Page 3 of 7
ACS Medicinal Chemistry Letters
against S. aureus. Thus, the 4-phenol derivative 79a was only
The general synthetic routes to the oxadiazoles have been described
previously.4-5 The same approach was used to access the analogs
reported herein, except our focus was on modification of ring D,
which necessitated diversification of the structures at the stage of
the diphenyl ether construction. This was achieved either by direct
nucleophilic substitution and Ullmann coupling reactions of nitriles
12–14 or direct nucleophilic substitution and acylation of 4-
hydroxybenzonitrile 15 (Figure 2). In many cases, ring D variations
could be introduced by using commercially available substituted
phenols (16–48), carboxylic acid derivatives (50, 51, 53 and 54) or
halides (49 and 52). The substituents of several of these analogs
could be used as synthetic handles for further elaboration, the
reactions of which are described in the Supporting Information. In
some cases where ring D was replaced by non-annular substituents,
the requisite ethers were purchased (55–58).
1
2
3
4
5
6
7
8
modestly active (MIC = 16 g/mL), and 2-benzyl alcohol 88a
exhibited no activity. Similarly, benzoic acid derivatives 80a,
81a, and 82a were either poorly active or inactive, and 2-
carboxamide 83a and 2-aniline 84a were inactive. Introduction
of polar substituents onto an aliphatic system also resulted in
compounds with no activity, as observed with 4-
aminocyclohexane derivatives 111a, and 112a. Electron-
withdrawing substituents introduced onto ring D also resulted
in compounds with no activity, as evidenced by 4-benzonitriles
85a and 85b, as well as 4-methylsulfonate 86b and 4-acetate
87b. One exception is 2-nitro derivative 71a, which had an MIC
of 4 g/mL against S aureus.
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
Replacement of the phenyl ring at position D with
heteroaromatic systems was also detrimental to activity. Thus,
benzothiazole derivative 106b was modestly active (MIC = 16
g/mL), while pyridine derivative 107a was inactive (MIC
>128 g/mL). We observed a similar trend with heteroaliphatic
ring systems, such as 4-piperidines (108a, 108b, 109b, 110a),
3-piperidines (117a, 117b, 118a, 118b), 3-azetidines (119a), as
well as the pyrrolidines and tetrahydrofuran discussed above.
Finally, we introduced more complex aliphatic amine-
containing ring systems, giving quinuclidine derivatives 120a
and 120b and tropine derivatives 121a and 121b, however,
these exhibited poor activity against S. aureus (MIC ≥32
g/mL).
We established the SAR for the oxadiazoles by screening
them for antibacterial activity against the aforementioned
ESKAPE panel of bacteria. The analogs reported herein show
activity against Gram-positive bacteria. We specifically
focused our attention on the analogs that exhibited activity
against S. aureus (Figure 3).
In general, hydrophobic substituents, and especially halogens,
were well-tolerated in ring D. Addition of an extra fluorine
(59b) or chlorine (69b) at the 3-position retained activity in the
4-chloropyrazole series, as did substitution by two chlorines at
these same positions (60b). Replacement of the fluorine by an
iodine, either at position 4 (61b, 75a), position 3 (74a), or
position 2 (67a) also retained activity overall. The same trend
was observed with introduction of a bromine atom at position 4
(63a) or position 3 (73a). Introduction of a trifluoromethoxy or
methoxy group at position 4 of the ring (62a, 68a) also retained
activity, as did the presence of a 3-trifluoromethyl ketone
substituent (70a). An azide group at any position of the ring also
maintained a favorable MIC against S. aureus (64a, 65a, 66a).
Interestingly, a benzylic bromide at position 2 (76a) or position
3 (77a) of the ring did not result in a noticeable increase in MIC,
but introduction of this substituent at position 4 (89a)
completely abolished activity.
Compounds 59b, 60b, 61b, 62a, 69b, 78a, and 72c were
selected and further evaluated for in vitro toxicity using the
XTT assay with HepG2 cells and hemolysis of red blood cells
(Table 1). Compound 78a, in which the ring D was replaced
with trifluoromethoxy caused 16% hemolysis at a concentration
of 64 µg/mL; no further studies were done with this compound.
Plasma protein binding was between 90% to 99%. Although
high, it is within the range for 43% of the 1500 most frequently
prescribed drugs,8 including many antibiotics on the market
such as daptomycin, oxacillin, teicoplanin, rifampicin, and
clindamycin.9-12 Compounds 59b, 60b, 61b, 62a, and 69b
showed >1% hemolysis or < 35 for the ratio of XTT/MIC and
were not studied further.
We also explored the effect of tethering rings C and D at their
respective 2-positions to form a dibenzofuran, and we found
that the resulting compounds where ring A was a phenol were
inactive when ring D was unsubstituted (93a) or substituted
with a 4-trifluoromethyl group (94a), 4-fluoro (90a), 4-
methoxy (91a), or 4-trifluoromethoxy (92b). Likewise,
expanding the furan portion with a heteroatom to form
dibenzodioxine 105a resulted in loss of activity.
Table 1. In vitro evaluation of oxadiazoles
Human
MIC
µg/mL
plasma
protein
XTT IC50
(μg/mL)
Hemolysis
%
binding %
8a
2
0.5
1
<1
1.3
1.8
3.1
2.1
2.9
16
98.2 3.2
98.9 1.6
97.8 0.3
90.5 0.3
99.8 0.1
91.2 0.6
ND
75.7 7.3
17.7 1.0
17.2 1.0
30.7 0.8
16.9 0.9
21.0 0.9
84.9 0.9
37.0 0.9
Replacing ring D with linear alkyl substituents resulted in
only one active compound, the trifluoromethoxy derivative 78a,
which contains a phenol for ring A. However, changing ring A
to a 4-chloropyrazole (78b) or an indole (78c) resulted in loss
of activity. The reasons for this are not clear at present. Other
alkyl substituents introduced as surrogates for ring D, such as
methyl (96a, 96b) and tert-butyl (97a) gave compounds with
poor activity, as did sulfonates (98a, 99a), alkylcarbamates
(100a) and amines (101a), and amides (102a, 103a, 104a).
59b
60b
61b
62a
69b
78a
72c
1
2
2
1
1
<1
98.8 0.7
We observed that introducing partially or fully saturated
systems at ring D gave compounds with good activity in a few
cases. Introduction of a nitrogen atom to give 3-pyrrolidines
113a and 114b, or of an oxygen atom to give 3-tetrahydrofuran
115a, decreased or abolished the activity. As a general trend,
introduction of hydrogen-bond-donating substituents at ring D
resulted in compounds with decreased antimicrobial activity
aData from Spink et al., reproduced for comparison
Compound 72c was evaluated in a broader panel of Gram-
positive bacteria and compared to the previous lead oxadiazole
8 and the standards vancomycin and linezolid (Table 2).
Oxadiazole 72c showed better activity against MRSA strains
3
ACS Paragon Plus Environment