Phosphinate inhibitors of the D-glutamic acid-adding enzyme of peptidoglycan biosynthesis

Article abstract of DOI:10.1021/jo951780a

We report the synthesis and initial evaluation of the first effective inhibitors of the D-glutamic acid-adding enzyme (UDP-N-acetylmuramoyl-L-alanine:D-glutamate ligase or MurD). This enzyme plays a key role in bacterial peptidoglycan biosynthesis and is therefore a target for antibiotic design. Phosphinic acid 3 is a dipeptide analog linked to uridine diphosphate by a hydrophobic spacer. It is a good inhibitor of the enzyme (IC₅₀ = 0.68 μM) as it closely resembles the tetrahedral intermediate that is presumed to form in the ligation reaction. Compound 4 lacks the terminal UMP group, and compound 5 lacks both the linker and UDP functionalities. These are less effective inhibitors of the enzyme with IC₅₀ values of 29 μM and >1 mM, respectively. Preincubation of the enzyme in the presence of inhibitor 3 and ATP does not result in irreversible inhibition or in the formation of a slowly decomplexing species, suggesting that the phosphinic acid is not phosphorylated in the active site.

Full text of DOI:10.1021/jo951780a

1756
J . Org. Chem. 1996, 61, 1756-1760
P h osp h in a te In h ibitor s of th e D-Glu ta m ic Acid -Ad d in g En zym e of
P ep tid oglyca n Biosyn th esis
Martin E. Tanner,*,† Sabine Vaganay,‡ J ean van Heijenoort,‡ and Didier Blanot‡
Department of Chemistry, University of British Columbia, Vancouver, British Columbia, V6T 1Z1,
Canada and URA 1131 du CNRS, Universite´ de Paris-Sud, 91405 Orsay, France
Received October 2, 1995X
We report the synthesis and initial evaluation of the first effective inhibitors of the D-glutamic
acid-adding enzyme (UDP-N-acetylmuramoyl-L-alanine:D-glutamate ligase or MurD). This enzyme
plays a key role in bacterial peptidoglycan biosynthesis and is therefore a target for antibiotic design.
Phosphinic acid 3 is a dipeptide analog linked to uridine diphosphate by a hydrophobic spacer. It
is a good inhibitor of the enzyme (IC50 ) 0.68 µM) as it closely resembles the tetrahedral intermediate
that is presumed to form in the ligation reaction. Compound 4 lacks the terminal UMP group,
and compound 5 lacks both the linker and UDP functionalities. These are less effective inhibitors
of the enzyme with IC50 values of 29 µM and >1 mM, respectively. Preincubation of the enzyme
in the presence of inhibitor 3 and ATP does not result in irreversible inhibition or in the formation
of a slowly decomplexing species, suggesting that the phosphinic acid is not phosphorylated in the
active site.
In tr od u ction
Many antibiotics, such as penicillins, cephalosporins,
vancomycin, and D-cycloserine, are known to interfere
with bacterial peptidoglycan construction.1 Peptidogly-
can is a network of polysaccharides that are cross-linked
by D-amino acid-containing polypeptides. It serves to
impart rigidity to the cell wall and protects bacteria from
osmotic lysis. The UDP-N-acetylmuramic acid (UDP-
MurNAc)2 pentapeptide 1 (Figure 1) is a common precur-
sor in the biosynthesis of peptidoglycan in both Gram-
positive and Gram-negative bacteria, and therefore
enzymes involved in its biosynthesis serve as attractive
targets for the design of broad spectrum antibiotics.3
Compound 1 is biosynthesized from UDP-MurNAc to
which is sequentially added L-Ala, D-Glu, meso-Dap (or
L-Lys), and finally the dipeptide D-Ala-D-Ala by a series
of ATP-dependent amino acid ligases. In this paper we
describe the syntheses of, and initial studies on, the first
examples of effective mechanism-based inhibitors of the
second of these ligases, the D-glutamic acid-adding en-
zyme (Figure 1).
F igu r e 1. The structure of UDP-MurNAc pentapeptide 1 and
the reaction catalyzed by the D-glutamic acid-adding enzyme.
ATP-dependent amide-forming enzymes are thought
to operate by catalyzing an initial phosphorylation of the
acid carboxylate (Figure 2).4-7 The resulting acyl phos-
phate is subsequently attacked by the amine to produce
a tetrahedral intermediate which ultimately collapses to
the amide and inorganic phosphate. In several cases it
has been found that appropriately substituted phosphinic
acids act as slow-binding inactivators of these enzymes.6a,8,9
The remarkable feature of this inhibition is that the
enzyme promotes the transfer of the γ-phosphate of ATP
onto the phosphinate anion to produce ADP and a
phosphorylated inhibitor. This process has been con-
firmed by X-ray diffraction analysis of the phosphorylated
†University of British Columbia.
‡Universite´ de Paris-Sud.
(5) D-Ala-D-Ala ligase: (a) Shi, Y.; Walsh, C. T. Biochemistry 1995,
34, 2768. (b) Fan, C.; Moews, P. C.; Walsh, C. T.; Knox, J . R. Science
1994, 266, 439. (c) Wright, G. D.; Walsh, C. T. Acc. Chem. Res. 1992,
25, 468. (d) Mullins, L. S.; Zawadzke, L. E.; Walsh, C. T.; Raushel, F.
M. J . Biol. Chem. 1990, 265, 8993.
X Abstract published in Advance ACS Abstracts, February 1, 1996.
(1) Gale, E. F.; Cundliffe, E.; Reynolds, P. E.; Richmond, M. H.;
Waring, M. J . The Molecular basis of Antibiotic Action, 2nd ed.; J ohn
Wiley and Sons: London, 1981.
(2) Abbreviations: UDP ) uridine diphosphate; UMP ) uridine
monophosphate; meso-Dap ) meso-diaminopimelic acid; TEA ) tri-
ethylamine; HOBt ) 1-hydroxybenzotriazole; DCI ) desorption chemi-
cal ionization; LSI ) liquid secondary ion.
(6) Glutathione synthetase: (a) Hiratake, J .; Kato, H.; Oda, J . J .
Am. Chem. Soc. 1994, 116, 12059. (b) Kato, H.; Tanaka, T.; Yamaguchi,
H.; Hara, T.; Nishioka, T.; Katsube, Y.; Oda, J . Biochemistry 1994,
33, 4995. (c) Meister, A. In The Enzymes; 3rd ed., Boyer, P. D., Ed.;
Academic Press: New York, 1974; Vol. 10, pp 671-697.
(7) The multienzyme systems involved in microbial peptide biosyn-
theses often utilize an alternative thiotemplate mechanism. An initial
formation of an aminoacyl adenylate is followed by the formation of a
thioester and then an amide. For leading references see: Stein, T.;
Vater, J .; Kruft, V.; Wittmann-Liebold, B.; Franke, P.; Panico, M.; Mc
Dowell, R.; Morris, H. R. FEBS Lett. 1994, 340, 39. Preliminary studies
indicate that the D-glutamic acid-adding enzyme is not following such
a mechanism.11
(3) van Heijenoort, J . In Bacterial Cell Wall; Ghuysen, J .-M.,
Hakenbeck, R., Eds.; Elsevier Science B.V.: Amsterdam, 1994; pp 39-
54.
(4) Glutamine synthetase: (a) Meek, T. D.; J ohnson, K. A.; Vil-
lafranca, J . J . Biochemistry 1982, 21, 2158. (b) Meek, T. D.; Villafranca,
J . J . Biochemistry 1980, 19, 5513. (c) Midelfort, C. F.; Rose, I. A. J .
Biol. Chem. 1976, 251, 5881. (d) Meister, A. In The Enzymes; 3rd ed.,
Boyer, P. D., Ed.; Academic Press: New York, 1974; Vol. 10, pp 699-
754.
0022-3263/96/1961-1756$12.00/0 © 1996 American Chemical Society

Phosphinate Inhibitors of Peptidoglycan Biosynthesis
J . Org. Chem., Vol. 61, No. 5, 1996 1757
Ch a r t 1
hedral intermediate 2 is formed in the enzyme active site
and that analogs of this intermediate would therefore
inhibit the enzyme. With this goal in mind, inhibitor 3
was designed. Inhibitor 3 contains a phosphinic acid that
has a tetrahedral geometry at the “dipeptide” center,
which could be enzymatically phosphorylated to give a
close analog of intermediate 2. It also retains the charged
UDP moiety which is likely to be important for binding.
The muramoyl residue has been replaced by a hydropho-
bic linker of appropriate length for ease of synthesis. The
truncated inhibitors 4 and 5 are also of interest in that
they can provide information on the importance of the
UDP and the linker/phosphate groups to the observed
inhibition. We have synthesized compounds 3-5 as a
mixture of diastereomers about the “peptide” chiral
centers and tested them as inhibitors of the purified
D-glutamic acid-adding enzyme from Escherichia coli.
F igu r e 2. The proposed mechanism followed by ATP-depend-
ent amide-forming enzymes and the general structure of a
phosphorylated phosphinate inhibitor.
inhibitor-enzyme complex in the cases of D-Ala-D-Ala
ligase5b and glutathione synthetase.6a The resulting
phosphoryl phosphinate moiety closely mimics the tet-
rahedral intermediate formed in the normal reaction
pathway and is tightly bound by the enzyme. In the case
of D-Ala-D-Ala ligase, an enzyme-inhibitor decomplex-
ation half-life of 17 days was measured, which renders
this noncovalent inhibition effectively irreversible.9a
The D-glutamic acid-adding enzyme (UDP-N-acetyl-
muramoyl-L-alanine:D-glutamate ligase or MurD) cata-
lyzes the addition of D-glutamate to the free carboxylate
of UDP-MurNAc-L-Ala (Figure 1).10 Although relatively
few mechanistic studies11 have been performed with this
enzyme, and no effective inhibitors are known,12 it is
reasonable to assume that the mechanism outlined in
Figure 2 is followed. This would mean that the tetra-
Resu lts
Compounds 3-5 were synthesized from the known
phosphonous methyl ester 6 (Scheme 1).9c,13 Treatment
of racemic 6 with sodium methoxide and dimethyl 2-me-
thylenepentanedioate14 provided a good yield of com-
pound 7 as a mixture of diastereomers. Hydrogenolysis
of the carbobenzyloxy protecting group provided the free
amine which was unstable to heat or prolonged storage
at room temperature and was used directly in subsequent
reactions. Acylation of the amine with acetyl chloride
gave compound 8 which was deprotected under basic
conditions to afford compound 5. Compound 5 prepared
in this fashion was found to be an equimolar mixture of
the four stereoisomers possible from the two “peptide”
chiral centers (two pairs of enantiomers) as indicated by
the presence of two phosphinate 31P NMR signals of
approximately equal intensity.
Hydrogenation of 7 followed by a DCC/1-hydroxyben-
zotriazole-mediated coupling of the resulting amine and
acid 9 provided compound 10 in good yield. Deprotection
of 10 was brought about by hydrogenation over PtO2 at
50 psi followed by treatment with NaOH. This produced
compound 4 which was purified by size exclusion chro-
matography. Compound 3 was prepared using the
standard Khorana coupling conditions15 with the tri-
octylammonium salt of 4 and uridine 5′-monophospho-
morpholidate in pyridine and was purified by preparative
(8) Phosphinothricin: (a) Abell, L. M.; Villafranca, J . J . Biochemistry
1991, 30, 6135. (b) Colanduoni, J . A.; Villafranca, J . J . Bioorg. Chem.
1986, 14, 163. (c) Bayer, E.; Gugel, K. H.; Hagele, K.; Hagenmaier,
H.; J essipow, S.; Konig, W. A.; Zahner, H. Helv. Chim. Acta 1972, 55,
224.
(9) D-Ala-D-Ala ligase inhibition: (a) McDermott, A. E.; Creuzet, F.;
Griffin, R. G.; Zawadzke, L. E.; Ye, Q. Z.; Walsh, C. T. Biochemistry
1990, 29, 5767. (b) Duncan, K.; Walsh, C. T. Biochemistry 1988, 27,
3709. (c) Parsons, W. H.; Patchett, A. A.; Bull, H. G.; Schoen, W. R.;
Taub, D.; Davidson, J .; Combs, P. L.; Springer, J . P.; Gadebusch, H.;
Weissberger, B.; Valiant, M. E.; Mellin, T. N.; Busch, R. D. J . Med.
Chem. 1988, 31, 1772. See also references 5a and 5b.
(10) (a) Pratviel-Sosa, F.; Mengin-Lecreulx, D.; van Heijenoort, J .
Eur. J . Biochem. 1991, 202, 1169. (b) Michaud, C.; Blanot, D.; Flouret,
B.; van Heijenoort, J . Eur. J . Biochem. 1987, 166, 631. (c) Nathenson,
S. G.; Strominger, J . L.; Ito, E. J . Biol. Chem. 1964, 239, 1773.
(11) Vaganay, S.; Tanner, M. E.; van Heijenoort, J .; Blanot, D.
Microbial Drug Resistance: Mechanisms, Epidemiology, and Diseases,
in press.
(13) Baylis, E. K.; Campbell, C. D.; Dingwall, J . D. J . Chem. Soc.
Perkin Trans. 1 1984, 2845.
(12) (a) Auger, G.; Deprun, C.; van Heijenoort, J .; Blanot, D. J . Prakt.
Chem./ Chem. Z. 1995, 337, 351. (b) Pratviel-Sosa, F.; Acher, F.;
Trigalo, F.; Blanot, D.; Azerad, R.; van Heijenoort, J . FEMS Microbiol.
Lett. 1994, 115, 223. See also reference 10b.
(14) Cookson, R. C.; Smith, S. A. J . Chem. Soc. Perkin Trans. 1 1979,
2447.
(15) Roseman, S.; Distler, J . J .; Moffatt, J . G.; Khorana, H. G. J .
Am. Chem. Soc. 1961, 83, 659.

1758 J . Org. Chem., Vol. 61, No. 5, 1996
Tanner et al.
Sch em e 1
layer chromatography followed by size exclusion chro-
matography. It is interesting to note that the phosphate
coupling reaction can be successfully carried out in the
presence of several carboxylates and a phosphinate
functionality. Compounds 3 and 4 were also isolated as
equimolar mixtures of the four possible stereoisomers as
indicated by the phosphinate signals in their 31P NMR
spectra. Attempts to separate the diastereomers by
reverse-phase HPLC were unsuccessful.16
Compounds 3-5 were tested for their ability to inhibit
the addition of D-[14C]Glu to UDP-MurNAc-L-Ala cata-
lyzed by recombinant E. coli D-glutamic acid-adding
enzyme. The concentration of inhibitor necessary to
reduce the rate of addition by one-half (IC50 value) was
determined in the presence of 25 µM UDP-MurNAc-L-
Ala (Km ) 7.5 µM) and 25 µM D-Glu (Km ) 55 µM).
Compound 3 was found to be a very good inhibitor of the
enzyme with an IC50 ) 0.68 µM. Compound 4 was a
weaker inhibitor with an IC50 ) 29 µM. Compound 5
was a poor inhibitor of the enzyme (IC50 > 1 mM), and
the addition of UDP-N-acetylglucosamine (1 mM) did not
significantly increase the inhibitory activity (IC50 ≈ 1
mM). Preincubation of the enzyme and 3, with and
without added ATP, followed by dilution into an assay
mixture and kinetic analysis, showed no signs of ir-
reversible inhibition or slow decomplexation rates.
than that observed. The absence of any irreversible
inhibition or slow decomplexation rates indicates that 3
is behaving differently than other reported phosphinate
inhibitors and is possibly not being phosphorylated in the
active site. An ATP/3 complex may closely resemble an
ADP/2 complex and account for the observed inhibition.
Compound 4 was also a good inhibitor of the enzyme
but needed to be present at concentrations 43-fold higher
than 3 in order to show comparable inhibition. This
indicates that it is important, but not crucial, to retain
the terminal UMP functionality when designing inhibi-
tors of this enzyme. This also agrees well with the
observation that 1-phospho-MurNAc-L-Ala is a substrate
for the enzyme and has a Km value 36-fold higher than
the natural substrate, UDP-MurNAc-L-Ala.10b Com-
pound 5 was a poor inhibitor of the enzyme since it lacked
both the hydrophobic spacer and the UDP functionality.
Attempts at increasing the potency of 5 by including
UDP-N-acetylglucosamine in the medium (which should
be accommodated in the active site as a coinhibitor
without any steric problems) met with limited success.
This is likely due to the entropic costs of binding two
separate ligands, and points out the importance of
maintaining a covalent bond between the “sugar” and
“peptide” portions when designing inhibitors of this
enzyme.17 Analogs of 3 that contain a muramoyl sugar
in place of the lipophilic spacer will likely act as very
potent inhibitors of, and may be phosphorylated by, the
D-glutamic acid-adding enzyme.
Discu ssion
The observation that compound 3 is a good inhibitor
of the D-glutamic acid-adding enzyme is consistent with
the notion that the enzyme will strongly bind species
containing a tetrahedral geometry and an oxyanion
between the two amino acid units (as in the putative
intermediate 2). It also indicates that the muramoyl
sugar is not a crucial recognition element for the enzyme
and that hydrophobic linkers are accommodated in its
place. Since the sample consisted of an equimolar
mixture of four stereoisomers, and since the one most
closely mimicking the L-Ala-D-Glu stereochemistry would
presumably be bound the strongest, the actual IC50 for
the best stereoisomer may be as much as four-fold lower
Exp er im en ta l Section
Gen er a l P r oced u r es. Chemicals were purchased from
Aldrich Chemical Co. or Sigma Chemical Co. UDP-MurNAc-
L-Ala was prepared according to Michaud et al.10b D-[14C]Glu
was obtained after treatment of D,L-[14C]Glu (NEN, Du Pont
de Nemours, Les Ulis, France; 2 GBq/mmol) with L-glutamate
decarboxylase and purification by paper rheophoresis. Ion
exchange and gel chromatographic supports were from Bio-
Rad Laboratories. Silica gel (230-400 mesh) was from BDH
Inc. and preparative layer chromatography plates (2 mm,
silica gel 60 F254) were from Merck. Pyridine, CH2Cl2, and TEA
were distilled under N2 from CaH2. MeOH was distilled under
(16) Flouret, B.; Mengin-Lecreulx, D.; van Heijenoort, J . Anal.
Biochem. 1981, 114, 59.
(17) J encks W. P. Catalysis in Chemistry and Enzymology; Dover
Publications, Inc.: New York, 1975; pp 615-806.

Phosphinate Inhibitors of Peptidoglycan Biosynthesis
J . Org. Chem., Vol. 61, No. 5, 1996 1759
N2 from magnesium methoxide. DMF was dried over 4 Å
molecular sieves and vacuum degassed before use. All reac-
tions performed in organic solvents were carried out under 1
atm of argon unless otherwise specified. All compounds were
prepared for microanalysis by heating at 70 °C (0.1 mmHg)
for 48 h, except compound 3 which was dried at rt (0.1 mmHg)
for 48 h. Microanalyses were carried out in the microanalytical
laboratory at the University of British Columbia by Mr. Peter
Borda.
Met h yl [1-[(Ben zyloxyca r b on yl)a m in o]et h yl](2,4-d i-
ca r bom eth oxybu tyl)p h osp h in a te (7). To a solution of
methyl [1-[(benzyloxycarbonyl)amino]ethyl]phosphinate (6)
(1.78 g, 6.9 mmol) in 15 mL of dry MeOH at 0 °C was added
a solution of sodium methoxide in MeOH (3.8 mL of a 2 M
solution). The reaction was stirred for 5 min and neat
dimethyl 2-methylenepentanedioate (1.1 mL, 6.9 mmol) was
added dropwise. The reaction was stirred at 0 °C for 15 min
and then allowed to warm to rt and stirred for an additional
4 h. The reaction was quenched with 1 M HCl (150 mL) and
extracted twice with CH2Cl2. The combined organic extracts
were dried over MgSO4, and the solvent was removed in vacuo.
The residue was purified by silica gel chromatography (MeCN-
EtOAc 1:9 v/v) to give 2.5 g (84%) of 8 as a colorless oil: 1H
NMR (CDCl3) δ 7.34 (m, 5H), 5.34-5.08 (3 brd, 1H), 5.10 (s,
2H), 4.05 (brm, 1H), 3.68 (m, 9H), 2.83 (brm, 1H), 2.35-1.17
(m, 6H), 1.39-1.27 (m, 3H); 31P NMR (CDCl3) δ 52.84 (s), 52.03
(s); DCI-MS (NH3) 430 (M + H+, 100%); Anal. Calcd for
C19H28NO8P: C, 53.14; H, 6.57; N, 3.26. Found: C, 52.83; H,
6.66; N, 3.40.
Meth yl [1-Aceta m id oeth yl](2,4-d ica r bom eth oxybu tyl)-
p h osp h in a te (8). A solution of methyl [1-[(benzyloxycarbo-
nyl)amino]ethyl](2,4 dicarbomethoxybutyl)phosphinate (7) (363
mg, 0.85 mmol) in 5 mL of MeOH containing 5% Pd/C (70 mg)
was stirred under 1 atm of hydrogen at room temperature for
11 h. The mixture was filtered, and the solvent was removed
under reduced pressure to give 245 mg of a colorless oil. This
was dissolved in 10 mL of dry CH2Cl2, and triethylamine (0.5
mL) and acetyl chloride (0.070 mL, 0.98 mmol) were added.
The reaction was stirred for 2 h at rt, and the solvent was
removed in vacuo. The residue was taken up in minimal
EtOAc and filtered to remove the insoluble salts. This solution
was loaded directly onto a silica gel column and eluted with
MeOH-EtOAc (1:19 v/v). This procedure afforded 200 mg
(71%) of 8 as a colorless oil: 1H NMR (CDCl3) δ 6.47-6.05 (2
brd, 1H), 4.42 (m, 1H), 3.76-3.65 (m, 9H), 2.83 (brm, 1H),
2.36-2.16 (m, 3H), 2.02 (s, 3H), 2.01-1.80 (m, 3H), 1.39-1.26
(m, 3H); 31P NMR (CDCl3) δ 53.38 (s), 52.54 (s), 52.45 (s), 52.37
(s); DCI-MS (NH3) 338 (M + H+, 100%); Anal. Calcd for
C13H24NO7P: C, 46.29; H, 7.17; N, 4.15. Found: C, 46.29; H,
7.09; N, 4.03.
80%), 318 (M - 2Na+ + 3H+, 70%), 384 (M + Na+, 70%); Anal.
Calcd for C10H15NO7PNa3•H2O: C, 31.68; H, 4.52; N, 3.69.
Found: C, 31.95; H, 4.19; N, 3.75.
5-Ca r boxyp en tyl Dip h en yl P h osp h a te (9). To a solution
of 6-hydroxyhexanoic acid18 (3.3 g, 25 mmol) in 50 mL of dry
pyridine was slowly added diphenyl phosphorochloridate (6.7
g, 25 mmol). The mixture was stirred at rt for 12 h, and the
solvent was removed in vacuo. The residue was dissolved in
CH2Cl2 and extracted twice with 1 M HCl. The organic layer
was dried over MgSO4, and the solvent was removed in vacuo.
The residue was purified by silica gel chromatography (MeOH-
CH2Cl2 1:19 v/v) to give 4.6 g (50%) of 9 as a colorless oil: 1H
NMR (CDCl3) δ 7.4-7.1 (m, 10H), 4.23 (dt, 2H, J ) 7.5 Hz, J
) 6.4 Hz), 2.29 (t, 2H, J ) 7.2 Hz), 1.64 (m, 4H), 1.38 (m, 2H);
31P NMR (CDCl3) δ -12.02 (s); DCI-MS (NH3) 365 (M + H+,
100%); Anal. Calcd for C18H21O6P: C, 59.34; H, 5.81. Found:
C, 58.97; H, 5.90.
Meth yl [1-[6-[(Diph en oxyph osph in yl)oxy]h exan am ido]-
eth yl](2,4-d ica r bom eth oxybu tyl)p h osp h in a te (10). A sol-
ution of methyl [1-[(benzyloxycarbonyl)amino]ethyl](2,4 dicar-
bomethoxybutyl)phosphinate 7 (1.04 g, 2.4 mmol) in 10 mL of
MeOH containing 5% Pd/C (200 mg) was stirred under 1 atm
of hydrogen at rt for 11 h. The mixture was filtered, and the
solvent was removed in vacuo to give 702 mg of the amine as
a colorless oil. In a separate flask, a solution of 5-carboxy-
pentyl diphenyl phosphate (9) (0.862 g, 2.4 mmol) in CH2Cl2/
DMF (4:1 v/v, 12.5 mL) was prepared. To this solution was
added 1-hydroxybenzotriazole•H2O (0.351 g, 2.6 mmol) and
dicyclohexylcarbodiimide (0.513 g, 2.5 mmol). The mixture
was stirred at rt for 1 h during which time a white precipitate
formed. A solution of the amine in DMF (15 mL) was then
added, and the mixture was allowed to stir for 24 h at rt. The
reaction was filtered and the filtrate was partitioned between
water (120 mL) and CH2Cl2 (120 mL). The organic layer was
back-extracted with water (120 mL) and then dried over
MgSO4. The solvent was removed in vacuo, and the residue
was dissolved in EtOAc (5 mL) and applied to a column (100
mL) of silica gel. The column was washed with EtOAc (300
mL) and the product was eluted with MeOH-EtOAc (2:23 v/v)
to give 1.34 g (88%) of 10 as a colorless oil: 1H NMR (CDCl3)
δ 7.37-7.13 (m, 10H), 6.24-5.85 (4 brd, 1H), 4.43 (m, 1H),
4.23 (dt, 2H, J ) 7.6 Hz, J ) 6.6 Hz), 3.72-3.60 (m, 9H), 2.82
(brm, 1H), 2.30 (t, 2H, J ) 7.2 Hz), 2.15 (m, 3H), 1.98-1.58
(m, 7H), 1.41-1.23 (m, 5H); 31P NMR (CDCl3) δ 53.10 (s), 52.40
(s), 52.32 (s), -11.96 (s); DCI-MS (NH3) 642 (M + H+, 100%);
Anal. Calcd for C29H41NO11P2: C, 54.29; H, 6.44; N, 2.18.
Found: C, 54.05; H, 6.38; N, 2.20.
[1-(6-P h osp h on oh exa n a m id o)eth yl](2,4-d ica r boxybu -
tyl)p h osp h in a te Tr isod iu m Sa lt (4). A solution of com-
pound 10 (0.635 g, 0.99 mmol) in 15 mL of MeOH containing
PtO2 monohydrate (160 mg) was shaken under 50 psi of
hydrogen at rt for 20 h. The mixture was filtered, and the
solvent was removed in vacuo to give 513 mg of the phosphate
as a colorless oil. This was dissolved in 0.5 M NaOH (20 mL)
and stirred at rt for 3 h. The solution was brought to pH 5 by
the addition of AG 50W-X8 resin (H+ form, 20-50 mesh) and
filtered. The pH was adjusted to 7 with 0.5 M NaOH, and
the volume was reduced to 10 mL under reduced pressure. To
this solution was added 10 mL of ethanol and 10 mL of acetone.
The sample was cooled at -20 °C overnight, and the resulting
oil was obtained by decanting the solvent. The oil was
dissolved in water (6 mL) and applied, in three separate
portions, to a column (250 mL) of Bio-Gel P-2 (fine). The
product was eluted with water, and the fastest moving
fractions were lyophilized to dryness to give 457 mg (90%) of
4 (trisodium salt) as a white solid: 1H NMR (D2O) δ 3.95 (m,
1H), 3.80 (dt, 2H, J ) 6.6 Hz, J ) 6.6 Hz), 2.54 (brm, 1H),
2.25 (t, 2H, J ) 7.2 Hz), 2.17 (t, 2H, J ) 8.2 Hz), 1.86 (m, 2H),
1.59 (m, 6H) 1.34 (m, 2H) 1.18 (dd, 3H, J ) 7.3, J ) 13.2 Hz);
31P NMR (D2O) δ 39.64 (s), 39.53 (s), 0.79 (s); +LSIMS
(thioglycerol/H2O matrix) 536 (M + Na+, 100%), 558 (M +
(1-Acetam idoeth yl)(2,4-dicar boxybu tyl)ph osph in ic acid
(5). A solution of methyl (1-acetamidoethyl)(2,4 dicarbometh-
oxybutyl)phosphinate (8) (320 mg, 0.95 mmol) in 10 mL of 0.5
M NaOH was stirred at rt for 3 h. The solution was
neutralized by the addition of AG 50W-X8 resin (H+ form, 20-
50 mesh) and filtered. The pH was then readjusted to 10 with
2 M NaOH, and the solution was applied to a column (10 mL)
of AG 1-X8 (formate form, 100-200 mesh). The column was
washed with water (50 mL) and 0.5 M formic acid (50 mL),
and the product was eluted with 4 M formic acid (80 mL). The
latter fraction was evaporated to dryness in vacuo to yield 280
mg (99%) of the free acid 5 as a hygroscopic foam. For use in
inhibition studies the trisodium salt was prepared by dissolv-
ing the free acid in 15 mL of water and adjusting the pH to
7.5 with 0.5 M NaOH. The volume was reduced to 5 mL under
reduced pressure, and 10 mL of ethanol and 2 mL of acetone
were added. The sample was cooled at -20 °C overnight, and
the resulting oil was obtained by decanting the solvent. The
oil was dissolved in 5 mL of water and lyophilized to dryness
to yield the salt as a white solid: 1H NMR of free acid (D2O)
δ 4.11 (m, 1H), 2.71 (brm, 1H), 2.39 (t, 2H, J ) 7.3 Hz), 2.12-
1.75 (m, 4H), 1.96 (s, 3H), 1.23 (dd, 3H, J ) 7.4, 14.5 Hz); 31
P
NMR of free acid (D2O) δ 48.67 (s), 48.38 (s). Anal. Calcd for
C10H18NO7P (free acid): C, 40.68; H, 6.15; N, 4.74. Found:
C, 41.00; H, 6.26; N, 4.65. +LSIMS of trisodium salt (thiogly-
cerol/H2O matrix) 340 (M - Na+ + 2H+, 100%), 362 (M + H+,
(18) This compound was prepared by refluxing a solution of capro-
lactone in 4 N NaOH for 1 h, washing with CH2Cl2, acidifying with
HCl, and extracting with ether. The organic phase was removed in
vacuo, and the resulting clear oil was used without further purification.

1760 J . Org. Chem., Vol. 61, No. 5, 1996
Tanner et al.
2Na+ - H+, 95%), 514 (M + H+, 50%), 580 (M + 3Na+ - 2H+,
45%). Anal. Calcd for C14H24NO11P2Na3•0.5H2O: C, 32.20; H,
4.82; N, 2.68. Found: C, 32.24; H, 4.56; N, 2.57.
followed by treatment with AG 50W-X8 resin and Chelex 100
to afford 11 mg of pure 3 as a white solid: 1H NMR (D2O) δ
7.94 (d, 1H, J ) 8.1 Hz), 5.95 (d, 1H, J ) 2.9 Hz), 5.93 (d, 1H,
J ) 8.1 Hz), 4.33 (d, 2H, J ) 3.6 Hz), 4.24 (m, 1H), 4.18 (m,
2H), 3.91 (m, 1H), 3.88 (dt, 2H, J ) 6.5 Hz, J ) 6.6 Hz), 2.43
(brm, 1H), 2.24 (t, 2H, J ) 7.4 Hz), 2.14 (t, 2H, J ) 8.6 Hz),
1.67 (m, 2H), 1.59 (m, 6H) 1.34 (m, 2H) 1.18 (dd, 3H, J ) 7.7,
13.7 Hz); 31P NMR (D2O) δ 39.90 (s), 39.62 (s), -10.57 (d, J )
21.3 Hz), -11.29 (d, J ) 21.1 Hz); +LSIMS (thioglycerol/H2O
matrix) 864 (M + H+, 100%), 842 (M - Na+ + 2H+, 80%), 886
(M + Na+, 70%), 908 (M + 2Na+ - H+, 35%) 820 (M - 2Na+
+ 3H+, 35%); HR+LSIMS Calcd for C23H34N3O19P3Na5 864.0488.
Found: 864.0505; Anal. Calcd for C23H33N3O19P3Na5•5H2O:
C, 28.97; H, 4.55; N, 4.41. Found: C, 29.10; H, 4.65; N, 4.14.
In h ibition Stu d ies. The enzyme used in this study was
purified from overproducing strain E. coli J M83 (pMLD58) as
described elsewhere.11 The concentrations of stock solutions
of 3 were calculated from A262 measurements using ꢀ ) 9890
M-1 cm-1. The D-glutamic acid-adding activity was assayed
by following the appearance of UDP-MurNAc-L-Ala-D-[14C]Glu
in a mixture containing, in a final volume of 50 µL, 0.1 M Tris/
HCl, pH 8.6, 5 mM MgCl2, 5 mM ATP, 25 µM UDP-MurNAc-
L-Ala, 25 µM D-[14C]Glu (0.88 KBq), enzyme, and inhibitor. The
mixture was incubated for 30 min at 37 °C, and the reaction
was stopped by the addition of 10 µL glacial acetic acid. The
radioactive substrate and product were separated by HPLC
and quantitated by on-line scintillation counting as previously
described.12a Preincubation experiments were carried out for
10 min in the presence of 3 mM 3 and 5 mM ATP. The
samples were then diluted 25 fold and assayed directly.
[1-[(6-Ur id in ed ip h osp h o)h exa n a m id o]eth yl](2,4-d ica r -
b oxyb u t yl)p h osp h in a t e P en t a sod iu m Sa lt (3). A solu-
tion of compound 4 (0.192 g, 0.37 mmol) in 10 mL of water
was acidified by the addition of AG 50W-X8 resin (H+ form,
20-50 mesh) and filtered. Pyridine (6 mL) was added, and
the volume was decreased to 5 mL under reduced pressure. A
solution of trioctylamine (0.820 mL, 1.9 mmol) in 10 mL of
pyridine was added, and the solvent was removed under
reduced pressure. Three successive additions of pyridine (15
mL) followed by evaporation to dryness under reduced pres-
sure were used to remove the residual water. In a separate
flask a sample of uridine 5′-monophosphomorpholidate (0.308
g, 0.45 mmol) was dried by three successive additions of
pyridine (10 mL) followed by evaporation under reduced
pressure. The morpholidate was dissolved in 15 mL of
pyridine and added to the trioctylammonium salt of 4. The
reaction was stirred at rt for 48 h. An additional sample of
uridine 5′-monophosphomorpholidate (0.150 g, 0.22 mmol) was
dried as before and added to the reaction in 5 mL of pyridine.
The reaction was stirred for an additional 24 h, and the volume
was decreased to 3 mL under reduced pressure. Preparative
layer chromatography (2-isopropanol-water-ammonium hy-
droxide, 7:3:1 v/v) and extraction (2-isopropanol-water, 4:1
v/v) of the slowest band visible under UV light afforded a
mixture containing 3, 4, and uridine 5'-monophosphate. This
mixture was applied to a column (250 mL) of Bio-Gel P-2 (fine)
and eluted with water. The fastest moving UV active fractions
were collected, concentrated under reduced pressure, passed
through a column (5 mL) of AG 50W-X8 resin (Na+ form, 20-
50 mesh) and a column (5 mL) of Chelex 100 (Na+ form, 50-
100 mesh), and then lyophilized to dryness. This afforded 60
mg of 3 (17% yield calculated as pentasodium salt pentahy-
drate) containing a 15% impurity of 4 as a white solid.
Selected fractions (20 mg) containing <5% of 4 were further
purified on a second column (250 mL) of Bio-Gel P-2 (fine)
Ack n ow led gm en t. We thank the University of
British Columbia, the Natural Sciences and Engineering
Research Council of Canada (NSERC), and the Centre
National de la Recherche Scientifique (URA 1131) for
financial support.
J O951780A