Investigation of mechanism of nitrogen transfer in glucosamine 6- phosphate synthase with the use of transition state analogs

Article abstract of DOI:10.1006/bioo.1997.1077

Several structural analogs of putative tetrahedral intermediates of the reaction catalyzed by the glutamine amide transfer domain of Candida albicans glucosamine 6-phosphate synthase have been designed and synthesized. Esters and amides of γ-phosphonic and γ-sulfonic analogs of glutamine and glutamic acid were tested as potential inhibitors of the enzyme. N-substituted amides 9 and 15 were found to be the strongest inhibitors in the series. Structure- activity relationship studies led to conclusions supporting the possibility of a direct nucleophilic attack of the glutamine amide nitrogen on an electrophilic site of the enzyme-bound fructose 6-phosphate as the most likely mechanism of nitrogen transfer in glucosamine 6-phosphate synthase.

Full text of DOI:10.1006/bioo.1997.1077

BIOORGANIC CHEMISTRY 25, 283–296 (1997)
ARTICLE NO. BH971077
Investigation of Mechanism of Nitrogen Transfer in
Glucosamine 6-Phosphate Synthase with the Use of
Transition State Analogs
Sławomir Milewski,*,1 Maria Hoffmann,† Ryszard Andruszkiewicz,* and
Edward Borowski*
*Department of Pharmaceutical Technology and Biochemistry and
†Department of Organic Chemistry, Technical University of Gdan´sk, 11/12 Narutowicza Str.,
80-952 Gdan´sk, Poland
Received September 18, 1997
Several structural analogs of putative tetrahedral intermediates of the reaction catalyzed
by the glutamine amide transfer domain of Candida albicans glucosamine 6-phosphate synthase
have been designed and synthesized. Esters and amides of Ͳ-phosphonic and Ͳ-sulfonic analogs
of glutamine and glutamic acid were tested as potential inhibitors of the enzyme. N-substituted
amides 9 and 15 were found to be the strongest inhibitors in the series. Structure–activity
relationship studies led to conclusions supporting the possibility of a direct nucleophilic attack
of the glutamine amide nitrogen on an electrophilic site of the enzyme-bound fructose 6-
phosphate as the most likely mechanism of nitrogen transfer in glucosamine 6-phosphate
synthase.
1997 Academic Press
INTRODUCTION
Glucosamine 6-phosphate synthase, EC 2.6.1.16, is a member of the family of
enzymes transferring an amino group from the Ͳ-amide function of -glutamine to
L
different acceptor molecules. The amidotransferase family comprises two subfamil-
ies known as G-type and F-type, depending on location of the catalytic cysteinyl
residue (1). GlcN-6-P2 synthase belongs to the latter subfamily, together with three
other proteins, namely phosphoribosylpyrophosphate (PRPP) amidotransferase,
glutamate synthase, and asparagine synthetase (1), however differs from other
members in its apparent inability to use exogenous ammonia as an alternative
nitrogen donor (2) and catalyzes almost exclusively an irreversible reaction:
L-glutamine ϩ
D-fructose 6-phosphate Ǟ
D
-glucosamine 6-phosphate ϩ -glutamate.
L
1 To whom correspondence should be addressed. Fax: ϩ48 58 47 26 94. E-mail: milewski@altis.chem.
pg.gda.pl.
2 Abbreviations used: GlcN-6-P,
D-glucosamine 6-phosphate; Fru-6-P, D-fructose 6-phosphate;
Ͳ-Glu(P), (3-amino-3-carboxypropyl)-phosphonic acid; GLUPA, Ͳ-glutamyl-p-nitroanilide; A2pr, 2,3-
diaminopropanoic acid; THF, tetrahydrofurane; DMF, dimethylformamide; DMSO, dimethylsulfoxide;
MNDO, modified neglect of diatomic overlap.
283
0045-2068/97 $25.00
Copyright 1997 by Academic Press
All rights of reproduction in any form reserved.

284
MILEWSKI ET AL.
In the absence of Fru-6-P, the enzyme is capable of catalyzing the glutamine hydroly-
sis, but the velocity of this reaction is several orders of magnitude lower than that
of GlcN-6-P formation (3).
The molecular mechanism of the complex reaction catalyzed by GlcN-6-P syn-
thase has been extensively investigated and special attention has been paid to the
mechanism of nitrogen transfer (4–8). There is little doubt that in G amidotransfer-
ases the amide nitrogen labilization is mediated by a catalytic triad Cys–His–Asp
in a manner similar to that of cysteine proteases (9). The presence of such a system,
previously postulated also for F amidotransferases (10), was ruled out and Cys-1
was identified as the only catalytic residue participating in nitrogen transfer in PRPP
amidotransferase (11) and GlcN-6-P synthase (12).
Badet-Denisot et al. in a recent paper proposed three possible mechanisms of
nitrogen transfer in GlcN-6-P synthase (7). The authors finally excluded a possibility
of nitrogen labilization as a result of cyclization of glutamine into pyroglutamic
acid, but two other possibilities can be still considered (Fig.1):
—a ‘‘classical’’ cleavage of the glutamine amide bond upon the action of Cys-1
nucleophile, followed by hydrolysis of the acyl-enzyme intermediate (Mechanism
1), based on the previous proposition of Chmara et al. (4). A transient release of
ammonia which is subsequently transferred to Fru-6-P in an unknown way is likely
to occur in this mechanism.
—a direct nucleophilic attack of the glutamine amide nitrogen on an electrophilic
site of the enzyme-bound acceptor (Mechanism 2) in a manner similar to that
proposed previously for asparagine synthetase (13). In this mechanism Cys-1 partici-
pates only in releasing of glutamate.
Various phosphonic, sulfonic and boronic derivatives have been previously pro-
posed as analogs of tetrahedral intermediates formed transiently during different
enzymatic reactions involving hydrolysis of amide or ester bonds, especially those
catalyzed by proteases and esterases. Since presence of tetrahedral intermediates
is also postulated in both possible mechanisms of nitrogen transfer carried out by
GlcN-6-P synthase, we synthesized several Ͳ-phosphono- and Ͳ-sulfonic analogs of
glutamine and glutamic acid—potential mimics of putative tetrahedral intermedi-
ates I, II, and III. The synthesis and inhibitory properties of these compounds
toward Candida albicans GlcN-6-P synthase are described in the present paper.
MATERIALS AND METHODS
Reagents
L
-Methionine sulfoxide 10, L-methionine sulfone 11, and -SR-methionine sulfox-
L
imine 13 were purchased from Sigma. Other reagents were of the finest grade
commercially available. Diazomethane was generated from N-nitroso-N-methylurea
and used immediately for esterification.
Synthetic Chemistry
General. Thin-layer chromatography (TLC) was performed on Kieselgel 60 plates
(Merck) using following solvent systems: solvent A (iPrOH:NH4OH:H20, 8:1:1),
solvent B (nBuOH:AcOEt:AcOH:H2O, 1:1:1:1), solvent C (CHCl3:MeOH, 9:1),

NITROGEN TRANSFER IN GlcN-6-P SYNTHASE
285

286
MILEWSKI ET AL.
solvent D (nBuOH:Pyr:AcOH:H2O, 6:2:3:3), or solvent E (nPrOH:H2O, 1:1).
1
Amino acids were visualized with ninhydrin. Melting points are uncorrected. H-
NMR and 31P-NMR spectra were recorded on a Gemini Varian 500 MHz spec-
trometer.
DL-[3-Amino-3-carboxypropyl]phosphonic acid 1 was synthesised according to
the previously described procedure (14). Acylation of 1 with benzyloxycarbonyl
chloride under conditions described elsewhere (15) afforded DL-[3-[(benzyloxycar-
bonyl)amino]-3-carboxypropyl]phosphonic acid 1a. Analytical results for com-
pounds 1 and 1a (TLC, mp, and NMR) were satisfactory and consistent with
literature data.
Synthesis of dimethyl-DL-[3-[(benzyloxycarbonyl)amino]-3-methoxycarbonylpro-
pyl]phosphonate 2a. An ice-cooled solution of 1 mmol of 1a in 5 ml of methanol
was treated portionwise with an etheral solution of diazomethane, until the solution
became slightly yellow. Volatile components of the mixture were removed under
reduced pressure. The residue was dissolved in 10 ml of ethyl acetate and washed
subsequently with 5% aqueous HCl solution and water. The organic layer was dried
over anhydrous MgSO4 . Removal of the drying agent and the solvent afforded
1
ester 2a as an oil. Yield, 88%; TLC (solvent C) Rf ϭ 0.5; H-NMR (CDCl3) ͳ:
1.70–2.30 (m, 4H), 3.72, 3.73 (two d, 6H, J ϭ 11 Hz), 3.75 (s, 3H), 4.30–4.50 (m,
1H), 5.12 (s, 2H), 5.55 (d, 1H), 7.35 (s, 5H).
Synthesis of dimethyl-DL-[3-amino-3-methoxycarbonylpropyl]phosphonate 2. A
0.25-mmole of 2a sample was dissolved in 5 ml of methanol and 15 mg of 10%
Pd/C was added. The mixture was then stirred under H2 gas for 2 h at room
temperature. After filtration and removal of the solvent under reduced pressure,
a viscous oily product 2 was obtained. Yield, 90%; TLC (solvent A) Rf ϭ 0.57; 1H-
NMR (CF3COOD) ͳ: 2.15–2.70 (m, 4H), 3.92–4.02 (m, 9H), 4.53 (t, 1H).
Synthesis of dimethyl-DL-[3-[(benzyloxycarbonyl)amino]-3-carboxypropyl]phos-
phonate 3a. Selective saponification of carboxyl methyl ester 2a was accomplished
by the slow addition of an equimolar amount of 1 M NaOH to the ice-cooled
ethanolic solution of 2a. The mixture was then left for 1 h at room temperature.
Subsequently, the solvent was removed under reduced pressure; the remaining
aqueous phase was acidified with 6 M HCl at 0ЊC and extracted several times with
ethyl acetate. Combined organic extracts were dried over MgSO4 . Removal of the
drying agent and the solvent afforded a crystalline product 3a. Yield, 80%; mp
120–122ЊC; TLC (solvent A) Rf ϭ 0.5; 1H-NMR (CDCl3) ͳ 1.75–2.30 (m, 4H), 3.73,
3.76 (two d, 6H, J ϭ 11 Hz), 4.30–4.50 (m, 1H), 5.11 (s, 2H), 5.71 (d, 1H), 7.35 (s, 5H).
Synthesis of dimethyl-DL-[3-amino-3-carboxypropyl]phosphonate 3. Hydrogeno-
lysis of 3a under conditions described above for the conversion of 2a into 2 afforded
viscous oily product 3. Yield, 99%; TLC (solvent A) Rf ϭ 0.20, (solvent B) Rf ϭ
1
0.34; H-NMR (CF3COOD) ͳ: 1.80–2.50 (m, 4H), 3.66 (d, 6H, J ϭ 11 Hz), 4.31
(m, 1H)
Synthesis of methyl-DL-[3-[(benzyloxycarbonyl)amino]-3-methoxycarbonylpro-
pyl]phosphonic acid 4a. Dimethyl-DL-[3-[(benzyloxycarbonyl]-3-methoxycarbon-
ylpropyl]phosphonate 2a (0.5 mmol) was dissolved in t-butylamine (8 ml) and heated
at refluxed temperature for 4 days. The reaction mixture was concentrated in vacuo
to provide the product as a white salt (208 mg). The salt was dissolved in methanol

NITROGEN TRANSFER IN GlcN-6-P SYNTHASE
287
and acidified with cation exchange resin (Dowex 50WX8, Hϩ form), 200–400 mesh.
The Dowex resin was removed by filtration and the filtrate was concentrated in
1
vacuo to afford oily product 4a. Yield, 90%; TLC (solvent A) Rf ϭ 0.5; H-NMR
(CDCl3) ͳ: 1.65–2.35 (m, 4H,), 3.70 (d, 3H, J ϭ 11 Hz), 3.75 (s, 3H ), 4.35–4.50
(m, 1H ), 5.11 (s, 2H ), 5.40–5.70 (bs, 1H), 7.35 (s, 5H).
Synthesis of methyl-DL-[3-amino-3-methoxycarbonylpropyl]phosphonic acid 4.
Hydrogenolysis of 4a under conditions described above for conversion of 2a into
1
2 afforded 4 as a viscous oily product. Yield, 89%; TLC (solvent A) Rf ϭ 0.15; H-
NMR (CF3COOD) ͳ: 1.80–2.40 (m, 4H), 3.63 (d, 3H, J ϭ 11 Hz), 3.72 (s, 3H), 4.23
(m, 1H).
Synthesis of methyl-DL-[3-[(benzyloxycarbonyl)amino]-3-carboxypropyl]phos-
phonic acid 5a. t-Butylamine salt of 4a (0.5 mmol) was dissolved in 1 ml 1 M NaOH
and the solution heated for 30 min. Afterward the mixture was acidified with cation
exchange resin (Dowex 50WX8, Hϩ form). The resin was removed by filtration and
the filtrate was concentrated in vacuo. The residue was crystallized from ethyl
1
acetate/ether. Yield, 90%; TLC (solvent A) Rf ϭ 0.20; mp 109–110ЊC; H-NMR
(DMSO) ͳ: 1.50–2.00 (m, 4H), 3.53 (d, 3H, JPH ϭ 11 Hz), 3.95–4.12 (m, 1H), 5.04
(s, 2H), 7.36 (s, 5H), 7.70 (d, 1H, J ϭ 8 Hz).
Synthesis of methyl-DL-[3-amino-3-carboxypropyl]phosphonic acid 5. Hydro-
genolysis of 5a under conditions described above for conversion of 2a into 2 afforded
1
5 as a viscous oily product. Yield, 99%; TLC (solvent B) Rf ϭ 0.20; H-NMR
(DMSO ϩ CF3COOD) ͳ: 1.50–2.20 (m, 4H), 3.58 (d, 3H, JPH ϭ 11 Hz), 3.90–4.30
(m, 1H).
General procedure for the synthesis of methyl DL-[3-[(benzyloxycarbonyl)amino]-
3-methoxycarbonylpropyl]-SR-phosphonamidates: 6a, 7a, 8a, 9a. Oxalyl chloride
(0.066 g, 0.52 mmol) was added dropwise to a solution of 4a (115 mg, 0.334 mmol)
and DMF (1.25 Ȑl, 0.017 mmol) dissolved in CH2Cl2 at 0ЊC. The solution was stirred
at 0ЊC for 20 min and then warmed to room temperature and stirred for 1.5 h.
The reaction mixture was concentrated, dissolved in toluene (2 ml), and then
reconcentrated to remove the volatile reagents. This left the phosphonochloridate
as a yellow oil which was used immediately in a reaction with ammonia or amine.
The amine reagent, dimethylamine, ammonia (both as a 2 M solution in THF),
ethyl amine, or diethylamine (4 mmol), cooled to 0ЊC, was added to the phospho-
nochloridate (1 mmol) dissolved in 5 ml of CH2Cl2 . The reaction mixture was
allowed to warm to room temperature and stirred overnight. The resulting solution
was concentrated in vacuo. The oily residue was dissolved in ethyl acetate and
washed successively with 5% NaHCO3 , 5% KHSO4 , and water, dried with MgSO4 ,
filtered, and concentrated to afford the product.
General procedure for the synthesis of methyl DL-[3-amino-3-methoxycarbonylpro-
pyl]-SR-phosphonamidates: 6, 7, 8, 9. Hydrogenolysis of 6a, 7a, 8a, 9a under condi-
tions described above for conversion of 2a into 2, afforded 6, 7, 8, or 9 as viscous
oily products.
6a: DL-Methyl [3-[(benzyloxycarbonyl)amino]-3-methoxycarbonylpropyl-SR-
1
phosphonamidate. Yield, 55%; TLC (solvent C) Rf ϭ 0.35; H-NMR (CDCl3) ͳ:
1.70–2.60 (m, 6H), 3.71, 3.72 (two d, 3H, JPH ϭ 11 Hz), 3.79 (s, 3H), 4.40–4.60 (m,
1H), 5.14 (s, 2H), 5.65–5.80 (br, 1H), 7.38 (s, 5H).

288
MILEWSKI ET AL.
6: DL-Methyl [3-amino-3-methoxycarbonylpropyl]-SR-phosphonamidate. Yield
quant.; TLC (solvent A) Rf ϭ 0.45, (solvent B), Rf ϭ 0.32; 31P-NMR (CD3OD) ͳ:
1
39.56, 39.54; H-NMR (CD3OD) ͳ: 1.70–2.20 (m, 4H), 3.67 (d, 3H, J ϭ 11 Hz),
3.78 (s, 3H), 3.60–3.80 (m, 1H).
7a: DL-Methyl N,N-dimethyl-[3-[(benzyloxycarbonyl)amino]-3-methoxycar-
bonylpropyl]-SR-phosphonamidate. Yield 70.5%; TLC (solvent C) Rf ϭ 0.55;
1
31P-NMR (CDCl3) ͳ: 37.46, 37.43; H-NMR (CDCl3) ͳ: 1.60–2.10 (m, 4H), 2.67
(d, 6H, JPH ϭ 8,9 Hz), 3.57, 3.58 (two d, 3H, JPH ϭ 11 Hz), 3.75 (s, 3H), 4.30–4.50
(m, 1H), 5.12 (s, 2H), 5.63–5.72 (bd, 1H), 7.35 (s, 5H).
7: DL-Methyl N,N-dimethyl-[3-amino-3-methoxycarbonylpropyl]-SR-phospho-
namidate. Yield, 95%; TLC (solvent A) Rf ϭ 0.55, (solvent B) Rf ϭ 0.38; 31P-NMR
(CD3OD) ͳ: 40.09, 40.02; 1H-NMR (CD3OD) ͳ: 1.70–2.10 (m, 4H), 2.69, 2.70 (two
d, 6H, JPH ϭ 9 Hz), 3.59 (d, 3H, JPHϭ 11 Hz), 3.74 (s, 3H), 3.50–3.80 (m, 1H).
8a: DL-Methyl N-ethyl-[3-[(benzyloxycarbonyl)amino]-3-methoxycarbonylpro-
pyl]-SR-phosphonamidate. Yield, 70%; TLC (solvent C), Rf ϭ 0.46; 1H-NMR
(CDCl3) ͳ: 1.13, 1.25 (two t, 3H, J ϭ 7 Hz), 1.65–2.30 (m, 4H), 2.90 (dq, 2H,
JPH ϭ 11 Hz, JHH ϭ 7 Hz), 3.62, 3.64 (two d, 3H, JPH ϭ 11 Hz), 3.75 (s, 3H),
4.35–4.50 (m, 1H), 5.11 (s, 2H), 5.70–5.85 (br, 1H), 7.35 (s, 5H).
8: DL-Methyl N-ethyl-[3-amino-3-methoxycarbonylpropyl]-SR-phosphonami-
1
date. Yield quant.; TLC (solvent A), Rf ϭ 0.55, (solvent B) Rf ϭ 0.42; H-NMR
(CD3OD) ͳ: 1.14, 1.30 (two t, 3H, J ϭ 7Hz), 1.65–2.20 (m, 4H), 2.95, 4.15 (two dq,
2H), 3.62 (d, 3H, JPH ϭ 11 Hz), 3.74 (s, 3H), 3.50–3.75 (m, 1H).
9a: DL-Methyl N,N-diethyl-[3-[(benzyloxycarbonyl)amino]-3-methoxycarbon-
ylpropyl]-SR-phosphonamidate. Yield, 51%; TLC (solvent C), Rf ϭ 0.6; 31P-NMR
1
(CDCl3) ͳ: 36,67; H-NMR (CDCl3) ͳ: 1.09, 1.30 (two t, 6H, J ϭ 7 Hz), 1.60–2.20
(m, 4H), 3.06, 4.10 (two dq, 4H, JPH ϭ 11 Hz, JHH ϭ 7 Hz), 3.55, 3.56 (two d, 3H,
J ϭ 11 Hz), 3.75 (s, 3H), 4.30–4.50 (m, 1H), 5.12 (s, 2H), 5.75 (br d), 7.35
(s, 5H).
9: DL-Methyl N,N-diethyl-[3-amino-3-methoxycarbonylpropyl]-SR-phospho-
namidate. Yield, 95%; TLC (solvent A), Rf ϭ 0.64, (solvent B), Rf ϭ 0.48; 31P-NMR
1
(CD3OD) ͳ: 39.26, 39.15; H-NMR (CD3OD) ͳ: 1.13, 1.33 (two t, 6H), 1.65–2.10
(m, 4H), 3.05, 4.10 (two dq, 4H, JHH ϭ 6 Hz, JPH ϭ 12 Hz), 3.55 (d, 3H, J ϭ 11
Hz), 3.73 (s, 3H), 3.50–3.80 (m, 1H).
Synthesis of N3-methanesulfonyl-
L
-2,3-diaminopropanoic acid trifluroacetate 12.
Methyl N2-tert-butoxycarbonyl-
L-2,3-diaminopropanoate (16) (1 mmol) was dis-
solved in 5 ml of THF and the solution was cooled to 0ЊC. Triethylamine (0.14 ml)
was added, followed by methanesulfonyl chloride (0.07 ml). The reaction mixture
was stirred for 6 h at room temperature and then diluted with 100 ml of ethyl
acetate. The organic layer was washed with 10% KHSO4, brine, dried over MgSO4 ,
filtered, and evaporated to give 0.27 g (88%) of oily methyl N2-tert-butoxycarbonyl-
N3-methanesulfonyl-
-2,3-diaminopropanoate. Saponification of the methyl ester
L
with 1 M NaOH (0.9 ml, 2 h) and removal of the t-butoxycarbonyl protecting group
with cold trifluoroacetic acid (5 ml), afforded an oily product which on trituration
with ethyl ether gave N3-methanesulfonyl-
-2,3-diaminopropanoic acid trifluroa-
L
cetate as an amorphous powder. Yield, 40%; TLC (solvent D), Rf ϭ 0.52; 1H-NMR
(D2O) ͳ: 3.2 (s, 3H), 3.3–3.5 (m, 2H), 4.2 (m, 1 H).

NITROGEN TRANSFER IN GlcN-6-P SYNTHASE
289
Synthesis of DL-3-amino-3-carboxypropanesulfonamide 14. DL-Homocysteic acid
(5 mmol) and thionyl chloride (5 ml) were heated together under reflux for 3 h.
An excess of thionyl chloride was evaporated and the residue containing the crude
homocysteic acid chloride was stirred with a cold 25% aqueous ammonia (25 ml)
for 6 h. The reaction mixture was concentrated under reduced pressure and the
crude product was purified on a Dowex 50WX8 ion exchange resin (Hϩ form). The
column was washed with water and then with 0.5 M ammonia solution. Fractions
containing the title compound were collected and evaporated to dryness to give
the pure product. Yield, 50%; mp 245–250ЊC [lit. 247ЊC (17)]; TLC (solvent E),
1
Rf ϭ 0.5; H-NMR (D2O) ͳ: 2.0–2.4 (m, 2H), 2.9–3.2 (m, 2H), 4.2 (t, 1H).
Synthesis of DL-3-amino-3-carboxy-N,N-diethyl-propanesulfonamide 15. The
crude homocysteic acid chloride obtained as described above was treated with a
cold 50% aqueous N,N-diethyl amine. Reaction conditions and product isolation
was performed analogously as for the compound 14. Final product was obtained
1
as an amorphous powder. Yield, 35%; TLC (solvent D), Rf ϭ 0.6; H-NMR (D2O)
ͳ: 1.1 (m, 6H), 2.2 (m, 2H), 2.9 (m, 2H), 3.3 (m, 4H), 4.2 (m, 1H).
Enzymology
Candida albicans GlcN-6-P synthase overproduced by Saccharomyces cerevisiae
YRSC-65 was purified to apparent homogeneity as described previously (18). Two
assays of GlcN-6-P synthase activity were used: (a) determination of GlcN-6-P
formation by the modified Elson–Morgan procedure (19) and (b) continuous quanti-
tation of glutamate in a coupled assay with glutamate dehydrogenase in the presence
of acetylpyridine adenine dinucleotide (2). In both assays, reactions were carried out
in 0.1 M Tris–HCl buffer, pH 6.85, containing 1 mM EDTA and 1 mM dithiothreitol.
Reaction mixtures contained 0.1–0.2 ȐM GlcN-6-P synthase and 7.5 mM Fru-6-P.
Glutamine concentrations were either fixed (10 mM, for IC50 determinations) or
variable (0.5–5 mM, for Ki determinations).
Amidohydrolase activity of the enzyme was determined using 1 mM GLUPA as
a substrate, in the same buffer as shown above. Fru-6-P (7.5 mM) was either present
or omitted from incubation mixtures and GlcN-6-P synthase was present at 1–2
ȐM. Formation of p-nitroaniline was followed spectrophotometrically at 420 nm.
All the measurements were done in triplicate. Inhibitory constants were deter-
mined from the secondary plots of kapp versus inhibitor concentration, derived from
Lineweaver–Burk plots.
RESULTS AND DISCUSSION
Synthesis
Structures of compounds synthesized by us and commercially available methio-
nine derivatives tested as potential inhibitors of GlcN-6-P synthase are shown in
Fig. 2. N-benzyloxycarbonyl derivative of Ͳ-Glu(P) was converted into N-protected
methyl triester 2a using diazomethane. Hydrogenolysis of 2a yielded directly methyl
triester 2. Carboxyl methyl ester in 2a could be selectively saponified under alkaline

290
MILEWSKI ET AL.
FIG. 2. Structures of transition state analogs tested as potential inhibitors of GlcN-6-P synthase.
conditions to afford, after subsequent N-deprotection, the diester 3. Alternatively,
treatment of dimethyl phosphonate 2a with t-butylamine under reflux, followed by
acidification, resulted in selective hydrolysis to monomethyl phosphonate 4a. This
compound was either N-deprotected to give diester 4, saponified, and N-deprotected
to afford monoester 5, or used as a starting substrate for the synthesis of methyl
Ͳ-Glu-(P) amidates 6–9. For this purpose, 4a was first converted into phosphono-
chloridate upon treatment with oxalyl chloride. Aminolysis of this compound fol-
lowed by N-deprotection yielded required products 6–9.
All our attempts to prepare P-unprotected phosphonamidates were unsuccessful
due to an extreme instability of these compounds, resulting in a rapid decomposition
in aqueous solutions. This is in agreement with previous literature reports confirming
the sensitivity of the phosphonamidate bond to hydrolysis, especially under acidic
conditions (20, 21). However, other authors demonstrated good enzyme inhibitory
properties of monomethyl phosphonamidates designed as potential transition state
analog inhibitors of HIV-1 protease (22).
Selectively protected diamino acid, methyl N2-tert-butoxycarbonyl-
-2,3-diami-
L
nopropanoate, was acylated with methanesulfonyl chloride. Saponification of the
methyl ester and deprotection of the Ͱ-amino function with TFA afforded N3-
methanesulfonyl- -2,3-diaminopropanoic acid 12.
L
The sulfonamides 14 and 15 were prepared from commercially available (DL)-
homocysteic acid which gave corresponding acid dichloride (both carboxyl and
sulfonyl groups were converted) by heating with thionyl chloride. Immediate treat-
ment of the dichloride with an excess of either aqueous ammonia or N,N-diethyl-

NITROGEN TRANSFER IN GlcN-6-P SYNTHASE
291
TABLE 1
Inhibition of GlcN-6-P Synthase by Ͳ-Glu(P) and Its Ester Derivatives
GlcN-6-P synthetic activity
Amidohydrolase activity,
IC50 (mM)
Compound
IC50 (mM)
Ki (mM)
1
2
3
4
5
26
Ͼ50
Ͼ50
2.6
12 (6.0)
ND
ND
1.2 (0.6)
0.9 (0.45)
22
Ͼ50
Ͼ50
2.1
1.8
2.0
Note. ND, not determined. In parentheses Ki values corrected for one active enantiomer.
SD was Ϯ5%.
amine solution at elevated temperature led to the formation of required products
14 or 15 in moderate yields. Under these conditions sulfonyl chloride was readily
converted into sulfonamide, whereas carboxyl chloride was mainly hydrolyzed.
Crude preparations, slightly contaminated by disubstitution products, were finally
purified by ion-exchange chromatography.
Inhibition Studies
Compounds 1–15 were tested as potential inhibitors of GlcN-6-P-forming and
amidohydrolyzing activities of pure C. albicans GlcN-6-P synthase. Results obtained
for the former activity were essentially the same when either of two assay systems
was used: fixed point colorimetric determination of GlcN-6-P or continuous coupled
quantitation of glutamate, so that a more convenient and cheaper colorimetric
determination was routinely used. The enzyme exhibited KM ϭ 1.56 mM for -
L
1
1
1
glutamine, kcat ϭ 1150 minϪ , and kcat/KM ϭ 12.290 MϪ sϪ . Amidohydrolase activity
of GlcN-6-P synthase was measured with an artificial substrate, Ͳ-glutamyl p-nitro-
anilide. The kinetic parameters of this reaction were as follows: KM for GLUPA,
0.67 mM, kcat ϭ 1.52 minϪ , kcat/KM ϭ 38 MϪ sϪ1 (in the presence of 7.5 mM Fru-6-
P); kcat decreased to 1.09 minϪ1 when Fru-6-P was omitted from incubation mixtures.
Cumulative data including IC50 and Ki values are summarized in Tables 1–3.
Inhibitory constants were determined for those compounds which were found active
in initial screening. Since most of the tested compounds were racemic mixtures, Ki
values determined in kinetic experiments were corrected, taking into account a
1
1
highly probable action of enantiomer
is fully justified in light of the fact that
L
as an actual inhibitor. Such an assumption
-glutamine is an extremely poor substrate
D
for GlcN-6-P synthase (data not shown). Generally none of the studied compounds
showed as high inhibitory potency as could be expected for a proper transition
state analog. Inhibitory constants for the most active compounds 9 and 15 were
only less than an order of magnitude lower than the KM for
L-glutamine. All the
active compounds inhibited GlcN-6-P synthase competitively with respect to
L-Gln.
The representative graph obtained for compound 9 is shown in Fig. 3. On the other
hand the inhibition with respect to Fru-6-P was either uncompetitive or mixed.

292
MILEWSKI ET AL.
TABLE 2
Inhibition of GlcN-6-P Synthase by Amide Derivatives of Ͳ-Glu(P)
GlcN-6-P synthetic activity
Amidohydrolase activity,
Compound
IC50 (mM)
Ki (mM)
IC50 (mM)
6
7
8
9
25
11 (2.75)
4.4 (1.1)
1.3 (0.325)
0.94 (0.235)
30
7.8
4.4
3.6
7.1
6.1
3.2
Note. In parentheses Ki values corrected for one, possibly active diastereoisomer. SD
was Ϯ5%.
Compounds 1–5 (Table 1) could be considered as structural analogs of putative
tetrahedral intermediate III. In this hypothetical compound, one of the oxygen
atoms linked to phosphorus is negatively charged. Among the Ͳ-Glu(P) esters
tested, compounds 4 and 5 are the best mimics of this structure and the best
inhibitors of the enzyme. Presence of two potentially negatively charged oxygen
atoms (compound 1) or absence of any charge (compounds 2 and 3) markedly
lowers the inhibitory potency. Nevertheless, the inhibitory constant for 5 is much
higher than could be expected for a transition state analog. Undoubtedly, the best
mimic of III known so far is a covalent adduct formed upon the reaction of Ͳ-
glutamic semialdehyde with the active site cysteine, described previously by Badet-
Denisot et al. (7) and Bearne and Wolfenden (23). Inhibitory constant of this
8
compound against Escherichia coli GlcN-6-P synthase was as low as 3 ϫ 10Ϫ
M
(23). It is therefore not very surprising that our phosphonate and sulfonate analogs,
being unable to react with cysteine, show much lower affinity to the enzyme active
site. However, chemical reactivity is not an absolute requirement for an effective
transition state analog inhibitor mimicking a tetrahedral intermediate, as was pre-
TABLE 3
Inhibition of GlcN-6-P Synthase by Sulfonic Analogs of Glutamate and Glutamine
GlcN-6-P synthetic activity
Amidohydrolase activity,
Compound
IC50 (mM)
Ki (mM)
IC50 (mM)
10
11
12
13
14
15
Ͼ50
Ͼ50
Ͼ50
2.8
ND
ND
ND
1.5 (0.75)
2.8 (1.4)
0.6 (0.3)
Ͼ50
Ͼ50
Ͼ50
2.1
5.5
1.6
7.2
1.2
Note. ND, not determined. In parentheses Ki values corrected for one active enantiomer.
SD was Ϯ5%.

NITROGEN TRANSFER IN GlcN-6-P SYNTHASE
293
FIG. 3. Competitive inhibition of GlcN-6-P synthase by compound 9 in respect to
L
-glutamine. Inhibitor
concentrations were as follows: (᭿) none; (᭹) 0.5 mM; (᭡) 1.0 mM; (᭜) 4.5 mM. (Inset) A secondary
plot of kapp ϭ f([I]).
viously demonstrated by a high inhibitory activity of -methionine sulfone with
L
respect to glutamate synthase (24). Nevertheless, it seems sure that the presence
of III in a molecular mechanism of nitrogen transfer in GlcN-6-P synthase cannot
be questioned. It should be noted, however, that this intermediate is common for
both alternative mechanisms shown in Fig. 1.
Rather unexpected results were obtained for amide derivatives of Ͳ-Glu(P). Data
obtained for these compounds, summarized in Table 2, are corrected, taking into
account a highly probable possibility that only one of the four possible diastereo-
isomers could be an actual inhibitor. The closest structural analog of a putative
tetrahedral intermediate I, i.e., compound 6, appeared to be the poorest inhibitor
of GlcN-6-P synthase. Presence of a steric hindrance at the amide nitrogen enhanced
the inhibitory potency of Ͳ-Glu(P) amidates, resulting in a relatively low inhibitory
constant exhibited by the most substituted compound 9. The same inhibitory potency
order was observed when Ͳ-Glu(P) amidates 6–9 were tested as inhibitors of amido-
hydrolase activity of GlcN-6-P synthase.
This tendency was also confirmed for Ͳ-sulfonic analogs of glutamate and gluta-
mine (Table 3). Homocysteine sulfonamide 14 was previously reported to be a
strong inhibitor of glutamate synthase, an amidotransferase belonging to the same
subfamily as GlcN-6-P synthase. Masters and Meister demonstrated the high inhibi-
tory potency of this compound (Ki ϭ 3.6 ȐM) against the S. cerevisiae enzyme and

294
MILEWSKI ET AL.
FIG. 4. ‘‘Concerted attack’’ as the most probable mechanism of initiation of nitrogen transfer in
GlcN-6-P synthase.
suggested its close structural analogy to the putative tetrahedral transition state
(24). In our hands this compound was only a moderate inhibitor of C. albicans GlcN-
6-P synthase, poorer than methionine sulfoximine 13. N-substituted homocysteine
sulfonamide 15 was again the best inhibitor in the series, thus confirming the
tendency observed for Ͳ-Glu(P) amidates 6–9. No inhibitory effect was demon-
strated for methionine sulfoxide 10, methionine sulfone 11, and A2prSO2CH3 12,
potential analogs of transition state III, having no negatively charged oxygen atoms
linked to sulfur. Relatively good inhibitory activity of methionine sulfoximine 13
is therefore worth mentioning. Although in this compound no substituents are
present at the sulfur-linked nitrogen, hybridization of the latter atom is the same
as that of the respective nitrogen in the putative transition state II.
We believe that N-substituted Ͳ-Glu(P) amidates 7, 8, 9, and homocysteine N,N-
diethyl sulfonamide 15 show less structural analogy to the transition state I than
the respective compounds 6 and 10. On the other hand, N-substitution makes them
more resembling an amino acid part of the intermediate II. The probability of
existence of intermediate I was also previously questioned using semiempirical
quantum mechanical MNDO energy calculations (8). In our opinion, results pre-
sented above provide new data supporting the possibility of a direct nucleophilic
attack of the glutamine amide nitrogen according to mechanism 2 (Fig. 1) as a
more likely way of nitrogen transfer in GlcN-6-P synthase. However, the glutamine
amide nitrogen is undoubtedly a poor nucleophile. Its nucleophilicity may be suffi-
cient to attack aspartyl-AMP in asparagine synthetase (13), but not necessarily other,
less electrophilic acceptors. Therefore, we believe that the most likely pathway of
nitrogen transfer in GlcN-6-P synthase should be initiated with a concerted attack
according to mechanism 3 (Fig. 4). Approaching of the Cys-1 to glutamine amide
carbonyl would enhance the nucleophilicity of the amide nitrogen and facilitate its

NITROGEN TRANSFER IN GlcN-6-P SYNTHASE
295
simultaneous attack on the neighbouring electrophilic centre. It should be noted
˚
that a 20-A-long, solvent inaccessible channel, connecting the glutamine and the
phosphoribosylpyrophosphate binding sites, has been recently revealed in phospho-
ribosylpyrophosphate amidotranferase. This channel permits an interdomain trans-
fer of the NH3 intermediate (25). However, a very close proximity between the N-
terminal cysteine located at the glutamine binding site and the Fru-6-P domain has
been demonstrated in two independent studies on affinity labeling of the bacterial
GlcN-6-P synthase with N-iodoacetylglucosamine 6-phosphate and anhydro-1,2-
hexitol 6-phosphate (26, 27). Such a proximity should obviously facilitate the direct
nucleophilic attack of the glutamine amide nitrogen on an electrophilic site of the
enzyme-bound Fru-6-P.
Apparent inability of exogenous ammonia utilization, unique among amidotrans-
ferases (2), and marginal amidohydrolase activity of the enzyme in an absence of
Fru-6-P (3, this work) further support this assumption. At last, it should be noted
that the presence of the acceptor molecule, i.e., Fru-6-P, which is known to bind
to the enzyme active site before glutamine (5) and to bring about the conformational
change in the enzyme molecule (12), strongly enhances the velocity of -Gln hydroly-
L
sis (12) but has very little effect on hydrolysis rate of Ͳ-glutamyl p-nitroanilide (this
work). This is obvious that in the latter case, p-nitroaniline cannot be transferred
to the acceptor.
All these data taken together seem to favor mechanism 3. The recently revealed
molecular structure of the glutamine binding domain of E. coli GlcN-6-P synthase
does not provide any information undermining our hypothesis. Two amino acid
residues, Asn 98 and Gly 99, were unequivocally identified as those involved in
stabilization of oxyanionic tetrahedral intermediates (12). Respective residues, Asn
124 and Gly 125, are present in the C. albicans GlcN-6-P synthase, as can be
deduced from the gene sequence (28). These residues could be therefore involved
in stabilization of II and III.
Undoubtedly further studies are necessary to reach the conclusive evidence. One
of the possible approaches could be a design and synthesis of other structural
analogs of intermediate II that could be better mimics than bisubstrate analogs
designed and synthesized by Badet-Denisot et al. (7) and our compounds 9 and 15.
These studies are in progress in our laboratory.
ACKNOWLEDGMENTS
The authors appreciate the financial support of these studies by KBN Grant 3 T09A 133 08,
Foundation for Polish Science and (in part) Technical University of Gdansk.
´
REFERENCES
1. Zalkin, H. (1993) Adv. Enz. Rel. Areas Mol. Biol. 66, 203–309.
2. Badet, B., Vermoote, P., Haumont, P. Y., Lederer, F., and Le Goffic, F. (1987) Biochemistry
26, 1940–1948.
3. Badet-Denisot, M-A., Rene, L., and Badet, B. (1993) Bull. Soc. Chim. Fr. 130, 249–255.

296
MILEWSKI ET AL.
4. Chmara, H., Andruszkiewicz, R., and Borowski, E. (1986) Biochim. Biophys. Acta 870, 357–366.
5. Badet, B., Vermoote, P., and Le Goffic, F. (1988) Biochemistry 27, 2282–2287.
6. Golinelli-Pimpaneau, B., Le Goffic, F., and Badet, B. (1989) J. Am. Chem. Soc. 111, 3029–3034.
7. Badet-Denisot, M-A., Leriche, C., Massiere, F., and Badet, B. (1995) Bioorg. Med. Chem. Lett.
5, 815–820.
8. Tempczyk, A., Tarnowska, M., Liwo, A., and Borowski, E. (1992) Eur. Biophys. J. 21, 137–145.
9. Amuro, N., Paluh, J. L., and Zalkin, H. (1985) J. Biol. Chem. 260, 14844–14849.
10. Mei, B., and Zalkin, H. (1989) J. Biol. Chem. 264, 16613–16619.
11. Smith, J. L., Zaluzec, E. J., Wery, J-P., Niu, L., Switzer, R. L., Zalkin, H., and Satow, Y. (1994)
Science 264, 1427–1433.
12. Isupov, M. N., Obmolova, G., Butterworth, S., Badet-Denisot, M-A., Badet, B., Polikarpov,
Littlechild, J. A., and Teplyakov, A. (1996) Structure 4, 801–810.
13. Richards, N. G. J., and Schuster, M. S. (1992) FEBS Lett. 313, 98–102.
14. Wasielewski, C., and Antczak, K. (1981) Synthesis, 540.
15. Yang, C., Qamar, R., Norton, S. J., Cook, P. F., and Minter, D. E. (1994) Tetrahedron 50, 1919–1926.
16. Van Der Baan, J. L., Barnck, J. W. F. K., and Bickelhaupt, F. (1983) J. Antibiotics 36, 784–792.
17. Reisner, D. B. (1956) J. Am. Chem. Soc. 78, 5102–5105.
18. Milewski, S., Smith, R. J., Brown, A. J. P., and Gooday, G. W. (1996) in Chitin Enzymology
(Muzzarelli, R.A.A., Ed.), Vol. 2, pp. 439–446, Atec Edizioni, Grottamare.
19. Kenig, M., Vandamme, E., and Abraham, E. P. (1976) J. Gen. Microbiol. 94, 46–54.
20. Yamauchi, K., Ohtsuki, S., and Kinoshita, M. (1984) J. Org. Chem. 49, 1158–1163.
21. Jacobsen, N., and Barlett, P. (1981) J. Am. Chem. Soc. 103, 654–657.
22. McLeod, D. A., Brinkworth, R. I., Ashley, J. A., Janda, K. D., and Wirshing, P. (1991) Bioorg.
Med. Chem. Lett. 1, 653–658.
23. Bearne, S. L., and Wolfenden, R. (1995) Biochemistry 34, 11515–11520.
24. Masters, D. S., and Meister, A. (1982) J. Biol. Chem. 257, 8711–8715.
25. Krahn, J. M., Kim, J. H., Burns, M. R., Parry, R. J., Zalkin, H., and Smith, J. L. (1997) Biochemistry
36, 11061–11068.
26. Bearne, S. L. (1996) J. Biol. Chem. 271, 3052–3057.
27. Leriche, C., Badet-Denisot, M-A., and Badet, B. (1997) Eur. J. Biochem. 245, 418–422.
28. Smith, R. J., Milewski, S., Brown, A. J. P., and Gooday, G. W. (1996) J. Bacteriol. 178, 2320–2327.