Design and synthesis of substrate-mimic inhibitors of mycothiol-S-conjugate amidase from Mycobacterium tuberculosis

Article abstract of DOI:10.1016/j.bmcl.2006.10.031

The Staudinger reaction between a polymer-supported triphenylphosphine reagent and pseudo-disaccharide azides is successfully applied to synthesize a variety of substrate-mimic mycothiol analogs. Screening of this new group of analogs against the mycobacterial detoxification enzyme mycothiol-S-conjugate amidase (MCA) yielded several modest inhibitors (IC₅₀ values around 50 μM) and provided additional structure-activity relationships for future optimization of inhibitors of MCA and its homologs.

Full text of DOI:10.1016/j.bmcl.2006.10.031

Bioorganic & Medicinal Chemistry Letters 17 (2007) 444–447
Design and synthesis of substrate-mimic inhibitors of mycothiol-S-
conjugate amidase from Mycobacterium tuberculosis
Belhu B. Metaferia, Satyajit Ray, Jeremy A. Smith and Carole A. Bewley*
Laboratory of Bioorganic Chemistry, National Institute of Diabetes and Digestive and Kidney Diseases,
National Institutes of Health, Bethesda, MD 20892, USA
Received 22 August 2006; revised 3 October 2006; accepted 12 October 2006
Available online 17 October 2006
Abstract—The Staudinger reaction between a polymer-supported triphenylphosphine reagent and pseudo-disaccharide azides is suc-
cessfully applied to synthesize a variety of substrate-mimic mycothiol analogs. Screening of this new group of analogs against the
mycobacterial detoxification enzyme mycothiol-S-conjugate amidase (MCA) yielded several modest inhibitors (IC50 values around
50 lM) and provided additional structure–activity relationships for future optimization of inhibitors of MCA and its homologs.
ꢀ 2006 Published by Elsevier Ltd.
Mycothiol (MSH, 1, Fig. 1) is a low molecular weight
thiol1–3 found exclusively in actinomycetes,4 a group
that includes the pathogen Mycobacterium tuberculosis.
Mycothiol functions analogously to glutathione in
Gram-negative bacteria and eukaryotes and plays an
important role as a detoxifying agent by covalently
binding to toxins, electrophiles, and drugs to form myc-
othiol-S-conjugates (MSE).5 This conjugate subsequent-
ly is cleaved at the amide bond by a detoxification
enzyme mycothiol-S-conjugate amidase (MCA) to yield
a mercapturic acid conjugate that is ultimately exported
from the cell.5 Mycothiol also is involved in maintaining
a redox equilibrium within the cell that is critical for a
reducing cellular environment.4
are excellent choices for rational drug design and screen-
ing of small molecule inhibitors.
Previously, we and others have reported the total synthe-
sis of MSH and mycothiol bimane (Fig. 1, MSmB)7,8
which are key for performing structural studies of sub-
strates and inhibitors.9 In addition, we have identified
natural products and natural product-like synthetic
inhibitors that have shown inhibitory activities against
MCA.10–13 Here, as part of our ongoing effort to discover
new generations of anti-mycobacterial agents, we report
synthesis and evaluation of novel substrate-mimic inhib-
itors of MCA built upon a quinic acid-derived scaffold.
Our synthetic strategy to produce MSH-inspired sub-
strate-mimic analogs centered on incorporation of a qui-
nic acid template to replace the inositol ring of natural
MSH. This strategy is attractive for a number of rea-
sons: incorporation of quinic acid circumvents the
otherwise laborious and low-yielding transformations
to obtain optically pure myo-D-inositol-containing ana-
logs;14,15 quinic acid is commerically available as a single
isomer; it can be coupled to a glycosyl donor with high
efficiency to produce a variety of analogs; and the pres-
ence of a relatively unreactive tertiary hydroxyl group as
a site of attachment for linkers makes the synthetic
scheme amenable to parallel synthesis.
The absence of MSH in eukaryotes and Gram-negative
bacteria, along with the compelling evidence for in-
creased sensitivity of MSH-deficient mutants toward
electrophiles, free radicals, and antibiotics6, suggest that
enzymes involved in MSH biosynthesis and MSH-de-
pendent detoxification (namely MCA and biosynthetic
enzymes MshA–MshD) are potential targets for new
classes of antibiotics. Moreover, because inhibitors of
these enzymes may lead to new classes of compounds
that target with some specificity to mycobacteria, they
Retrosynthetic analysis (Scheme 1) of quinic acid-con-
taining analogs led us to envision the use of protected
acceptor 3, readily derived from quinic acid; and glyco-
syl donor 5 that could be obtained from D-mannopyra-
Keywords: Quinic acid; Mycothiol bimane; Myo-D-inositol; Polymer-
supported triphenyl phosphine; Pseudodisaccharide; Mycobacteria.
*
Corresponding author. Tel.: +1 301 594 5187; e-mail: caroleb@
mail.nih.gov
0960-894X/$ - see front matter ꢀ 2006 Published by Elsevier Ltd.
doi:10.1016/j.bmcl.2006.10.031

B. B. Metaferia et al. / Bioorg. Med. Chem. Lett. 17 (2007) 444–447
445
OH
O
HO
HO
HO
OH
HN
NHAc
O
HS
OH
HO
OH
OH
O
MSH (1)
O
OH
OH
O
HO
HO
N
N
HO
HO
HO
O
O
E
OH
NHAc
OH
HN
OH
NHAc
OH
OH
O
O
S
S
OH
OH
OH
O
O
(E = electrophile, drugs etc)
MSmB
MSE
Figure 1. Mycothiol (1, MSH), MCA substrate mycothiol bimane (MSmB), and mycothiol-S-conjugate (MSE).
O
O
HO
O
OH
OH
O
O
O
Glycosylation
OAc
HO
O
O
OH
Staudinger
ligation
O
OH
2
O
3
O
N3
OAc
AcO
Cl
AcO
OH
OAc
OAc
OAc
OAc
O
OAc
4
6
O
N3
AcO
5
Scheme 1. Retrosynthetic analysis.
nose derivative 4. Glycosylation of 3 with 5 would then
furnish the core structure 6 for subsequent Staudinger
amidation to produce diverse mycothiol analogs.
Cl
OAc
OAc
O
O
O
N3
O
O
O
OAc
O
5
O
OAc
OAc
AgOTf, DTBMP
OH
Mol. Sieves,CH2Cl2
O
O
N3
AcO
The forward synthesis of target analogs proceeded with
preparation of glycosylation counterparts 3 and 5. Qui-
nic acid was transformed to protected acceptor 3 in four
steps according to published procedures (Scheme 2).16,17
Treatment of acid 2 with 2,2-dimethoxypropane under
acidic conditions gave the protected lactone that upon
reduction with NaBH4 gave triol 7. Subsequent protec-
tion of the vicinal diol of 7 provided compound 3 in
65% overall yield.
O
3
6
Scheme 4. Glycosylation reaction.
with inversion of configuration at C-2 (Scheme 3).
Treatment of 8 with a,a-dichloromethyl methyl ether
(DCMME) and ZnCl2 in refluxing chloroform gave gly-
cosyl chloride 5 in good yield (Scheme 4).
Glycosyl donor 3,4,6-tri-O-acetyl-2-azido-2-deoxy-a-D-
glucopyranosyl chloride (5) was obtained starting from
The reaction between acceptor 3 and donor 5 was effect-
ed by activation with silver triflate in the presence of
2,6-di-tert-butyl-4-methylpyridine (DTBMP) in dichlo-
romethane to give pseudo-disaccharide 6 with high a
selectivity (5:1 a/b ratio) and 71% yield.19
commercially
available
1,3,4,6-tetra-O-acetyl-b-D-
mannopyranose (4) (Scheme 3). Following published
protocols,7,18 compound 4 was converted to its corre-
sponding triflate by treatment with Tf2O in Py/CH2Cl2,
and the triflate was reacted with sodium azide to give 8
Inspired by recent reports describing the use of a poly-
mer-supported triphenylphosphine reagent for derivati-
zation of azides, we explored a similar method to
derivatize 6 efficiently.20–22 Thus, azido pseudo-disac-
charide 6 was treated with polymer-supported triphenyl-
phosphine reagent followed by in situ trapping of the
iminophosphorane intermediate with 14 varied acid
chlorides (Scheme 5).23 This sequence produced the
O
OH
O
HO
HO
O
1. 2,2-dimethoxypropane
acetone, p-TsOH
OH
OH
2,2-dimethoxypropane
p-TsOH
2. NaBH4, EtOH
3. NaCl/H2O
OH
O
OH
O
HO
O
OH
O
2
3
7
Scheme 2. Preparation of acceptor 3.
Cl
OAc
1. Tf2O,Py
CH2Cl2
DCMME
OAc
OAc
AcO
O
OAc
OAc
AcO
N3
ZnCl2
O
OAc
O
N3
OAc
OAc
CHCl3, 600C
OAc
2. NaN3,DMF
OH
8
5
4
Scheme 3. Synthesis of donor 5.

446
B. B. Metaferia et al. / Bioorg. Med. Chem. Lett. 17 (2007) 444–447
Table 1. Yields of analogs 9a–9n and inhibition of mycothiol-S-
conjugate amidase
desired amide analogs 9a–n cleanly and with good yields
ranging from approximately 40 to 80% (Table 1).
Among the acid chlorides used for derivatization, reac-
tion with p-nitrobenzoyl chloride (to give 9f) and p-tri-
fluoromethyl benzoyl chloride (to give 9g) gave the
highest overall yields (ꢀ80%), while the reaction with
p-(N,N-dimethyl amino) benzoyl chloride (to give 9e)
gave the lowest yield (ꢀ40%) for its amide product.
These differences likely arise from the respective electron
withdrawing versus electron donating character of the
para substituents of the acid chlorides used to generate
9f/g versus 9e, and their respective activating or deacti-
vating effect on the carbonyl center that undergoes at-
tack by the iminophosphorane intermediate. The final
stages of the synthesis were accomplished by global
deprotection to give the desired mycothiol analogs 9a
through 9n.
HO
HO
HO
OH
O
OH
O
R
HN
OH
HO
O
Analog
R
% inhibition of MCAa
15
% yieldb
65
9a
O
9b
9c
9d
9e
9f
13
13
46
44
15
16
13
NI
15
10
NI
26
11
56
62
71
39
78
79
60
58
52
50
56
49
67
S
Compounds 9a–n were tested for inhibition of recombi-
nant MCA using a fluorescence detected HPLC assay
employing mycothiol bimane (MSmB, Fig. 1) or its
cyclohexyl thioglycoside analog13 as substrate as de-
scribed previously.11 Compounds were tested at two
concentrations, namely 100 lM and 50 lM, and results
from screening at 50 lM are summarized in Table 1.
Compounds 9d and 9e were found to be the most active
against MCA inhibiting the enzyme by approximately
45% at 50 lM to indicate that derivatization with nitro-
gen-containing moieties such as pyridinyl or para-amino
phenyl groups contributes favorably to binding of this
scaffold to MCA.
N
N
O2N
F3C
9g
9h
9i
Recent reports describing high density mutagenesis or
systematic gene silencing in M. tuberculosis have provid-
ed further evidence that MSH is essential for growth and
survival of this pathogen.24,25 Discovery and synthesis
of small molecule inhibitors and probes of MSH biosyn-
thesis and detoxification will thus continue to be of great
importance in relation to the need for new antitubercu-
lars. Previously, Knapp and co-workers demonstrated
that the cyclitol present in natural MSH can be replaced
by a cyclohexyl unit to obtain mycothiol conjugates and
analogs that can function, respectively, as substrates14
or inhibitors.15 On the other hand, through synthesis
S
N
9j
S
Cl
N
9k
9l
S
Cl
N
F
O
N
Cl
N
9m
9n
Cl
O
S
Cl
O
O
O
O
O
O
1.
O
PPh3
S
CH2Cl2
O
O
a Values are means of two experiments with final inhibitor concentra-
tion of 50 lM; standard deviations were on average 8%; (NI = less
than 10% inhibition).
O
O
OAc 2. RCOCl
CH2Cl2
OAc
OAc
O
R
O
N3
AcO
HN
AcO
OAc
O
b % overall yields for Scheme 5.
6
OH OH
HO
OH
1. NaOMe/MeOH
O
OH
2. TFA/H2O
CH2Cl2
O
R
HN
HO
OH
of MSH precursors D-GlcN-myo-D-Ins and D-GlcN-
myo-L-Ins we showed previously that the stereochemis-
try of the hydroxyl groups on the cyclitol are critical
as myo-L-Ins-containing substrates were tolerated by
the biosynthetic or detoxification enzymes.26
O
9a - 9n
Scheme 5. Staudinger amidation employing polymer-supported tri-
phenylphosphine reagent.

B. B. Metaferia et al. / Bioorg. Med. Chem. Lett. 17 (2007) 444–447
447
15. Knapp, S.; Amorelli, B.; Darout, E.; Ventocilla, C. C.;
Goldman, L. M.; Huhn, R. A.; Minnihan, E. C. J.
Carbohydr. Chem. 2005, 24, 103.
16. Matsuo, K.; Matsumoto, T.; Nishiwaki, K. Heterocycles
1998, 48, 1213.
17. Phoon, C. W.; Abell, C. J. Comb. Chem. 1999, 1, 485.
The work described here, which incorporates a new moi-
ety in place of myo-D-Ins, namely a quinic acid-derived
ring system, provides further insight into the specificity
of MSH-associated enzymes and their substrates. In
summary, starting from quinic acid and a suitable azido
glycosyl chloride intermediate, we have generated with
efficiency a small but diverse library of new MCA inhib-
itors by employing a Staudinger reaction using polymer-
supported triphenylphosphine. The discovery that two
of these compounds inhibit the M. tuberculosis enzyme
MCA with IC50 values around 50 lM provides a start-
ing point for future elaboration of these scaffolds, which
may yield more potent MCA inhibitors and provide
additional structure–activity relationships.
´
18. Pavliak, V.; Kovac, P. Carbohydr. Res. 1991, 210, 333.
19. General procedure for glycosylation: In a flame-dried two-
necked round bottomed flask under nitrogen, 5 (1.70 g,
4.88 mmol) and 3 (630 mg, 2.44 mmol) were dissolved in
CH2Cl2 (12 mL) and cooled to ꢁ40 ꢁC. To this solution,
2,6-di -tert-butyl-4-methylpyridine (1.5 g, 7.32 mmol) and
˚
1 g of freshly dried and powdered 4 A molecular sieves
were added and stirred. Silver triflate (ACROSS Organics)
(1.25 g, 4.88 mmol) in four portions was added over a 2-h
period in the dark and the reaction was carriedout for
12 h. After filtering through a Celiteꢂ pad, the reaction
mixture was diluted with ethyl acetate; washed with 10%
aqueous sodium thiosulfate, water, and brine; dried over
anhydrous sodium sulfate and evaporated to dryness to
give 6 (71% yield). Compound (6): Rf = 0.23 (40% EtOAc/
Acknowledgments
We thank Drs. John Lloyd for HRMS data and Son
Lam for contributory discussions. This work was sup-
ported in part by the Intramural Research Program,
NIH (NIDDK), and the AIDS Targetted Antiviral
Program of the Office of the Director (C.A.B.).
22
hexanes); ½aꢂD +50.7ꢁ (c 0.6, CHCl3); 1H NMR (300 MHz,
CDCl3, d)1.35 (s, 3H), 1.38 (s, 3H), 1.39 (s, 3H), 1.40 (s,
3H), 1.89 (dd, J = 4.8, 5.1 Hz, 2H), 2.03 (s, 3H), 2.08 (s,
6H), 2.18 (m, 2H), 3.32 (dd, J = 3.6 Hz, 1H), 3.78 (q,
J = 8.5, 9.0 Hz, 2H), 4.03 (m, 3H), 4.17 (m, 1H), 4.31 (m,
2H), 5.03 (t, J = 9.3, 9.9 Hz, 1H), 5.44 (t, J = 10.2,
10.5 Hz, 1H), 5.49 (d, J = 3.6 Hz, 1H); 13C NMR
(75 MHz, CDCl3) d 25.99, 27.08, 27.91, 28.31, 35.62,
38.59, 61.06, 62.14, 68.12, 68.81, 70.47, 73.61, 74.23, 74.54,
78.79, 79.24, 97.10, 109.29, 110.19, 169.86, 170.22, 170.80;
HRMS (ES+) m/z: [M+H]+ calcd for C25H38N3O12:
572.2455; found 572.2480, D = 4.3 ppm.
Supplementary data
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/
j.bmcl.2006.10.031
20. Root, Y. Y.; Bailor, M. S.; Norris, P. Synth. Commun.
2004, 34, 2499.
21. He, Y.; Hinklin, R. J.; Chang, J.; Kiessling, L. L. Org.
Lett. 2004, 6, 4479.
References and notes
22. Lindsley, C. W.; Zhao, Z.; Newton, R. C.; Leister, W. H.;
Strauss, K. A. Tetrahedron Lett. 2002, 43, 4467.
23. General procedure for amidation: To an ice cooled solution
of 6 (20 mg, 0.035 mmol) in dichloromethane (1 mL) was
1. Sakuda, S.; Zhou, Z.; Yamada, Y. Biosci. Biotechnol.
Biochem. 1994, 58, 1347.
2. Spies, H. S. C.; Steenkamp, D. J. Eur. J. Biochem. 1994,
224, 203.
added
a
polymer-supported
triphenylphosphine
3. Newton, G. L.; Bewley, C. A.; Dwyer, T. J.; Horn, R.;
Aharonowitz, Y.; Faulkner, D. J.; Fahey, R. C. Eur. J.
Biochem. 1995, 230, 821.
4. Fahey, R. C. Annu. Rev. Microbiol. 2001, 55, 333.
5. Newton, G. L.; Av-Gay, Y.; Fahey, R. C. Biochemistry
2000, 39, 10739.
6. Rawat, M.; Newton, G. L.; Ko, M.; Martinez, G. J.;
Fahey, R. C.; Av-Gay, Y. Antimicrob. Agents Chemother
2002, 46, 3348.
´
7. Nicholas, G. M.; Kovac, P.; Bewley, C. A. J. Am. Chem.
Soc. 2002, 124, 3492.
8. Lee, S.; Rosazza, J. P. N. Org. Lett. 2004, 6, 365.
9. Mahadevan, J.; Nicholas, G. M.; Bewley, C. A. J. Org.
Chem. 2003, 68, 3380.
10. Nicholas, G. M.; Eckman, L.; Ray, S.; Hughes, R. O.;
Pfefferkorn, J. A.; Barluenga, S.; Nicolaou, K. C. Bioorg.
Med. Chem. Lett. 2002, 12, 2487.
11. Nicholas, G. M.; Eckman, L. L.; Newton, G. L.; Fahey,
R. C.; Ray, S.; Bewley, C. A. Bioorg. Med. Chem. 2003,
11, 601.
12. Fetterolf, B.; Bewley, C. A. Bioorg. Med. Chem. Lett.
2004, 14, 3785.
(2.04 mmol/g loading, 26 mg). The mixture was gently
stirred for approximately 5 min at room temperature after
which acid chloride (0.07 mmol) was added and the
mixture stirred until no additional nitrogen gas was
liberated. The mixture was gently refluxed at 40 ꢁC for
12 h, cooled to room temperature, filtered by gravity, and
washed with dichloromethane (2· 3 mL). To the filtrate
was added acid chloride scavenger MP-Trisamine
(ꢀ3 mmol/g loading, 58.3 mg, 0.175 mmol) and the mix-
ture was gently stirred for 2 h at room temperature. The
mixture was again filtered (gravity), washed with dichlo-
romethane (3· 5 mL), and concentrated in vacuo. The
resulting crude products were purified by preparative TLC
to give the respective protected amides which were
subsequently treated with 0.5 M NaOMe in MeOH
followed by TFA/CH2Cl2/H2O (9:1:1) to give the final
deprotected analogs 9a–n. Spectral data for each of the
final products and for six of the 14 protected amides are
available online as Supporting Information.
24. Sassetti, C. M.; Rubin, E. J. Proc. Natl. Acad. Sci. U.S.A.
2003, 100, 12989.
25. Sareen, D.; Newton, G. L.; Fahey, R. C.; Buchmeier, N.
J. Bacteriol. 2003, 185, 6736.
26. Nicholas, G. M.; Eckman, L. L.; Kovac, P.; Otero-
Quintero, S.; Bewley, C. A. Bioorg. Med. Chem. 2003, 11,
2641.
13. Nicholas, G. M.; Bewley, C. A. Curr. Med. Chem. Anti-
Infective Agents 2004, 3, 221.
14. Knapp, S.; Gonzalez, S.; Myers, D. S.; Eckman, L.;
Bewley, C. A. Org. Lett. 2002, 4, 4337.