli
S. aureus
EC50
(ꢁM)
0.028
0.11
0.029
11
IC50
(ꢁM)
0.68
12
IC50
(ꢁM)
0.47
5.1
EC50
(ꢁM)
0.24
1.2
ꢂTm
(°C)
24
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
cpd
6
20
8
0.36
1400
0.11
5.5
1.7
0.33
570
16
7
1000
0.11
4.5
16
9
0.30
0.66
58
0.29
0.23
3.0
11
1
10
10
11
12
13
14
3
9.2
3.5
7.1
5.8
5.2
3.9
2.7
1.2
0.7
0.6
0.1
ꢀ1
380
9.4
530
2.1
>1500
570
570
7.7
4.8
4.9
14
28
400
14
240
5.6
12
4.2
900
380
15
16
17
18
19
25
8.9
>1500
>1500
0.97
>1500
280
11
5.7
10
6.1
1.4
7.6
0.36
0.39
100
ACS Paragon Plus Environment
8

Page 9 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Cell growth inhibition results. With these results on UPPS inhibition in hand, we next
investigated the activity of all 16 compounds against two bacteria, the Gramꢀnegative E. coli,
and the Gramꢀpositive, S. aureus. Cell growth inhibition results are shown in Table 1. As can be
seen in Table 1, there are 4 compounds with <1 ꢁM activity against both organisms 1, 6, 7, and
10, with similar results for 6-8 having been reported previously.9,11,12,17 In most cases, these four
compounds have good activity against UPPS, the exception being 10 (where UPPS IC50 values
are ~5 ꢁM). However, the potent UPPS inhibitor 19 had no cell activity, presumably because it
has poor permeability and in previous work21 we found only one related (bisphosphonate)
inhibitor having activity (EC50 = 33 ꢁM), against E. coli. These results appear to be in accord
with the idea that UPPS could be a target for the most potent cell growth inhibitors (some of
which have been shown to inhibit cell wall biosynthesis). There is, however, only a very poor
overall correlation between enzyme pIC50 (= ꢀlog10 IC50) and cell growth inhibition pEC50 values
with R2 = 0.25, pꢀvalue = 0.05 for E. coli/EcUPPS, and R2 = 0.16, pꢀvalue = 0.12 for the S.
aureus/SaUPPS results (Supporting Information Table S1). It thus seemed that there might be
another, more important target, DNA, since as noted in the Introduction, in other work it has
been shown that bisamidines bind to ATꢀrich DNA, both in bacteria11 as well as to kinetoplast
DNA, in trypanosomatid parasites such as T. brucei.17
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
DNA binding: calorimetry and crystallography. We thus next investigated the binding
of the 16 compounds shown in Table 1 to a DNA dodecamer duplex containing an ATꢀrich
central domain: (CGCGAATTCGCG)2, used previously to study the binding of Hoechst
33258,22 propamidine23 (pentamidine but with a C3 linker) and other ammonium compounds.24
We used differential scanning calorimetry (DSC) to study the DNAꢀligand interaction, and DSC
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 47
1
2
3
4
5
6
thermograms for all 16 compounds are shown in Figure 2 (the results for the compounds having
the largest ∆Tm values being shown at top left).
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
Figure 2. DSC thermograms for the 16 compounds shown in Figure 1 binding to the DNA
dodecamer duplex (CGCGAATTCGCG)2. The ∆Tm values are indicated. ApoꢀDNA
thermograms of are shown in red and DNA/ligand (1:1) thermograms are shown in blue.
ACS Paragon Plus Environment
10

Page 11 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The changes in the DNA thermal unfolding temperatures, ∆Tm (where Tm is the
maximum in the Cp vs. T thermogram) are given in Table 1 and the DSC thermogram results are
shown in Figure 2, ranked according to their ∆Tm values. In most cases there was a single
component observed, although in several instances there were two components (folded +
unfolded), due presumably to solubility/incomplete mixing deficiencies. The two strongest
binding ligands were 6 (∆Tm = 24 ˚C) and 8 (∆Tm = 20 ˚C), followed by 7 (∆Tm = 16 ˚C) and 9
(netropsin, ∆Tm = 16˚C). 1, the 5 membered biphenyl bisamidine had ∆Tm = 11 ˚C, essentially
the same as the 6ꢀmembered ring analog 10 (∆Tm = 10 ˚C). The thioamide analog of 1, 11, had
worse binding (∆Tm = 7.1 ˚C). Pentamidine (3) was also a poor binder (∆Tm = 2.7 ˚C). Clearly,
for most potent binding, the presence of a bisamidine is required, together with, perhaps, central
groups that might Hꢀbond to the DNA bases. What is particularly interesting about these DSC
results is that the most potent DNA binders are also some of the most the potent UPPS inhibitors,
as well as the most potent cell growth inhibitors, suggesting, perhaps, dual target inhibition.
To see how some of the bisamidines actually bound to the d(CGCGAATTCGCG)2
duplex, we used Xꢀray crystallography, coꢀcrystallizing the DNAꢀligand complexes as described
in the Experimental Section. Full crystallographic data acquisition and structure refinement
details are given in Table 2. Ligand electron density maps and ligandꢀprotein interactions are
given in Supporting Information Figures S1ꢀ4.
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
As can be seen in Figure 3, 1, 6 and 10 all bound to the minor groove of the ATꢀrich
dodecamer, and in each case there are Hꢀbond contacts between the nitrogen atoms in the
ligands, either in the amidine ring, amide bond (1 and 10) or in the indole group (6), with the
bases in the DNA minor groove. Although there are only 3 structures, more interactions correlate
with stronger binding, that is, a bigger ∆Tm: 6 has 6 Hꢀbond contacts and a ∆Tm = 24 ˚C; 1 and
ACS Paragon Plus Environment
11

Journal of Medicinal Chemistry
Page 12 of 47
1
2
3
4
10 have only 3ꢀ4 interactions and a ∆Tm ~ 10 ˚C, Figure 3 and Supporting Information Figures
5
6
S2ꢀ4.
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
ACS Paragon Plus Environment
12

Page 13 of 47
Journal of Medicinal Chemistry
1
2
3
4
Table 2. Data collection and refinement statistics for DNA duplex (CGCGAATTCGCG)2.
5
6
7
DNA/1
DNA/6
DNA/10
8
9
Crystals
(4U8B)
(4U8C)
(4U8A)
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
Data collection
Space group
P212121
P212121
P212121
a, b, c (Å)
23.77, 39.39, 65.36
24.79, 39.95, 65.79
25.18, 40.14, 65.63
Resolution (Å)
No. of reflections
Completeness (%)
R-merge
50.0ꢀ1.31 (1.33ꢀ1.31) 50.0ꢀ1.24 (1.26ꢀ1.24) 50.0ꢀ1.48 (1.52ꢀ1.48)
15,205 (731)
98.7 (100.0)
0.076 (0.66)
29.8 (5.4)
18,244 (760)
95.0 (83.2)
0.068 (0.55)
32.8 (3.6)
11,466 (533)
98.3 (97.1)
0.050 (0.39)
44.5 (5.9)
I/σ(I)
Multiplicity
13.4 (14.3)
13.8 (12.1)
13.8 (12.9)
Refinement statistics
Rꢀwork/Rꢀfree (%)
Geometry deviations
Bond lengths (Å)
Bond angles (°)
24.2/24.6
27.6/33.3
26.6/32.2
0.019
2.76
0.015
2.33
0.015
2.84
B average (Å2)/No. of nonꢀH atoms
DNA
19.7/486
20.2/43
39.5/40
22.8/486
30.5/17
26.3/36
26.1/486
21.6/30
22.9/42
Water
Ligand
Values in parentheses correspond to the highestꢀresolution shells.
ACS Paragon Plus Environment
13

Journal of Medicinal Chemistry
Page 14 of 47
1
2
3
4
5
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
Figure 3. Xꢀray structures of DNA dodecamer duplex (CGCGAATTCGCG)2 showing
bisamidines bind to minor groove and interact with the nucleobases. (A) 1 (cyan) binds to the
DNA minor groove and (B) interacts with A5, A6, C9 and T20 (pink). (C) 6 (yellow) binds to
the DNA minor groove and (D) interacts with T7, T8, C9, G10, T19, and T20 (pink). (E) 10
(green) binds to the DNA minor groove and (F) interacts with T8, C9, and A17 (pink). The
ACS Paragon Plus Environment
14

Page 15 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
largest number of bisamidine contacts correlates with the largest ∆Tm value (shown in
parentheses).
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
Models for cell growth inhibition. These DSC ∆Tm results now enable us to test the
hypothesis that the cellꢀbased activity we see could be due to DNA binding, to UPPS binding, or
to both DNA as well as UPPS binding. We first used a simple heatꢀmap approach, Figure 4A,
from which it can be seen that S. aureus cell growth inhibition can be quite potent and is most
strongly correlated with ∆Tm (both in red/orange), with a weaker correlation with SaUPPS
inhibition (Figure 4A). The trends are similar with E. coli (Figure 4A). The quantitative
correlation between cell growth inhibition and either pIC50 (UPPS) or DNA ∆Tm results are
however poor, as shown in Table S1, varying from 0.16 to 0.60. We thus next employed the
basic approach derived earlier25 in which we used Eqn. 1:
pEC50 (cell) = a•A + b•B + c•C + d
(Eqn. 1)
where in this case A = UPPS pIC50, B = DNA ∆Tm and C is an (optional) computed descriptor
(selected from 307 descriptors we computed in MOE26). With the UPPS pIC50 and DNA ∆Tm
data together, the correlations improved to R2 = 0.73, p = 0.0002 for S. aureus and R2 = 0.53, p=
0.008 for E. coli, Table S1. These results improved further with addition of a computed
descriptor (vsurf_ID5 for S. aureus; GCUT_SMR_3 for E. coli) to R2 = 0.89, p = 4×10ꢀ6 for S.
aureus and R2 = 0.78, p= 0.0003 for E. coli (Table S1). The ∆Tm term was the major contributor
to the S. aureus model while both UPPS inhibition and DNA binding contributed to the E. coli
model. The experimental versus computed cell pEC50 results are shown in Figures 4B, C. There
was no significant correlation between pIC50 (EcUPPS) and ∆Tm (R2 = 0.02, p = 0.6), or between
pIC50 (SaUPPS) and ∆Tm (R2 = 0.003, p = 0.8) when data for all 16 compounds was used.
ACS Paragon Plus Environment
15

Journal of Medicinal Chemistry
Page 16 of 47
1
2
3
4
5
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
Figure 4. Heatꢀmap and correlation plot. (A) Enzyme and cell growth inhibition and DNA
binding heatꢀmap. Red = strong activity; yellow = moderate activity; green = weak/no activity.
(B) Correlation plot for S. aureus experimental and predicted cell activities based on UPPS IC50
and DNA ꢁTm results in addition to 1 mathematical descriptor (vsurf_ID5). (C) Correlation plot
for E. coli experimental and predicted cell activities based on UPPS IC50 and DNA ꢁTm results
in addition to 1 mathematical descriptor (GCUT_SMR_3).
These results suggest, then, an extended model for the potent S. aureus growth inhibition
reported previously6 for 1 in which there is multiꢀtarget inhibition (DNA binding as well as
UPPS inhibition), consistent with the observed synergistic interaction seen with meticillin in a
meticillinꢀresistant strain of S. aureus, as illustrated schematically in Figure 5.
ACS Paragon Plus Environment
16

Page 17 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
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
Figure 5. Schematic illustration of combination multiꢀtargeting in S. aureus. 1 inhibits UPPS as
well as binding to ATꢀrich bacterial DNA and synergizes with meticillin (FICI = 0.25) since
UPPS inhibition decreases flux though the isoprenoid biosynthesis pathway, required for
peptidoglycan formation (via undecaprenyl phosphate, Lipid I and Lipid II).
X-ray structures of SaUPPS and EcUPPS: progress and puzzles. Finally, we sought
to investigate how some of the UPPS inhibitors bound to UPPS. The SaUPPS protein was of
most interest since inhibitors such as 1 bind tightly to SaUPPS and are active in vivo. At present,
there have been no reports of the structures of SaUPPS with (or without) inhibitors, although
there are two reported structures of SaUPPS with its FPP substrate.6,27 Unfortunately, we were
not successful in obtaining any bisamidineꢀbound SaUPPS Xꢀray structures, but we did obtain
the structure of apoꢀUPPS as well as that of a Sꢀthiolo farnesyl diphosphate (FSPP) inhibitor
(IC50 = 1.9 ꢀM). Full crystallographic data acquisition and structure refinement details (PDB ID
codes 3WYI, 4U82) are given in Table 3. Ligand electron density maps and ligandꢀprotein
interactions are given in Supporting Information Figures S5 and S6. The apoꢀSaUPPS structure
(Figure 6A) has a 1.4 Å Cα residue mean square deviation (rmsd) from the EcUPPS apo
structure (PDB ID code 3QAS), while the FSPPꢀSaUPPS structure (Figure 6B) has a 1.1 Å rmsd
ACS Paragon Plus Environment
17

Journal of Medicinal Chemistry
Page 18 of 47
1
2
3
4
5
6
7
8
9
from the FSPPꢀEcUPPS structure (PDB ID code 1X06). These two proteins have a 43%
sequence identity and 62% similarity as calculated by BLAST analysis.28 These strong sequence
and structural similarities suggest that SaUPPS bisamidine inhibitors should bind in a similar
manner to that seen with EcUPPS, to i.e. sites 2, 4 as reported previously.6 The question then
arises as to why we were unsuccessful in obtaining SaUPPS bisamidineꢀbound structures. One
likely possibility is that since in the apoꢀSaUPPS structure, residues 77ꢀ84 (shown in red in
Figure 6A, B) form a loop that blocks entrance to the binding pocket, at least in the crystal. In the
FSPPꢀinhibitor structure, the loop is displaced, as shown in Figure 6C, D, and FSPP, Mg2+ and
sulfate are clearly visible. The molecular basis for this pocket closure in the apo form is not
obvious, but may simply be due to lattice packing effects (since both SaUPPS and EcUPPS are
inhibited by bisamidines, in solution). The inability to obtain a bisamidineꢀbound SaUPPS
structure was also reflected in our inability to obtain a structure with another large inhibitor 19,
consistent with a latticeꢀpacking effect (since 19 is also a good SaUPPS inhibitor, in solution).
We did succeed, however, in obtaining a structure of 19 bound to EcUPPS, Figure 7.
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
ACS Paragon Plus Environment
18

Page 19 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
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
Figure 6. Xꢀray structures of S. aureus UPPS. (A) Closed structure found in apoꢀSaUPPS, 77ꢀ84
in red (PDB ID code 3WYI). (B) Protein surface representation of apo structure. (C) Ligand
bound UPPS showing FSPP, sulfate and Mg2+, open pocket, residues 77ꢀ84 in red (PDB ID code
4U82). (D) Protein surface representation of FSPPꢀbound structure.
As reported previously,29 small bisphosphonates bind to EcUPPS with their lipophilic
sideꢀchains binding to one or more of sites 1, 2, 3 or 4,29 as shown for the inhibitor BPHꢀ629 ([2ꢀ
(3ꢀ(dibenzofuranꢀ2ꢀylꢀphenyl)ꢀ1ꢀhydroxyꢀ1ꢀphosphonoꢀethyl]ꢀphosphonic acid) in Figure 7A.
However, with the bisamidine 14 (Figure 7B) as well as the diacetylenic amine species 18
(which would be expected to protonated), only a single molecule bound to UPPS, and these
molecules were found to span both sites 2 and 4, the cationic groups being relatively solvent
ACS Paragon Plus Environment
19

Journal of Medicinal Chemistry
Page 20 of 47
1
2
3
4
5
6
7
8
9
exposed, while the aromatic groups were buried in the hydrophobic interior of the protein. With
19, we found that sites 1ꢀ3 were all occupied, Figure 7C. Full crystallographic data acquisition
and structure refinement details (PDB ID code 3WYJ) are given in Table 3. Ligand electron
density maps and ligandꢀprotein interactions are given in Supporting Information Figure S7ꢀ8.
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
Figure 7. Xꢀray structures of E. coli UPPS showing ligand binding sites. (A) The
bisphosphonate shown binds to sites 1ꢀ4 (PDB ID code 2E98). (B) Bisamidine 14 binds to sites 2
and 4 (PDB ID code 4H2J). (C) The tetraphosphonate 19 binds to sites 1ꢀ3 (PDB ID code
3WYJ).
ACS Paragon Plus Environment
20

Page 21 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
Table 3. Data collection and refinement statistics for UPPS.
SaUPPS/Apo
SaUPPS/FSPP
EcUPPS/19
6
7
Crystals
(3WYI)
(4U82)
(3WYJ)
8
9
Data collection
Space group
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
P3121
P41212
P212121
a, b, c (Å)
62.08, 62.08, 133.49
25.0ꢀ2.0 (2.07ꢀ2.00)
20,923 (2,047)
100.0 (100.0)
0.066 (0.50)
32.8 (3.6)
57.08, 57.08, 158.62
63.71, 69.00, 108.89
Resolution (Å)
No. of reflections
Completeness (%)
R-merge
50.0ꢀ1.66 (1.69ꢀ1.66) 25.0ꢀ2.1 (2.18ꢀ2.10)
32,040 (1,573)
100.0 (99.8)
0.068 (0.70)
34.6 (4.6)
28,793 (2,803)
99.7 (99.7)
0.049 (0.36)
39.1 (6.0)
I/σ(I)
Multiplicity
8.5 (8.2)
14.3 (14.3)
5.8 (5.4)
Refinement statistics
Rꢀwork/Rꢀfree (%)
Geometry deviations
Bond lengths (Å)
Bond angles (°)
19.0/22.7
18.7/22.0
20.0/24.1
0.015
1.20
0.026
2.42
0.006
1.15
B average (Å2)/No. of nonꢀH atoms
Protein
36.1/1894
55.4/382
ꢀ
19.4/1930
23.7/84
36.6/3504
53.6/426
57.4/60
Water
Ligand
25.3/25
Ramachandran plot
Most favored (%)
Allowed (%)
Disallowed (%)
96.5
3.5
0
97.9
2.1
0
97.2
2.5
0.2
Values in parentheses correspond to the highestꢀresolution shells.
ACS Paragon Plus Environment
21

Journal of Medicinal Chemistry
Page 22 of 47
1
2
3
4
Conclusions
5
6
7
8
The results we have reported here are of interest for a number of reasons. First, we
synthesized and tested a range of amidine and bisamidine compounds against E. coli and S.
aureus UPPS finding quite potent activity with some bisamidines. Second, we found that several
bisamidines also bound to ATꢀrich DNA, shifting Tm by as much as 24 ˚C. Third, we developed
computational models for cell growth inhibition using DNA ∆Tm and UPPS pIC50 values
together with one mathematical descriptor that had R2 = 0.78ꢀ0.90. Fourth, we determined the
structures of 3 bisamidines bound to a synthetic DNA (CGCGAATTCGCG)2 finding minor
groove binding, with the largest ∆Tm correlating with the largest number of ligandꢀDNA
interactions. Fifth, we solved three UPPS structures. Overall, the most important finding is that
the compounds with best cell growth inhibition bind to ATꢀrich DNA in addition to inhibiting
UPPS. This is a potentially important observation since it could lead to novel therapeutic leads
that inhibit bacterial cell wall biosynthesis (targeting UPPS), synergize with existing drugs acting
in these pathways (as does 1 with meticillin in a meticillinꢀresistant S. aureus strain), and also
inhibit DNA replication: multipleꢀtargeting leads that could in some cases help reverse
resistance. It is, however, also possible that there may be toxicity against a potential host with
such multiꢀtarget inhibitors although EC50 values against mammalian cells (manuscript in
preparation) for compound 1 are >100 ꢀM, leading to selectivity indices with S. aureus of > 300.
Moreover, 1 showed no adverse sideꢀeffects in the S. aureus mouse model of infection6 and 7
and 8 have been reported to be effective in mice models of infection with both Bacillus anthracis
as well as Yersinia pestis.11
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
ACS Paragon Plus Environment
22

Page 23 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
Experimental section
Chemical Syntheses: General Methods. All chemicals were reagent grade. Pentamidine
6
7
8
9
1
13
and netropsin were purchased from Aldrich. H NMR and C NMR spectra were obtained on
1
Varian (Palo Alto, CA) Unity spectrometers at 400 and 500 MHz for H and at 100 and 125
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
MHz for 13C. Elemental analyses were carried out in the University of Illinois Microanalysis
Laboratory. HPLC/MS analyses were performed by using an Agilent LC/MSD Trap XCT Plus
system (Agilent Technologies, Santa Clara, CA) with an 1100 series HPLC system including a
degasser, an autosampler, a binary pump, and a multiple wavelength detector. All final
compounds were ≥95% pure as determined by elemental analysis or analytical HPLC/MS
analysis and were also characterized by 1H NMR and HRMS.
N4,N4'-bis(3-(4,5-dihydro-1H-imidazol-2-yl)phenyl)-[1,1'-biphenyl]-4,4'-bicarboxamide (1)
To a mixture of 4,4'ꢀdiphenyl dicarbonyl chloride (1.39 g, 5 mmol), 3ꢀaminobenzonitrile
(1.18 g, 10 mmol) in anhydrous THF (20 mL) was added Et3N (2.1 mL, 15 mmol) and the
mixture stirred at room temperature overnight. After filtration, the white solid was washed with
water (20 mL) and ethyl acetate (10 mL) and then dried. Sodium hydrosulfide hydrate (100 mg),
ethylenediamine (2 mL), and dimethylacetamide (10 mL) were then added and stirred overnight
at 140 °C. Upon solvent removal, the solid was washed thoroughly with water and then ethyl
acetate (10 mL). To a suspension of the crude product in 10 mL of water were added two
equivalents of methanesulfonic acid. Removal of water afforded the final product 1 as its
methanesulfonic acid salt (1.44 g, 40%). 1H NMR (DMSOꢀd6, 500 MHz) δ: 10.68 (s, 2 H), 10.52
(s, 4 H), 8.50 (s, 2 H), 8.12 (d, J = 9.0 Hz, 4 H), 8.02–7.98 (m, 2 H), 7.96 (d, J = 9 Hz, 4 H),
7.68–7.58 (m, 4 H), 4.00 (s, 8 H), 2.36 (s, 6 H). HRMS (ESI): m/z [M + H]+ calculated for
ACS Paragon Plus Environment
23

Journal of Medicinal Chemistry
Page 24 of 47
1
2
3
4
5
6
7
8
9
C32H29N6O2 529.2361, found 529.2352. Purity of the product determined by HPLC
(Phenomenex C6ꢀPhenyl 110A. 100×2 mm, 3 ꢁm, 300 nm, retention time = 5.6 min): 98.9%.
1,4-bis(6-(1,4,5,6-tetrahydropyrimidin-2-yl)-1H-indol-2-yl)benzene (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
Terephthalaldehyde (1.34 g, 10 mmol) and 4ꢀmethylꢀ3ꢀnitrobenzonitrile (3.24 g, 20
o
mmol) were added to a roundꢀbottom flask and heated to 150 C until the compounds melted.
Piperidine (1.5 mL) and sulfolane (10 mL) were added and the resulting solution stirred at 150
oC overnight then cooled to room temperature to yield an orange solid (3.4 g, 80%) which was
washed with methanol (15 mL ×3) and dried. The solid, 4,4'ꢀ(1E,1'E)ꢀ2,2'ꢀ(1,4ꢀ
phenylene)bis(etheneꢀ2,1ꢀdiyl)bis(3ꢀnitrobenzonitrile) was then suspended in triethyl phosphate
o
(30 mL) and heated to reflux (160 C) for 72 h until the mixture turned from orange to yellow.
The suspension was filtered and the yellow solid washed with methanol (15 mL ×3) and dried
(1.4 g, 55%). The solid 2,2'ꢀ(1,4ꢀphenylene)bis(1Hꢀindoleꢀ6ꢀcarbonitrile) was suspended in
propaneꢀ1,3ꢀdiamine (15 mL) and heated to 130 oC for 24 h. The suspension was filtered and the
solid washed with water (15 mL ×3) and methanol (15 mL ×3). The final product 6 was a yellow
1
solid. H NMR (DMSOꢀd6, 500 MHz) δ: 7.99 (s, 4 H), 7.79 (s, 2 H), 7.48 (d, J = 8.0 Hz, 2 H),
7.42 (d, J = 8.0 Hz, 2 H), 6.99 (s, 2 H), 3.37 (s, 8 H), 1.72 (m, 4 H). HRMS (ESI): m/z [M + H]+
calculated for C30H29N6 473.2454, found 473.2450. Purity of the product determined by HPLC
(Phenomenex C6ꢀPhenyl 110A. 100x2 mm, 3 ꢁm, 230 nm, retention time = 5.8 min): 99.1%.
1,4-bis(6-(4,5-dihydro-1H-imidazol-2-yl)-1H-indol-2-yl)benzene (7)
Terephthalaldehyde (1.34 g, 10 mmol) and 4ꢀmethylꢀ3ꢀnitrobenzonitrile (3.24 g, 20
o
mmol) were added to a roundꢀbottom flask and heated to 150 C until the compounds melted.
Piperidine (1.5 mL) and sulfolane (10 mL) were added and the resulting solution stirred at 150
oC overnight then cooled to room temperature to yield an orange solid (3.4 g, 80%) which was
ACS Paragon Plus Environment
24

Page 25 of 47
Journal of Medicinal Chemistry
1
2
3
4
washed with methanol (15 mL ×3) and dried. The solid 4,4'ꢀ(1E,1'E)ꢀ2,2'ꢀ(1,4ꢀ
5
6
phenylene)bis(etheneꢀ2,1ꢀdiyl)bis(3ꢀnitrobenzonitrile) was then suspended in triethyl phosphate
7
8
9
o
(30 mL) and heated to reflux (160 C) for 72 h until the mixture turned from orange to yellow.
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 suspension was then filtered and the yellow solid washed with methanol (15 mL ×3) and
dried (1.4 g, 55%). The solid 2,2'ꢀ(1,4ꢀphenylene)bis(1Hꢀindoleꢀ6ꢀcarbonitrile) was suspended in
o
ethaneꢀ1,2ꢀdiamine (15 mL) and heated to 130 C for 24 h. The suspension was filtered and the
solid washed with water (15 mL ×3) and methanol (15 mL ×3). The final product 7 was a yellow
solid. 1H NMR (DMSOꢀd6, 500 MHz) δ: 11.73 (s, 2 H), 7.97 (s, 4 H), 7.85 (s, 2 H), 7.50 (s, 4 H),
7.00 (s, 2 H), 3.28 (s, 4 H). HRMS (ESI): m/z [M + H]+ calculated for C28H25N6 445.2141, found
445.2139. Purity of the product determined by HPLC (Phenomenex C6ꢀPhenyl 110A. 100×2
mm, 3 ꢁm, 230 nm, retention time = 5.5 min): 99.3%.
1H-Indole, 2,2'-(1,2-ethenediyl)bis[6-(4,5-dihydro-1H-imidazol-2-yl) dihydrochloride (8)
The compound was obtained from the Drug Synthesis and Chemistry Branch,
Developmental Therapeutics Program, Division of Cancer Treatment and Diagnosis, National
Cancer Institute. 1H NMR (methanolꢀd4, 400 MHz): 7.92 (s, 2H), 7.71 (d, 8.4 Hz, 2H), 7.45 (d, J
= 8.8 Hz, 2H), 7.33 (s, 2H), 6.78 (s, 2H), 4.08 (s, 4H). HRMS (ESI): m/z [M + H]+ calculated for
C24H23N6 395.1984, found 395.1986. Purity of the product determined by HPLC (Phenomenex
C6ꢀPhenyl 110A. 100×2 mm, 3 ꢁm, 230 nm, retention time = 5.6 min): 97.2%.
N4,N4'-bis(3-(1,4,5,6-tetrahydropyrimidin-2-yl)phenyl)biphenyl-4,4'-dicarboxamide (10)
To a mixture of 4,4'ꢀdiphenyl dicarbonyl chloride (1.39 g, 5 mmol), 3ꢀaminobenzonitrile
(1.18 g, 10 mmol) in anhydrous THF (20 mL) was added Et3N (2.1 mL, 15 mmol) and the
mixture stirred at room temperature, overnight. After filtration, the white solid was washed with
water (20 mL) and ethyl acetate (10 mL) and then dried. Sodium hydrosulfide hydrate (100 mg),
ACS Paragon Plus Environment
25

Journal of Medicinal Chemistry
Page 26 of 47
1
2
3
4
5
6
7
8
9
propaneꢀ1,3ꢀdiamine (2 mL), and dimethylacetamide (10 mL) were then added and the mixture
stirred overnight at 140 °C. Upon solvent removal, the solid was washed with water and then
ethyl acetate (10 mL). To the suspension of the crude product in 10 mL of water were added two
equivalents of methanesulfonic acid. Removal of water afforded the final product as its
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
methanesulfonic acid salt (1.50 g, 40%). H NMR (DMSOꢀd6, 500 MHz) δ: 10.68 (s, 2 H), 9.95
(s, 4 H), 8.38 (s, 2 H), 8.14 (d, J = 9.0 Hz, 4 H), 7.98 (m, 6 H), 7.62 (t, 2 H), 7.42 (d, 2 H), 3.50
(s, 8 H), 2.29 (s, 4 H). HRMS (ESI): m/z [M + H]+ calculated for C34H33N6O2 557.2665, found
557.2658. Purity of the product determined by HPLC (Phenomenex C6ꢀPhenyl 110A. 100×2
mm, 3 ꢁm, 230 nm, retention time = 5.3 min): 99.6%.
N4,N4'-bis(3-(4,5-dihydro-1H-imidazol-2-yl)phenyl)biphenyl-4,4'-bis(carbothioamide) (11)
N4,N4'ꢀbis(3ꢀ(4,5ꢀdihydroꢀ1Hꢀimidazolꢀ2ꢀyl)phenyl)ꢀ[1,1'ꢀbiphenyl]ꢀ4,4'ꢀbicarboxamide
(1, 560 mg, 1 mmol) was suspended in pyridine (5 mL). Lawesson's reagent (2,4ꢀbis(4ꢀ
methoxyphenyl)ꢀ1,3,2,4ꢀdithiadiphosphetaneꢀ2,4ꢀdisulfide) (1.0 g, 2.5 mmol) was added and the
mixture heated to 150 oC for 72 h at which point the mixture had turned yellow. The suspension
was then filtered and the solid washed with methanol (15 mL ×3) and acetone (15 mL ×3) and
1
then dried under vacuum. The product 11 was obtained as a yellow solid. H NMR (DMSOꢀd6,
500 MHz) δ: 12.12 (s, 2 H), 10.60 (s, 4 H) 8.45 (s, 2 H), 7.90 (d, J = 9.0 Hz, 4 H), 8.00–7.98 (m,
2 H), 7.82 (d, J = 9 Hz, 4 H), 7.73 (m, 4 H), 4.00 (s, 8 H). HRMS (ESI): m/z [M + H]+ calculated
for C32H29N6S2 561.1895, found 561.1896. Purity of the product determined by HPLC
(Phenomenex C6ꢀPhenyl 110A. 100×2 mm, 3 ꢁm, 230 nm, retention time = 5.8 min): 97.6%.
N4-(3-(4,5-dihydro-1H-imidazol-2-yl)phenyl)-N4'-(3-(((2S,3R,4S,5S,6R)-3,4,5-trihydroxy-6-
(hydroxymethyl)tetrahydro-2H-pyran-2-yl)oxy)phenyl)-[1,1 -biphenyl]-4,4 –dicarboxamide
(12)
ACS Paragon Plus Environment
26

Page 27 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
NH
NBoc
NC
N
N
NO2
AcO
NO2
NH2
12a
12b
OAc
OAc
AcO
HO
OAc
OAc
OAc
OAc
12c
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
O
O
Cl3C
NH
OAc
AcO
OAc
AcO
OAc
OAc
H2N
OAc
O2N
O
O
O
O
OAc
12d
12e
O
NBoc
OAc
AcO
O
N
H
N
OAc
OAc
N
H
O
O
12f
O
NH TFA
OH
HO
O
N
H
N
OH
OH
N
H
O
O
12
3ꢀnitrobenzonitrile (3.6 g, 20 mmol) was treated with ethylene diamine (40 mL) and
o
NaHS (200 mg). The solution was heated at 110 C for 48 h and then quenched with water. The
residue was extracted with ethyl acetate, dried under vacuum and purified by flash
chromatography (silica gel, hexane/ethyl acetate = 2/1) to give the dihydroimidazole compound
12a (2.6 g, 60%). To a solution of 12a (2.2 g, 10 mmol) and NaHCO3 (3.6 mg, 40 mmol) in
dichloromethane (20 mL) and water (20 mL) was added (Boc)2O (4.5 g, 20 mmol) with vigorous
o
stirring at 0 C. Stirring was continued for 12 h at which point the mixture had to room
temperature. The dichloromethane phase was separated and concentrated to give the Boc
protected nitro compound. To the Boc protected nitro compound in ethyl acetate (40 mL) was
added Pd/C (10% palladium on charcoal, 145 mg) under nitrogen at room temperature. The
nitrogen was switched to H2 using a balloon and stirring was continued for 12 h at room
temperature. Filtration and concentration under reduced pressure gave the 3ꢀaminophenyl
ACS Paragon Plus Environment
27

Journal of Medicinal Chemistry
Page 28 of 47
1
2
3
4
5
6
7
8
9
imidazoline 12b. Glucose tetraacetate (3.5 g, 10 mmol) was treated with trichloroacetonitrile
(14.4 g, 100 mmol), DBU (0.3 g, 2 mmol) and CH2Cl2 (40 mL). The solution was stirred at room
temperature for 12 h and then washed with water. The residue was extracted with Et2O, dried
under vacuum and purified by flash chromatography (silica gel, hexane/ethyl acetate = 2/1) to
give the peracetyl glucopyranosyl trichloroacetimidate 12c (4.4 g, 90%). A mixture of 12c (4.0 g,
8 mmol), 3ꢀnitrophenol (1.1 g, 8 mmol), and freshly activated 4 Å molecular sieves in anhydrous
CH2Cl2 (30 mL) was stirred under N2 at 0 oC for 10 min, then BF3•Et2O (40 ꢁL, 0.8 mmol) was
added via syringe. Stirring continued for 3 h, then Et3N (2 mL) was added, and the resulting
mixture filtered through Celite. The filtrate was concentrated and purified by silica gel column
chromatography (hexane/ethyl acetate = 3/1) to give 3ꢀnitrophenol glycoside 12d (3.0 g, 80%).
To a solution of 12d (2.3 g, 5 mmol) in ethyl acetate (20 mL) was added Pd/C (10% palladium
on charcoal, 100 mg) under nitrogen at room temperature. The nitrogen was switched to H2 using
a balloon and stirring was continued for 12 h at room temperature. Filtration and concentration
under reduced pressure gave 3ꢀaminophenol glycoside 12e. To a solution of 12b (520 mg, 2
mmol), 12e (920 mg, 2 mmol) and Et3N (1 mL) in dry dichloromethane (20 mL) was added
[1,1'ꢀbiphenyl]ꢀ4,4'ꢀdicarbonyl dichloride (560 mg, 2 mmol) at 0 oC. Stirring was continued for 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
o
h at 0 C. The reaction mixture was washed with water (15 mL ×3), then concentrated under
reduced pressure to give the residue. Purification of the residue by chromatography (silica gel,
hexane/ethyl acetate = 2/1) gave product 12f (800 mg, 45%). To a solution of 12f (450 mg, 0.5
mmol) in MeOH (5 mL) was added MeONa (1 g) at room temperature. Stirring was continued
for 4 h at room temperature and all the volatile components were removed under reduced
pressure. The resulting residue was purified by chromatography (silica gel, ethyl acetate) to give
the nonꢀprotected glycoside. To a solution of the free glycoside (360 mg, 0.5 mmol) in
ACS Paragon Plus Environment
28

Page 29 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
dichloromethane (3 mL) was added trifluoroacetic acid (2 mL) with stirring at room temperature.
Stirring was continued for 4 h then all volatile components were removed under reduced
pressure. The resulting residue was purified by chromatography (silica gel, chloroform/MeOH =
5/1) to give the final product 12 (340 mg, 95%). 1H NMR (methanolꢀd4, 400 MHz) 8.48 (m, 1H),
8.06 (m, 4H), 7.96 (m, 5H), 7.61 (m, 2H), 7.55 (m, 1H), 7.37 (d, J = 8.0 Hz, 1H), 7.26 (t J = 8.0
Hz, 1H), 6.89 (dd, J = 1.6 Hz, 1H), 4.78 (s, 1H), 4.11 (s, 4H), 3.90 (dd, J = 12.0, 2.0 Hz, 1H),
3.71 (dd, J = 12.0, 7.6 Hz, 1H), 3.46ꢀ3.38 (m, 5H). HRMS (ESI): m/z [M + H]+ calculated for
C35H35N4O8, 639.2455; found 639.2446. Purity of the product determined by HPLC
(Phenomenex C6ꢀPhenyl 110A. 100×2 mm, 3 ꢁm, 210 nm, retention time = 5.8 min): 97.9%.
N4-(3-(4,5-dihydro-1H-imidazol-2-yl)phenyl)-N4'-(3-methoxyphenyl)-[1,1-biphenyl]-4,4–
dicarboxamide (13)
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
NH
NBoc
NC
NO2
N
N
NO2
NH2
13a
13b
NBoc
H2N
+
O
Cl
O
N
+
NH2
Cl
O
13b
TFA
NH
NBoc
N
N
O
NH
O
O
NH
HN
HN
O
13c
13
O
O
3ꢀnitrobenzonitrile (3.6 g, 20 mmol) was treated with ethylene diamine (40 mL) and
o
NaHS (200 mg). The solution was heated at 110 C for 48 h and then quenched with water. The
residue was extracted with ethyl acetate, dried under vacuum and purified by flash
chromatography (silica gel, hexane/ethyl acetate = 2/1) to give the dihydroimidazole compound
13a (2.6 g, 60%). To a solution of 13a (2.2 g, 10 mmol) and NaHCO3 (3.6 mg, 40 mmol) in
dichloromethane (20 mL) and water (20 mL) was added (Boc)2O (4.5 g, 20 mmol) with vigorous
ACS Paragon Plus Environment
29