Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 21 7841
oxyanion, shown in orange in Figure 7B. The additional
hydrogen bond made by compounds, 2 and 20 in M. tuber-
culosis Fba are consistent with their greater active site
binding affinity in M. tuberculosis Fba compared to H. pylori
Fba, shown in Table 1.
Enzymatic Test. Fructose bisphosphate and inhibitor, made
up at the appropriate concentration in a glycyl-glycine buffer
(0.1 M pH 7.4, 0.2 M in AcOK), NADH (0.12 mM), glycero-
phosphate dehydrogenase (11 U), triose-phosphate isomerase
(3 U), and aldolase (4 mU), were placed in a cuvette to give
a final volume of 1.2 mL. The decrease in absorbance of
NADH at 340 nm was monitored on a spectrophotometer
over 1-2 min.
Chemical Syntheses. The complete procedure is given in the
Supporting Informations. The determinations of purity of the
four synthesized compounds were performed with a Perkin-
Elemer HPLC system using a Hypersil ODS C18 analytical
column (250 mm ꢀ 4.6 mm) and a linear gradient solvent system:
0.1 M triethylammonium acetate buffer:CH3CN in ratios from
95/5 to 40/60 for 20 min with flow rate 1 mL/min. Peaks were
detected by UV absorption using a diode array detector. Reten-
tion times, min (purity): 1, 4.5 min (95%); 2, 6.8 (96%); 3, 8.7
(96%); 4, 10.2 (95%).
Multiple alignment of class II Fba (Table 2, Supporting
Information) supports greater active site similarity between the
Fba enzymes of C. albicans/Y. pestis and M. tuberculosis than
with the H. pylori Fba. Active site residues that contact
compounds 10, 1, and 2 in M. tuberculosis Fba are conserved
in C. albicans and Y. pestis Fba, including the residues homo-
logous to Gly-56 and Gln-280. Indeed, compounds 1 and 2
exhibit tighter inhibitions of C. albicans and Y. pestis Fbas
similar to M. tuberculosis Fba, as seen in Table 1, compared to
H. pylori Fba. These two residues rationalize the differential
response to active site binding by compounds 10, 1, and 2 within
class II Fbas. The modification of compound 2 by the addition
of lipophilic esters does not significantly impact binding affinity
for compounds 3 and 4 for all class II Fbas, shown in Table 1,
and would indicate that lipophilic esters do not interact with
active residues and likely point toward the bulk solution.
Compound 1 differs from aldolase substrate sedoheptulose-
1,7-bisphosphate in the absence of hydroxyl moieties at C4, C5,
and C6 atoms. In class I aldolase, KM values determined with
sedoheptulose-1,7-bisphosphate are comparable with those
determined for FBP.9 We postulate that 1, in lacking hydroxyl
moieties at atom positions C4, C5, and C6, reduces its ability to
participate in hydrogen bonding interactions with active site
residues of class I aldolase and, together with the apparent
coplanarity of hydroxamate moiety atoms in 1, would further
hinder optimal fit into class I aldolase active site and forms an
additional basis for high selectivity of 1 for class II aldolases.
Future design would exploit stereochemical differences intro-
duced at C4, C5, and/or C6 atoms with respect to FBP and
sedoheptulose-1,7-bisphosphate substrates in order to enhance
inhibitor potency.
Selected analytical data for compounds 1-4:
1 (Disodium Salt). 1H NMR (250 MHz, D2O): δ 4.5 (d, J =
6.5 Hz, 2H), 3.7 (q, J = 6.25 Hz, 2H), 3.5 (t, J = 6.25 Hz, 2H),
1.56-1.44 (m, 2H), 1.62-1.56 (m, 2H). 13C NMR (62.5 MHz,
BB, D2O): δ 171.0, 170.86, 65.0, 64.92, 61.8, 47.88, 27.0, 26.89,
22.11. 31P NMR (101.25 MHz, BB, D2O): δ 1.85, 0.95.
HR-MS (ESI negative): m/z 322.0088 (calcd for C6H14NO10-
P2: 322.0093).
2 (Monosodium Salt). 1H NMR (250 MHz, D2O): δ 4.49 (d,
J = 6.25 Hz, 2H), 3.52 (t, J = 6.5 Hz, 2H), 3.48 (t, J = 6.5 Hz,
2H), 1.52 (p, J = 6.5 Hz, 2H), 1.42 (t, J = 6.5 Hz, 2H). 13C NMR
(62.5 MHz, BB, D2O): δ 171, 61.63, 61.18, 47.90, 28.29, 22.18.
31P NMR (101.25 MHz, BB, D2O): δ 2.47.
HR-MS (ESI negative): m/z 242.0429 (calcd for C7H13NO7P:
242.0430).
1
3 (Monosodium Salt). H NMR (250 MHz, D2O): δ 4.5 (d,
J = 5.7 Hz, 2H), 4.05 (t, J = 5.85 Hz, 2H), 3.55 (t, J = 5.8 Hz,
2H), 2.29 (t, J = 7.5 Hz, 2H), 1.66-1.4 (m, 6H), 1.30-1.10 (m,
4H), 0.77 (t, J = 6.9 Hz, 3H). 13C NMR (62.5 MHz, BB, D2O): δ
177.56, 64.82, 61.51, 47.81, 34.02, 30.54, 25.02, 24.13, 22.46,
21.64, 13.21. 31P NMR (101.25 MHz, BB, D2O): δ 3.72.
HR-MS (ESI negative): m/z 340.1163 (calcd for C12H23NO8-
P: 340.1161).
Antibacterial Evaluation. Compounds 1-4 were assayed
by standard procedures18,19 for inhibition of growth of culti-
vated microorganisms: M. tuberculosis, C. albicans, Y. pestis.
No inhibition was observed at concentrations up to 1 mM.
4 (Monosodium Salt). 1H NMR (250 MHz, D2O): δ 4.51
(d, J = 4.0 Hz, 2H), 4.07-3.93 (m, 2H), 3.58 - 3.45 (m, 2H), 2.2
(t, J = 6 Hz, 2H), 1.66-1.42 (m, 6H), 1.29-1.05 (m, 16H), 0.75
(t, J = 6 Hz, 3H). 13C NMR (62.5 MHz, BB, D2O): δ 174.6,
64.37, 61.63, 34.11, 32.0, 29.85-29.3, 25.36, 24.86, 22.66, 22.53,
13.85. 31P NMR (101.25 MHz, BB, D2O): δ 3.78.
Conclusion
HR-MS (ESI negative): m/z 424.2113 (calcd for C18H35NO8-
P: 424.2100).
We have prepared and evaluated the most powerful and
selective inhibitor of class II Fba reported so far. Nonpho-
sphorylated derivatives of this compound, although less
potent, may serve as lead candidates for the synthesis of
prodrugs to be tested on cultivated pathogenic species. The
detailed structural analysis, in full accordance with the ki-
netics data, indicates that the design of potent hydroxamate
based inhibitors for class II Fbas is conditioned by the absence
or presence of specific nonhomologous active site residues.
These differences canbeexploited in the design of morepotent
and selective inhibitors for therapeutic gain. No inhibition of
growth of cultivated pathogens however was observed so far
with compounds 1 - 4, most probably as a consequence of the
presence of phosphate groups. We propose to synthesize
lipophilic prodrugs of the best inhibitors and/or products
bearing phosphomimetic groups.
Acknowledgment. We are very grateful to Dr. Jean-Michel
Bruneau (NOVEXEL, Romainville, France) for the kind
supply of recombinant Fba from C. albicans. J.S. was sup-
ported by funding from Natural Science and Engineering
Research Council (Canada) and Canadian Institutes for
Health Research. R.D. was supported by a scholarship from
the Region Ile de France. M.J. is supported by the National
Institute of Allergy and Infectious Diseases (NIAID), Na-
tional Institutes of Health (NIH) grant AI078126 and the
National Institute of Neurological Disorders and Stroke
(NINDS), NIH grant NS066438.
Supporting Information Available: Detailed syntheses of
compounds 1-4. Determinations of IC50 and Ki. Structure
solution of H. pylori Fba complexes with compounds 1 and
2. Multiple alignment of class II FBP aldolases. This ma-
terial is available free of charge via the Internet at http://
pubs.acs.org.
Methods
Enzymes. The four class II Fbas were purified recombinant
enzymes, expressed in E. coli. class I Fba from rabbit muscle,
was purchased from Sigma.