Symmetrical and unsymmetrical analogues of isoxyl; active agents against Mycobacterium tuberculosis

Article abstract of DOI:10.1016/j.bmcl.2006.06.095

Symmetrical and unsymmetrical analogues of the antimycobacterial agent isoxyl-have been synthesized and tested against Mycobacterium tuberculosis H37Rv and Mycobacterium bovis BCG, some showing an increased bactericidal effect. In particular, compounds 1-(p-n-butylphenyl)-3-(4-propoxy-phenyl) thiourea (10) and 1-(p-n-butylphenyl)-3-(4-n-butoxy-phenyl) thiourea (11) showed an approximate 10-fold increase in in vitro potency compared to isoxyl, paralleled by increased inhibition of mycolic acid biosynthesis in M. bovis BCG. Interestingly, these isoxyl analogues showed relatively poor inhibition of oleate production, suggesting that the modifications have changed the spectrum of biological activity.

Full text of DOI:10.1016/j.bmcl.2006.06.095

Bioorganic & Medicinal Chemistry Letters 16 (2006) 4743–4747
Symmetrical and unsymmetrical analogues of isoxyl;
active agents against Mycobacterium tuberculosis
Veemal Bhowruth,a, Alistair K. Brown,a, Robert C. Reynolds,b
Geoffrey D. Coxon,c Simon P. Mackay,c David E. Minnikina and Gurdyal S. Besraa,*
aSchool of Biosciences, University of Birmingham, Edgbaston, Birmingham B15 2TT, UK
bDrug Discovery Division, Southern Research Institute, Birmingham, Alabama 35225-5305c, USA
cDivision of Pharmaceutical Science, University of Strathclyde, 27 Taylor Street, Glasgow G4 0NR, UK
Received 31 May 2006; revised 30 June 2006; accepted 30 June 2006
Available online 27 July 2006
Abstract—Symmetrical and unsymmetrical analogues of the antimycobacterial agent isoxyl-have been synthesized and tested
against Mycobacterium tuberculosis H37Rv and Mycobacterium bovis BCG, some showing an increased bactericidal effect. In par-
ticular, compounds 1-(p-n-butylphenyl)-3-(4-propoxy-phenyl) thiourea (10) and 1-(p-n-butylphenyl)-3-(4-n-butoxy-phenyl) thiourea
(11) showed an approximate 10-fold increase in in vitro potency compared to isoxyl, paralleled by increased inhibition of mycolic
acid biosynthesis in M. bovis BCG. Interestingly, these isoxyl analogues showed relatively poor inhibition of oleate production,
suggesting that the modifications have changed the spectrum of biological activity.
ꢀ 2006 Elsevier Ltd. All rights reserved.
Mycobacterium tuberculosis continues to be one of the
most prevalent causes of morbidity and mortality world-
wide.1 The impact of this infection on nearly one-third
of the world’s population justifies extensive research into
new chemotherapeutic reagents.1 Eisman et al. first pub-
lished in vitro and in vivo data on the anti-tuberculosis
activity of substituted thioureas tested in mice and guin-
ea pigs.2 Subsequent modification and biological testing
produced a library of compounds, in which the thiourea
derivative 4,40-diisoamythio-carbanilide (isoxyl, thiocar-
lide, ISO [1]) was shown to have considerable anti-
mycobacterial activity in mice, guinea pigs and rabbits.3,4
Clinical use of ISO began in the 1960s to treat tubercu-
losis.5–8 Furthermore, ISO was shown to be effective
against various multi-drug resistant M. tuberculosis
isolates.9 In an attempt to understand the biological basis
for the activity of ISO, Phetsuksiri et al. investigated its
mode of action and reported that ISO, like isoniazid and
ethanionamide, strongly inhibited the synthesis of
mycolic acids.10
The fatty acid synthase I (FAS-I) in mycobacteria syn-
thesizes long-chain fatty acids, including stearic acid,
which are subsequently desaturated to oleic acid, a ubiq-
uitous constituent of mycobacterial membrane phos-
pholipids.11,12 Over-expression of the M. tuberculosis
putative fatty acid desaturases in Mycobacterium bovis
BCG identified desA3, which encodes the D9-acyl-CoA
desaturase responsible for the biosynthesis of oleic acid.
Over-expression of desA3 also resulted in increased
resistance of M. bovis BCG to ISO, identifying desA3
as a novel target.10
The crystal structure of the related D9-stearoyl-acyl
13
˚
carrier protein desaturase has been resolved to 2.4 A.
By using docking experiments, it has become possible
to suggest modifications to the ISO skeleton that may
improve the affinity of the inhibitor for the active site.
As a result an initial series of symmetrical and unsym-
metrical ISO analogues have been developed as poten-
tial antimycobacterial agents.
Abbreviations: ACP, acyl carrier protein; CoA, coenzyme A; DTT,
dithiothreitol; FAS, fatty acid synthase; FAMEs, fatty acid methyl
ester; MAMEs, mycolic acid methyl ester; MIC, minimum inhibition
concentration; SI, selectivity index.
Keywords: Mycobacterium tuberculosis; Mt; Mycobacterium bovis; Mb;
Mycolic acids; Isoxyl; Inhibitor.
Symmetrical ISO analogues were prepared using a four-
step sequence.14,15 4-Acetomidophenol is alkylated upon
base treatment with an appropriate halide substituent.
The acetamide functionality is deprotected to form the
*
Corresponding author. Tel.: +44 121 415 8125; fax: +44 121 415
5925; e-mail: g.besra@bham.ac.uk
These authors contributed equally to this work.
0960-894X/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.06.095

4744
V. Bhowruth et al. / Bioorg. Med. Chem. Lett. 16 (2006) 4743–4747
Table 1. Structures, in vitro and in vivo results of isoxyl analogues20–22
Compound
Structure
Yield M. tb
MIC99
M. bovis FAMEs and MAMEs Oleate
BCG
Cytoxicity
inhibitione,19 SI23
a
ab
Ketoc FAd
(lg/ml)18 MIC99
a
(lg/ml)18
1
53%
2.0
1.1
60.45 58.63 114.12 61.02
82
O
O
S
N
N
2
3
30% NDf
0.9
0.9
46.37 45.83 110.36 51.62
NDf
O
O
S
N
N
55%
47%
0.39
66.45 42.64 168.36 19.24
72
O
O
S
N
N
4
0.1
0.6
56.65 36.50 132.26 63.99
79.9
O
O
S
N
N
5
6
7
35%
0.2
0.5
0.5
0.4
49.50 41.95 114.40 90.76
43.73 33.92 123.10 94.25
61.27 44.05 136.04 79.96
44
NDf
89
O
O
S
N
N
35% NDf
O
O
S
N
N
55% <0.1
O
O
S
S
N
N
8
9
52% NDf
65% NDf
85% <0.1
0.075
62.66 33.98 132.86 NIg
38.78 30.51 121.47 14.82
53.79 37.40 130.65 NIg
NDf
NDf
100
N
N
0.6
O
S
N
N
10
0.06
O
S
N
N
Line missing

V. Bhowruth et al. / Bioorg. Med. Chem. Lett. 16 (2006) 4743–4747
4745
11
12
84% <0.1
0.7
45.51 26.90 118.61 NIg
87
O
S
N
N
70%
0.78
0.6
62.25 27.39 104.41 NIg
12.4
S
S
S
N
N
Table 1 (continued)
Compound
Structure
Yield M. tb
MIC99
M. bovis
BCG
FAMEs and
MAMEs
Oleate
inhi-
Cyto-
xicity
a
(lg/ml)18 MIC99
bitione,19 SI23
a
ab
Ketoc FAd
(lg/ml)18
13
59%
>6.25
0.7
54.42
35.76 131.38 17.75
NDf
O
S
S
N
N
14
15
16
48%
45%
50%
0.2
NDf
0.2
0.06
0.5
46.50
88.84
50.79
34.60 142.38 NIg
66.09 183.79 53.32
32.56 147.92 48.83
45.5
NDf
48
O
S
S
S
N
N
N
N
O
S
1.1
O
S
N
N
17
18
57%
48%
0.39
1.56
1.5
54.94
62.54
39.12 140.70 66.15
22.1
O
S
N
N
1.75
41.69 158.83 NIg
5.28
S
S
N
N
Line missing
amine under acidic reflux conditions. The resulting
product is coupled using ethanolic carbon disulfide, cat-
alyzed by sulfur to generate the symmetrical ISO ana-
logues. The products were recrystallized and fully
characterized by NMR (1H and 13C) and mass spec-
trometry (1–2 and 5–6) (Table 1).
Unsymmetrical ISO analogues were synthesized by
two methods. Isothiocyanate groups can be coupled
with amines to generate the isothiourea subunit indic-
ative of ISO analogues (Scheme 1). Commercially
available 4-n-butylphenyl isothiocyanate was used to
generate three novel unsymmetrical ISO analogues

4746
V. Bhowruth et al. / Bioorg. Med. Chem. Lett. 16 (2006) 4743–4747
R1
R2
R2
R1
tion of biphenyl to both positions (19) produced
S
poorer in vivo activity in comparison to ISO (1), however
inhibition was shown against, both mycolate and
oleate production. Interestingly the unsymmetrical
biphenyl ISO (18) failed to inhibit oleate production
leading to the possible conclusion that this compound
does not inhibit DesA3 desaturase activity but some
other enzyme specifically involved in the production of
mycolic acids.
+
NCS
N
N
NH2
H
H
Scheme 1.
S
N
N
N
R2
R2
R1
R1
N
S
+
NH2
NH2
N
N
Numerous compounds in this study have similar if not
better activities than that of ISO (see Table 1). Notably,
the use of an aliphatic C4 functionality to either or both
R1 and R2-positions increased the potency of the inhib-
itor (4, 7–8, 10–11 and 14). Compound 10 has been
shown to be the most effective inhibitor in this study
possessing a significant 10-fold increase in potency
against both M. bovis and M. tuberculosis H37Rv. It
shows similar effects with regard to inhibition of myco-
late production, but unusually does not inhibit oleate
production in vitro. This leads to the possibility that
compound 10 does not inhibit the D9-desaturase specif-
ically, but inhibits other enzymes involved in the pro-
duction of mycolic acids.
H
H
R 2
-20ºC
R1
20ºC
NH2
S
N
N
H
N
Scheme 2.
(9–11) (Table 1). 4-n-Butylphenyl isothiocyanate is
coupled with an amine substituent in ethanol at
50 ꢁC and the products recrystallized from hot etha-
nol. These compounds (9–11) were fully characterized
by NMR (1H and 13C) and mass spectrometry. More
recently, Sriram et al. prepared a series of isonicoti-
nyl hydrazones by reacting isonicotinyl hydrazide
(INH) with thiourea units derived from various
phenyl isothiocyanate substituents.16
Interestingly, the oxygen containing compounds pro-
duced in this study indicate that oxygen is required
for a more effective inhibitor of oleate biosynthesis.
Nearly all the compounds that have one or no oxy-
gens tend to be poor inhibitors of oleate biosynthesis,
whereas compounds containing two oxygens (2–7)
inhibit oleate biosynthesis, to the same degree if not
better than ISO (1). Compounds 5–7 all showed a
high degree of activity against the D9-desaturase
inhibitory oleate biosynthesis in vitro 91%, 94% and
80%, respectively.
Due to the lack of commercially available isothiocya-
nate substituents a second method was developed
(Scheme 2). Perkins et al. coupled two different amine
substituents with 1,1-thiocarbonyldiimidazole, by con-
trolling the temperature of the reaction.17 In a one-pot
reaction the first equivalent of the amine substituent is
reacted with 2 equiv of 1,1-thiocarbonyldiimidazole at
ꢀ20 ꢁC in acetonitrile. Once the amine has fully reacted,
the second equivalent of a different amine is added and
the temperature of the reaction is raised to 20 ꢁC. The
desired products were recrystallized and fully character-
ized by NMR (1H and 13C) and mass spectrometry (3, 4,
7 and 13–21) (Table 1).
In conclusion, the analogues in this study have provided
an insight into the functionalities and their roles in inhi-
bition of ISO. The new analogues 10 and 11 have all
show improved activities against M. tuberculosis to such
a degree that the MIC values are nearly the same or bet-
ter than those of other well-known anti-tuberculosis
inhibitors, such as isoniazid and rifampin.
As shown by Phetsuksiri et al. ISO is an anti-mycobac-
terial agent implicated in targeting the membrane-bound
D9-desaturase, DesA3 and unknown targets in mycolic
acid biosynthesis.10 Therefore, aryl thioureas present
themselves as anti-tuberculosis drugs worthy of further
development.
Acknowledgments
This work was supported by the Medical Research
Council; G.S.B. acknowledges support from Mr. James
Bardrick in the form of a Personal Research Chair and
as a former Lister Institute Jenner Research Fellow.
D.E.M. is a Leverhulme Trust Emeritus Research
Fellow. This work was supported by Tuberculosis
Antimicobacterial Acquisition and Coordinating Facility
(TAACF).
The use of an ISO analogue without functional groups
attached to the core thiourea unit (22) clearly showed
that, without modification, poor in vivo and in vitro
activity was observed; therefore the functional R1 and
R2-groups of ISO play a key role in the effectiveness
of the compound. The incorporation of R-group func-
tions (see Table 1) leads to an increase in activity of
the compounds and gives justification for functionality
at both positions. The use of a biphenyl at one or both
of the positions (16–18) was also tolerated in compari-
son to ISO in terms of in vivo activity. The incorpora-
References and notes
1. Dye, C. Lancet 2006, 367, 938.
2. Eisman, P. C.; Geftic, S. G.; Mayer, R. L. Proc. Soc. Exp.
Biol. Med. 1953, 82, 263.

V. Bhowruth et al. / Bioorg. Med. Chem. Lett. 16 (2006) 4743–4747
4747
3. Crowle, A. J.; Mitchell, R. S.; Petty, T. L. Am. Rev.
Respir. Dis. 1963, 88, 716.
4. Urbancik, R.; Trnka, L.; Kruml, J.; Polenska, H. Pathol.
Microbiol. (Basel) 1964, 27, 79.
5. Urbancik, B. Tubercle 1966, 47, 283.
6. Urbancik, B. Antibiot. Chemother. 1970, 16, 117.
7. Lambelin, G. Antibiot. Chemother. 1970, 16, 84.
8. Picone, A.; Di Vincenzo, M.; Russo, C. G. Ital. Chemioter.
1965, 12, 99.
9. Phetsuksiri, B.; Baulard, A. R.; Cooper, A. M.; Minnikin,
D. E.; Douglas, J. D.; Besra, G. S.; Brennan, P. J.
Antimicrob. Agents Chemother. 1999, 43, 1042.
10. Phetsuksiri, B.; Jackson, M.; Scherman, H.; McNeil, M.;
Besra, G. S.; Baulard, A. R.; Slayden, R. A.; DeBarber, A.
E., ; Barry, C. E., 3rd; Baird, M. S.; Crick, D. C.; Brennan,
P. J. J. Biol. Chem. 2003, 278, 53123.
11. Walker, R. W.; Barakat, H.; Hung, J. G. Lipids 1970, 5,
684.
12. Okuyama, H.; Kankura, T.; Nojima, S. J. Biochem.
(Tokyo) 1967, 61, 732.
13. Lindqvist, Y.; Huang, W.; Schneider, G.; Shanklin, J.
EMBO J. 1996, 15, 4081.
14. Hugershof, F. Chemische. Berichte 1899, 32.
15. Gutekunst, G. O.; Gray, H. L. J. Am. Chem. Soc. 1922,
44, 1741.
CSNHPhOCH2CH(CH3)2, 100%); HRMS calcd for
C21H28N2O2S (MH+): 395.1769, found 395.1783; mp
156–158 ꢁC.
21. 1-(p-n-Butylphenyl)-3-(4-n-butoxy-phenyl) thiourea. 4-n-
Butylphenyl isothiourea (1 ml, 5.8 mmol, 1 equiv) was
added to ethanol (30 ml) at room temperature before
addition of 4-butoxyaniline (0.96 g, 5.8 mmol, 1 equiv).
After stirring at 50 ꢁC for 2 h, the reaction mixture is
cooled to room temperature and the crude product
crashed out. The title compound was recrystallized using
1
hot ethanol to give 84% yield (1.75 g). H NMR (CDCl3,
300 MHz) dH: 0.85 (t, 3H, J = 7.4 Hz, OCH2CH2CH2CH 3),
1.05 (t, 3H, J = 7.1 Hz, CH2CH2CH2CH3), 1.25 (m, 2H,
CH2CH3), 1.30–1.50 (m, 2H, CH2CH2CH3), 1.70–1.85 (m,
2H, OCH2CH2CH2), 2.60(m, 2H, CH2CH2CH2CH3), 3.85
(t, 2H, J = 6.5 Hz, OCH2), 6.75 (d, 2H, J = 8.4 Hz, H-30,
H-50), 7.15–7.25 (m, 6H, H-2, H-6, H-3, H-5, H-20, H-60),
7.75 (s, 2H, NH); 13C NMR (CDCl3, 75 MHz) dc: 10.2
(CH2CH3), 13.6 (CH2CH3), 22.0 (CH2CH3), 22.2
(CH2CH3), 33.2 (CH2CH2CH2), 34.9 (OCH2CH2CH3),
69.5 (OCH2), 115.0 (C-30, C-50), 125.1 (C-3, C-5), 127.3
(C-2, C-6), 129.2 (C-20, C-60, C-4), 135.5 (C-1), 141.1
(C-4), 159.2 (C-10), 180.3 (NHCSNH); m/z (EI) 379.4 [M+
Na+] (75%); HRMS calcd for C21H28N2OS (MH+):
354.1653, found 354.1645; mp 137–140 ꢁC.
16. Sriram, D.; Yogeeswari, P.; Madhu, K. Bioorg. Med.
Chem. Lett. 2006, 16, 876.
22. 1-(p-n-Butoxyphenyl)-3-(p-methylbutoxy-phenyl)thiourea.
1,1-Thiocarbonyldiimidazole (434 mg, 2.42 mmol,
2
17. Perkins, J. J.; Zartman, A. E.; Meissner, R. S. Tetrahedron
Lett. 1999, 40, 1103.
equiv) is dissolved in anhydrous acetonitrile and left at
ꢀ20 ꢁC, while 4-(3-methylbutoxy)aniline (200 mg, 1.2
mmol, 1 equiv) was added dropwise to the solution. The
reaction is monitored by TLC until all the aniline
substituent is consumed. The second substituent is added,
4-butoxyaniline (200 mg, 1.2 mmol, 1.1 equiv) is then
added dropwise and left to come to room temperature
gradually. The acetonitrile is reduced in vacuo and the
organic layer is extracted with ethyl acetate. This is
acidified to pH 2 with 1 M HCl. The organic layer is then
washed with water, brine, dried and reduced in vacuo. The
compound is recrystallized cold in ethyl acetate in petrol to
18. Minimum inhibition concentration (MIC99) and whole cell
radiolabelling. MIC99 of ISO analogues against M. bovis
BCG were calculated by growth on solid media and whole
cell radiolabelling are as described previously.9,10
19. In vitro effect of ISO on oleic acid synthesis. The wild type
M. bovis BCG strain was grown in Sauton medium
supplemented with 0.025% tyloxapol. Cells were harvested
by centrifugation, resuspended in 0.25 M sucrose, pulse-
disrupted by probe sonication and centrifugation at
27,000g for 30 min at 4 ꢁC. The D9-stearoyl-CoA desat-
urase activity was assayed as described.10
1
give the title compound in 55% yield (232 mg). H NMR
20. 1,3-Bis-(p-2-methylpropoxyphenyl)thiourea. A solution of
4-(2-methylpropoxy)aniline (500 mg, 3.2 mmol, 1 equiv) in
ethanol (50 ml) was treated with carbon disulfide (0.18 ml,
3.2 mmol, 1 equiv) and sulfur (28 mg, 0.8 mmol, 0.25
equiv), and heated under reflux for 16 h. The ethanol was
removed in vacuo to yield the crude product as an off-
white solid that was recrystallized from methanol, to yield
the title compound in 30% yield (100 mg). 1H NMR
(CDCl3, 300 MHz) dH: 0.98 (d, 12H, J = 6.6 Hz,
CH(CH3)2), 2.12 (m, 2H, CH(CH3)2), 3.72 (d, 4H,
J = 6.5 Hz, OCH2CH), 6.90 (d, 4H, J = 8.8 Hz, H-3,
H-5), 7.37 (d, 4H, J = 8.8 Hz, H-2, H-6), 7.71 (s, 2H,
NH); 13C NMR (CDCl3, 75 MHz) dc: 21.30 (CH(CH3)2),
29.40 (CH(CH3)2), 75.9 (OCH2CH3), 116.41 (C-3, C-5),
128.5 (C-2, C-6), 135.2 (C-1), 158 (C-4), 181 (NHCSNH)
m/z (EI) 372 (M+ 33%), 109 (MH2+–CH2CH(CH3)2 and
(CDCl3, 300 MHz) dH: 0.95–1.05 (m, 9 H, CH3), 1.40 (m,
2H, CH2CH3), 1.65 (m, 2H, CH2CH(CH3)2), 1.70–1.80 (m,
3H, CH(CH3)2 and CH2CH2CH3), 3.95 (m, 4H, OCH2),
6.40 (d, 4H, J = 6.5 Hz, H-3, H-5), 6.60 (d, 4H,J = 6.7 Hz,
H-2, H-6), 7.65 (s, 1H, NH); 13C NMR (CDCl3, 75 MHz)
dc: 14.0 (CH3), 20.1 (CH2CH3), 24.5 (CH(CH3)2), 25.7
(CH(CH3)2), 35.6 (CH2CH2CH3 and CH2CH(CH3)2), 70.9
(OCH2), 115.6 (C-3, C-5), 127.8 (C-2, C-6), 131.2 (C-4),
155.6 (C-1), 180.4 (NHCSNH); m/z (EI) 409.4 [M+ Na+]
(75%); HRMS calcd for C22H30N2O2S (MH+): 386.5514,
found 386.5569; mp 145–147 ꢁC.
23. Compounds were screened by serial dilution to assess
toxicity to a VERO cell line, generally beginning at
10· the MIC. Selectivity index (SI) has defined as the
ratio of the measured IC50 in VERO cells to the MIC
value.