Thiahomoisocitrate: A highly potent inhibitor of homoisocitrate dehydrogenase involved in the α-aminoadipate pathway

Article abstract of DOI:10.1016/j.bmc.2007.12.002

Homoisocitrate dehydrogenase is involved in the α-aminoadipate pathway of l-lysine biosynthesis in higher fungi such as yeast and human pathogenic fungi. This enzyme catalyzes the oxidative decarboxylation of (2R,3S)-homoisocitrate into 2-ketoadipate using NAD⁺ as a coenzyme. A series of aza-, oxa-, and thia-analogues of homoisocitrate was designed and synthesized as an inhibitor for homoisocitrate dehydrogenase. Among them, thia-analogue showed strong competitive inhibitory activity as K_i = 97 nM toward homoisocitrate dehydrogenase derived from Saccharomyces cerevisiae. Kinetic studies suggested that the formation of the enolate intermediate played an important role in inhibition.

Full text of DOI:10.1016/j.bmc.2007.12.002

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry 16 (2008) 3372–3376
Thiahomoisocitrate: A highly potent inhibitor of homoisocitrate
dehydrogenase involved in the a-aminoadipate pathway
Takashi Yamamoto and Tadashi Eguchi*
Department of Chemistry and Materials Science, Tokyo Institute of Technology, O-okayama, Meguro-ku, Tokyo 152-8551, Japan
Received 29 September 2007; revised 1 December 2007; accepted 4 December 2007
Available online 8 December 2007
Abstract—Homoisocitrate dehydrogenase is involved in the a-aminoadipate pathway of L-lysine biosynthesis in higher fungi such as
yeast and human pathogenic fungi. This enzyme catalyzes the oxidative decarboxylation of (2R,3S)-homoisocitrate into 2-ketoad-
ipate using NAD+ as a coenzyme. A series of aza-, oxa-, and thia-analogues of homoisocitrate was designed and synthesized as an
inhibitor for homoisocitrate dehydrogenase. Among them, thia-analogue showed strong competitive inhibitory activity as
Ki = 97 nM toward homoisocitrate dehydrogenase derived from Saccharomyces cerevisiae. Kinetic studies suggested that the forma-
tion of the enolate intermediate played an important role in inhibition.
Ó 2007 Elsevier Ltd. All rights reserved.
1. Introduction
2. Results and discussion
2.1. Inhibitor design and synthesis
In some of bacteria and higher fungi such as yeast and
human pathogenic fungi as Candida albicans, Cryptococ-
cus neoformans, and Aspergillus fumigatus, and plant
pathogens like Magnaporthe grisea, L-lysine is biosyn-
thesized via the a-aminoadipate pathway.1 Since this
pathway is unique for these organisms, enzymes in-
volved in this pathway are considered to be a potential
target for new antifungal drugs.2 Homoisocitrate dehy-
drogenase (HICDH, EC 1.1.1.87) is involved in the
fourth step of the a-aminoadipate pathway and cata-
lyzes the conversion of homoisocitrate (HIC) to 2-
ketoadipate through the oxidation of a hydroxy group
by NAD+, the subsequent decarboxylation, and the final
protonation of enolate as shown in Figure 1.1 We have
been involved in designing substrate analogues and
inhibitors for HICDH.3 In this paper, we describe the
first successful example of design and synthesis of a
highly potent inhibitor for HICDH. Evaluation of the
synthesized inhibitor is also described.
It is well known that substitution of a heteroatom such
as sulfur and oxygen at the a-position of a ketone en-
hances the acidity of a-carbon and increases the stability
of its enolate form.4,5 We hence designed a potential
inhibitor for homoisocitrate dehydrogenase by introduc-
tion of a heteroatom into the 3-position of HIC as
shown in Figure 2. When such a compound is once ac-
cepted by HICDH and the enzyme reaction proceeds,
the stability of the intermediary enolate (or enol) in
the HICDH reaction would be increased by heteroatom
substitution and the enolate intermediate might reside in
the active site. Therefore, these compounds may act as
an inhibitor of HICDH. Based on this idea, we designed
and synthesized a series of aza-, oxa-, and thia-ana-
logues 1–3 as a potential inhibitor for HICDH.
The preparation of 1–3 was straightforward and is de-
picted in Scheme 1. Aza-analogue 1 was obtained in
82% yield by treatment of glycine with epoxyfumarate
under basic conditions. Synthesis of oxa-analogue 2
was achieved by treatment of diethyl meso-tartrate with
methyl iodoacetate in the presence of silver (I) oxide, fol-
lowed by alkaline hydrolysis to give 2 in 72% yield (2
steps). Thia-analogue 3 was prepared by the same proce-
dure for synthesis of 1 using thioglycolate instead of gly-
cine in 82% yield.
Keywords: Homoisocitrate dehydrogenase; Lysine biosynthesis; Inhib-
itor; Substrate analogue.
*
Corresponding author. Tel./fax: +81
eguchi@cms.titech.ac.jp
3
5734 2631; e-mail:
0968-0896/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2007.12.002

T. Yamamoto, T. Eguchi / Bioorg. Med. Chem. 16 (2008) 3372–3376
3373
Figure 1. Reaction catalyzed by HICDH.
strate activity (Km = 150 lM, kcat = 2.0 sÀ1), oxa- and
thia-analogues (2 and 3) revealed very low substrate
activity. Furthermore, oxa- and thia-analogues (2 and
3) were found to act as a strong competitive inhibitor
against HIC in ScHICDH reaction, which indicates that
they bind to the homoisocitrate binding site. Notably,
thia-analogue 3 appeared to significantly inhibit the
ScHICDH reaction (Ki = 97 nM). This is the first suc-
cessful example of design and synthesis of a highly po-
tent inhibitor for ScHICDH. Since substitution by
sulfur is known to be the most effective for stabilization
of enolate rather than oxygen and nitrogen,5 the inhibi-
tory activities of 1–3 were in good agreement to ability
for the stabilization effect as we anticipated. Similar
inhibitors by heteroatom substitution were previously
reported as 3-mercapto-2-ketoglutarate and 3-methyl-
mercapto-2-ketoglutarate toward isocitrate dehydroge-
nase,6 although details of the inhibition mechanism
have not been discussed.
Figure 2. Heteroatom substituted analogues.
2.2. Inhibitory activity
Since thia-analogue strongly inhibited the enzyme reac-
tion, our attention was thus focused to which step of the
ScHICDH reaction was slowed by this analogue. At
first, the pre-steady state kinetic analysis of thia-ana-
logue was performed. The enzyme reactions with thia-
analogue were carried out in the presence of different
concentrations of thia-analogue and a fixed concentra-
tion of ScHICDH and NAD+ by a stopped-flow instru-
ment. The two-phase reaction was observed under
various concentrations of thia-analogues as shown in
Figure 3. The rate of the NADH formation in the first
The synthesized racemic analogues 1–3 were subjected
to reaction with HICDH derived from Saccharomyces
cerevisiae (ScHICDH).3 HICDHs from Saccharomyces
species are known to be highly homologous to fungal
HICDHs in the amino acid level. The reactions were
monitored by measuring the formation of NADH from
NAD+ as described previously,3 and the reaction kinet-
ics were analyzed by double reciprocal plots. As shown
in Table 1, while aza-analogue 1 showed moderate sub-
Scheme 1. Synthesis of compounds 1–4.

3374
T. Yamamoto, T. Eguchi / Bioorg. Med. Chem. 16 (2008) 3372–3376
Table 1. Kinetic data of synthetic compounds 1–4 and HIC
as usual and that inhibition of ScHICDH reaction by
thia-analogue was caused by basically slow down after
the oxidation by NAD+, in which decarboxylation and
protonation are involved.
Ki (lM)
KM (lM)
kcat (sÀ1
)
1
—
10
150
20
2.0
0.59
nd
2
3
0.097
12a
nd
—
18b
When the enolate intermediate ii in Figure 2 plays an
important role in the inhibition mechanism, the enzyme
reaction product of thia-analogue would also inhibit
HICDH reaction after keto–enol tautomerization. Thus,
we examined inhibition activity of the plausible enzyme
reaction product 4 as a product inhibitor. Compound 4
was synthesized by adding thioglycolate to a slightly ex-
cess of bromopyruvate as shown in Scheme 1,8 and the
reaction product was purified by ion-exchange chroma-
tography. In a polar solvent such as acetonitrile and
water, compound 4 exists as a mixture of enol- and
keto-form in a 1:2 ratio in acetonitrile, which indicates
that sulfur-substitution actually stabilizes its enol form.
In the reaction of ScHICDH, compound 4 appeared to
be a strong uncompetitive inhibitor (Ki = 12 lM) at dif-
4
HIC
—
17b
nd: these values were not determined under these conditions because of
its low substrate activity.
a Uncompetitive inhibitor.
b Data from Ref. 3.
faster phase was found to be concentration dependent,
while the rate of second slower phase was apparently
independent of the concentration of thia-analogue.
The formation of NADH in all reactions finally reached
almost plateau at the same level. These results clearly
indicated that the oxidation of thia-analogue by
NAD+ proceeded in a concentration dependent manner
Figure 3. Time-course of the NADH formation in ScHICDH reaction; thia-analogue 12.5–120 lM, NAD 5.0 mM, ScHICDH 3.0 lM, 10 °C.
Figure 4. Lineweaver–Burk plot of the effect of compound 4 on ScHICDH reaction.

T. Yamamoto, T. Eguchi / Bioorg. Med. Chem. 16 (2008) 3372–3376
3375
ferent concentrations of HIC and a fixed concentration
of NAD+ as shown in Figure 4. The same uncompetitive
inhibition was observed at different concentrations of
NAD+ and a fixed concentration of HIC (data not
shown). It was reported that 2-ketoadipate, the product
of the original reaction, was shown to be an uncompet-
itive inhibitor for ScHICDH (Ki = 5.8 mM),7 so that
2-ketoadipate was suggested to be released from
ScHICDH before NADH. Therefore, since compound
4 was found to be an uncompetitive inhibitor as
2-ketoadipate, the binding of the intermediary enolate
ii to ScHICDH could participate in inhibition by thia-
analogue. In the case that the binding of the enolate
intermediate to ScHICDH is a predominant factor for
inhibition of thia-analogue, the inhibition constant of
thia-analogue is likely to be of a similar order of magni-
tude to that of compound 4, however it was not the case.
It seems less likely that the decarboxylation step is
slowed in this case. With regard to another possible
inhibition mechanism of thia-analogue, if the enol iv in
Figure 2 is spontaneously formed from the intermediate
i by the effect of the introduced heteroatom, this
compound cannot be decarboxylated, thereby residing
in the active site. Therefore, the possibility of spontane-
ous formation of the enol iv cannot be ruled out for the
inhibition mechanism at the moment. In a preliminary
test, thia-analogue did not affect growth of S. cerevisiae
probably due to its low permeability into cells. Further
derivatization and mechanistic analysis of thia-analogue
is underway in our laboratory.
NaOH (pH 7.8), 0.2 mM KCl, 5.0 mM MgCl2, and
5.0 mM NAD+. A reaction mixture including ScHI-
CDH (0.1 lg) and HIC (5–50 lM) or the alternative
analogues was pre-incubated for ca. 3 min and the reac-
tion was started by addition of NAD+ to the reaction
mixture. The formation of NADH was measured at
340 nm for 5 s. Data were graphically analyzed by
Lineweaver–Burk double reciprocal plots, and the ki-
netic parameters were estimated by Hanes plots or Dix-
on plots. Pre-steady state kinetic measurements were
performed at 10 °C by mixing A and B components;
component A, ScHICDH (3.0 lM) and thia-analogue
3 (12.5–120 lM), component B, 5.0 mM NAD+.
3.3. Synthesis
3.3.1. Aza-analogue 1. To a solution of epoxyfumaric
acid (20.1 mg, 0.152 mmol) in distilled water (1.0 ml)
was added 0.2 ml of 3.0 M aqueous sodium hydroxide,
followed by glycine (11.4 mg, 0.152 mmol). The mixture
was stirred for 10 h at 80 °C. The solution was evapo-
rated and the residue was chromatographed by gel-fil-
tration (Sephadex G-10, water). Evaporation of the
solvent gave sodium salt of 1 (31 mg, 82%): IR (KBr):
3469, 1616, 1396 cmÀ1 1H NMR (D2O) d 3.08 (d,
;
J = 15.8 Hz, 1H), 3.15 (d, J = 15.8 Hz, 1H), 3.33 (d,
J = 3.6 Hz, 1H), 4.05 (d, J = 3.6 Hz, 1H); 13C NMR
(D2O) d 178.2, 178.1, 176.9, 73.1, 65.6, 50.7. Anal. Calcd
for C6H7NNa2O7Æ3H2O: C, 23.62; H, 4.29; N, 4.59.
Found: C, 23.83; H, 4.20; N, 4.48.
In conclusion, a series of aza-, oxa-, and thia-analogues
of homoisocitrate was designed and synthesized as a po-
tential inhibitor for HICDH. The synthesized thia-ana-
logue showed strong competitive inhibitory activity
toward homoisocitrate dehydrogenase. This is the first
successful example of design and synthesis of a highly
potent inhibitor for HICDH. The development of a
highly potent inhibitor such as thia-homoisocitrate
may provide a way to understand more detailed enzyme
reaction mechanism of this class of enzymes.
3.3.2. Oxa-analogue 2. To a solution of diethyl meso-tar-
trate (1.30 g, 6.30 mmol) and silver (I) oxide (4.38 g,
18.9 mmol) in dry CH3CN (2.6 ml) was added methyl
iodoacetate (1.39 g, 6.94 mmol). The mixture was stirred
for 20 h at 40 °C. The mixture was filtered and the filtrate
was concentrated. The residue was chromatographed
over silica gel (hexane/ethyl acetate = 3:7) to afford triest-
1
er of 1 (1.60 g, 84%): IR (NaCl): 3094, 1754 cmÀ1; H
NMR (CDCl3) d 1.24 (t, J = 7.2 Hz, 3H), 1.26 (t,
J = 7.2 Hz, 3H), 3.72 (s, 3H), 3.84 (d, J = 8.0 Hz, 1H),
4.13–4.16 (m, 5H), 4.36 (d, J = 2.4 Hz, 1H), 4.43 (d,
J = 17.2 Hz, 1H), 4.64 (dd, J = 2.4, 8.0 Hz, 1H); 13C
NMR (CDCl3) d 170.7, 170.3, 168.0, 81.1, 72.1, 67.7,
61.9, 61.5, 52.0, 14.0, 14.0. Anal. Calcd for C11H18O8:
C, 47.48; H, 6.52. Found: C, 47.73; H, 6.65.
3. Experimental
3.1. General
1H and 13C NMR spectra were recorded on a JEOL LA-
400 spectrometer. IR spectra were recorded on a Horiba
FT-710 Fourier-transform infrared spectrometer. Ele-
mental analyses were performed with a Perkin-Elmer
2400 apparatus. Column chromatography was carried
out with Merck Kieselgel 60 (70–230 mesh, Merck). En-
zyme reactions were monitored by measuring the
NADH absorption at 340 nm on a Shimadzu UV-
160 A UV–vis recording spectrometer. Pre-steady state
kinetics was monitored on a Photal stopped-flow appa-
ratus (Otsuka Electronics).
To a solution of the obtained triester (600 mg, 2.16 mmol)
in THF (5.0 ml) and water (7.0 ml) was added aqueous so-
dium hydroxide (3.0 M, 2.5 ml). The mixture was stirred
for 3 h. The solution was evaporated to remove THF,
and the residue was chromatographed by gel-filtration
(Sephadex G-10, water). Evaporation of the solvent gave
sodium salt of 1 (401 mg, 86%): IR (KBr): 3421, 1602,
1417 cmÀ1; 1H NMR (D2O) d 3.73 (d, J = 15.2 Hz, 1H),
3.91 (d, J = 2.8 Hz, 1H), 3.93 (d, J = 15.2 Hz, 1H), 4.18
(d, J = 2.8 Hz, 1H); 13C NMR (D2O) d 177.8, 177.2,
176.0, 83.4, 73.5, 69.1. Anal. Calcd for C6H5Na3O8Æ3H2O:
C, 21.96; H, 3.38. Found: C, 22.19; H, 3.44.
3.2. Inhibitory activity
3.3.3. Thia-analogue 3. To a solution of epoxyfumaric
acid (78.2 mg, 0.592 mmol) in distilled water (2.0 ml)
was added 3.0 M of aqueous sodium hydroxide until
Kinetic measurements were performed at 36 °C in an as-
say mixture (total 700 ll) containing 50 mM Hepes–

3376
T. Yamamoto, T. Eguchi / Bioorg. Med. Chem. 16 (2008) 3372–3376
pH 12. Then, thioglycolic acid (54.5 mg, 0.592 mmol)
was added. The mixture was stirred for 3 h at 80 °C.
The solution was evaporated and the residue was chro-
matographed by gel-filtration (Sephadex G-10, 5 mM
NaOH aq). Evaporation of the solvent gave sodium salt
3.25 (s, 2H); 13C NMR (CD3CN) d 188.7, 171.1,
171.0, 163.6, 161.3, 138.7, 112.8, 37.6, 35.2, 33.8. Anal.
Calcd for C5H6O5S: C, 33.71; H, 3.39; S, 18.00. Found:
C, 33.68; H, 3.68; S, 18.28.
of 3 (140 mg, 82%): IR (KBr): 3442, 1596, 1396 cmÀ1
;
1H NMR (D2O) d 3.22 (d, J = 14.8 Hz, 1H), 3.26 (d,
J = 14.8 Hz, 1H), 3.43 (d, J = 6.0 Hz, 1H), 4.00 (d,
J = 6.0 Hz, 1H); 13C NMR (D2O) d 178.6, 177.4,
177.2, 73.9, 53.6, 37.1. Anal. Calcd for C6H5Na3O7SÆ3-
H2O: C, 20.94; H, 3.22; S, 9.32. Found: C, 20.93; H,
3.42; S, 9.06.
References and notes
1. Zabriske, M.; Jackson, M. Nat. Prod. Rep. 2000, 17, 85.
2. Palmer, D. J.; Balogh, H.; Ma, G.; Zhou, X.; Marko, S. G.;
Kaminskyj, S. W. Pharmazie 2004, 59, 2.
3. Yamamoto, T.; Miyazaki, K.; Eguchi, T. Bioorg. Med.
Chem. 2007, 15, 1346.
4. Bernasconi, C. F.; Kittredge, K. W. J. Org. Chem. 1998, 63,
1944–1953.
3.3.4. Enzyme reaction product 4. To a solution of 3-
bromopyruvic acid (600 mg, 3.59 mmol) in acetone
(7.0 ml) was added thioglycolic acid (191.9 ml,
2.76 mmol) at À78 °C. After the mixture was warmed
to 0 °C, triethylamine (2.33 ml, 16.6 mmol) was added.
The mixture was evaporated and the residue was chro-
matographed over ion-exchanged resin (DEAE Sepha-
dex A-20, 0–2.0 M formic acid) to give 4 (460 mg,
5. Bordwell, F. G.; Bares, J. E.; Bartmess, J. E.; Drucker, G.
F.; Gerhold, J.; McCollum, G. J.; Van Der Puy, M.; Vanier,
N. R.; Mattews, W. S. J. Org. Chem. 1977, 42, 326.
6. Plaut, W. E. G.; Aogaichi, T.; Gabriel, L. J. Arch. Biochem.
Biophys. 1986, 245, 114.
7. Lin, Y.; Alguindigue, S. S.; Volkman, J.; Nicholas, K. M.;
West, H. A.; Cook, F. P. Biochemistry 2007, 46, 890.
8. Marco, C.; Rinaldi, A.; Piccaluga, G.; Fadda, M. B. Mol.
Cell. Biochem. 1974, 3, 3.
86%): IR (KBr): 3093, 2979, 1691 cmÀ1
;
1H NMR
(CD3CN) d 6.42 (s, 1H), 3.78 (s, 2H), 3.51 (s, 2H),