GMP synthetase: Synthesis and biological evaluation of a stable analog of the proposed AMP-XMP reaction intermediate

Article abstract of DOI:10.1055/s-1999-3104

The enzyme guanosine monophosphate synthetase is part of the de novo purine synthesis pathway and is responsible for the t conversion of xanthosine monophosphate to guanosine monophosphate. The phosphate-linked adenyl-XMP 1 has been proposed as the interm

Full text of DOI:10.1055/s-1999-3104

LETTER
897
GMP Synthetase: Synthesis and Biological Evaluation of a Stable Analog of the
Proposed AMP-XMP Reaction Intermediate
Edward J. Salaski1, Hans Maag2*
Division of Bio-Organic Chemistry, Institute of Organic Chemistry, Syntex Discovery Research, 3401 Hillview Avenue, Palo Alto, CA
94304, USA
Fax +1(650)3542442; E-mail: hans.maag@roche.com
Received 1 February 1999
This paper is dedicated to Professor Albert Eschenmoser
probe into the structure of the reactive enzyme complex
Abstract: The enzyme guanosine monophosphate synthetase is part
through enzymatic as well as spectroscopic and X-ray
of the de novo purine synthesis pathway and is responsible for the
crystallographic studies.
conversion of xanthosine monophosphate to guanosine monophos-
phate. The phosphate-linked adenyl-XMP 1 has been proposed as
the intermediate in a stepwise mechanism for this conversion.
Direct observation of 1; either as produced in the enzymatic reac-
tion, or as a potential synthetic target is not readily feasible due to
the high lability of the phosphate linkage. We therefore undertook
the synthesis of the phosphonate 2 as a stable analog of this pro-
posed intermediate. The successful synthetic strategy involved the
coupling under Mitsunobu conditions of a 2-phosphorylmethyl
inosine with an adenosine derivative. This gave 2 after deprotection
and 5’ monophosphate formation of the initial coupling product.
Compound 2 inhibited GMP synthetase with a Ki of 0.56 µM.
Key words: GMP synthetase, AMP-XMP intermediate, phosphory-
lation
The enzyme guanosine monophosphate synthetase
(GMPS) carries out the final purine base modifying step
in the de novo synthesis of guanine nucleotides by the
conversion of xanthosine monophosphate (XMP) to gua-
nosine monophosphate (GMP). GMPS belongs to a group
of enzymes controlling cellular metabolism that operate at
elevated levels of activity in rapidly proliferating cells.3
Among these cell types are neoplasms and lymphocytes.
Inhibition of GMP synthetase would be expected to have
an antiproliferative effect similar to that shown by myco-
phenolic acid, a potent inhibitor of IMP dehydrogenase,
the enzyme responsible for the conversion of inosine
monophosphate (IMP) to XMP.4 GMPS, an amidotrans-
ferase, is a Mg2+ dependent enzyme utilizing ammonia de-
Scheme 1
rived from the deamination of glutamine to effect the
amination of XMP. In the proposed mechanism5-8
We envisaged, that 2 could be constructed either by a
(Scheme 1), XMP is activated by reaction with ATP to
phosphonate coupling between a 2-methylphosphonate
give the phosphate linked intermediate 1. This intermedi-
derivative of inosine, such as i, and adenosine in suitably
ate then reacts with ammonia to give the product guanine
protected form ii (route a) or by displacement of a leaving
group on the 2-methyl group of an inosine derivative iii
with an adenosine 5’-H-phosphonate iv (route b, Scheme
2). Both required the synthesis of 2-methyl modified
inosines, which were approached from the commercially
available 4-amino-5-imidazolecarboxamide α-D-ribose 3
(AICA Riboside) (Scheme 3).
nucleotide and AMP. In the absence of an ammonia
source, the hypothetical intermediate is not accumulated
but is hydrolyzed to XMP and AMP. While the intermedi-
ate 1 has not been observed directly, labeling studies using
18O have provided quite conclusive circumstantial evi-
dence for its existence as an intermediate.5-7
We undertook the synthesis of the phosphonate 2 as a sta-
The hydroxy groups of 3 were protected as silyl ethers in
ble analog of this proposed intermediate. A high affinity
of 2 for the enzyme would support the theory of 1 as the
enzymatic intermediate. In addition, 2 could be used as a
nearly quantitative yield. The purine system was then con-
structed in 51% overall yield by acylation of the amine
Synlett 1999, S1, 897–900 ISSN 0936-5214 © Thieme Stuttgart · New York

898
E. J. Salaski, H. Maag
LETTER
with benzyloxyacetyl chloride and subsequent cyclization Since the 2-methylhydroxyl group derived from debenzy-
with DBU in refluxing toluene to give 4. Attempted acy- lation of 4 could not be converted to a leaving group in the
lation / cyclization sequences with phosphonoacetate de- presence of a free lactam, the hypoxanthine ring of 4
rivatives or potential Arbuzov reaction substrates, such as was protected as the 2-(trimethylsilyl)ethyl ether via
bromoacetyl bromide, were unsuccessful.
Mitsunobu coupling with trimethylsilylethanol. Transfer
hydrogenolysis of the benzyl ether and mesylation of the
resulting alcohol 5 gave the relatively unstable mesylate 6
in 82% overall yield. The phosphonate group was intro-
duced by the treatment of this material with the sodium
anion of dimethyl phosphite. The fully protected phospho-
nate ester 7 was thus obtained in 25% overall yield from
AICA riboside 3.
Coupling according to route b (Scheme 2) was attempted
first. To this end, the partially protected adenosine 89 was
treated with 2-(trimethylsilyl)ethyl dichlorophosphite
(prepared from PCl3 and trimethylsilylethanol)10 to give
the mixed phosphite ester 9 as a pair of phosphorus
diastereomers. This material was treated with NaH and
then added to the mesylate 6. In contrast to the simple
dimethyl case 7, no coupled phosphonate was produced.
Heating the dimethyl phosphonate 7 in t-butylamine solu-
tion effected complete, selective demethylation, giving
the monoester as the t-butylammonium salt 10 in quanti-
tative yield.11 Conversion of 10 to the triethylammonium
salt, followed by exposure to standard phosphate coupling
conditions utilizing hindered arylsulfonyl chlorides, tetra-
zole and alcohol 1212, failed to produce any coupled prod-
uct. Generation of the acid chloride 11 in situ directly
from the triethylammonium salt 10 and quenching with
the alcohol 12 also produced none of the desired coupling
product.
Scheme 2
Scheme 3 a) i) TBDMSCl, imidazole, DMF, r.t. 3 days, 98%; ii) ben-
zyloxyacetyl chloride, 2,6-di-t-butyl-4-methylpyridine, CH2Cl2,
63%; b) DBU, toluene, reflux 16 h, 80%; c) i) Me3Si(CH2)2OH, Ph3P/
DEAD, THF, 81%; ii) HCOONH4, Pd/C, EtOH, reflux, 90%; d)
Ms2O, Et3N, CH2Cl2, 89%; e) dimethyl phosphite, NaH, THF, O°C
followed by r.t. 20 h, 88%.
Scheme 4 a) Me3Si(CH2)2OPCl2; H2O; b) i) NaH, ii) 6; c) t-BuNH2,
70°C, 3 days, 100%; d) i) Et3N, ii) 2,4,6-triisopropyl-benzenesulfonyl
chloride, tetrazole; e) Et3N; SOCl2, DMF (crude).
Synlett 1999, S1, 897–900 ISSN 0936-5214 © Thieme Stuttgart · New York

LETTER
GMP Synthetase: AMP-XMP Intermediate
899
Reversal of the nucleophilic and electrophilic components versed phase HPLC purification. The cleavage of the me-
of the coupling reaction, as in route a (Scheme 2), proved thyl phosphonate ester of highly polar 18 was achieved by
to be the successful strategy (Scheme 5). Salt 10 was con- heating the salt for 3 days in t-butylamine; isolation of the
verted to the free acid by brief treatment with an aqueous major product by HPLC provided a 34% yield of the tris-
KHSO3 solution. The acid was then coupled with the al- t-butylammonium salt of target compound 2.17
cohol 8 under Mitsunobu conditions13 to give a good yield
(87%) of the desired phosphonate 1314 as a 3:2 mixture of
diastereomers.
Scheme 6 a) t-BuNH2, 70°C, 14 h, 100%; b) POCl3/H2O, triethylp-
hosphate, O°C, 6 h, 35% crude; c) t-BuNH2, 70°C, 3 days, 35%.
Scheme 5 a) i) 5% aqueous KHSO3, O°C; ii) 8, Ph3P, DIAD, THF,
The stable phosphonate 2, along with the intermediates 17
and 18, was evaluated for inhibitory properties towards
human GMP synthetase.18 Only the stable phosphonate 2
had appreciable affinity for the enzyme with a Ki value of
87%; b) conc. NH4OH, MeOH, r.t. 19 h, 83%; c) TBAF, THF, r.t., 1
h, 83%; d) TFA/H2O = 4:1, O°C followed by r.t., 1 h, 100%.
Completion of the synthesis required the removal of all 0.56 µM (Table 1). The lack of affinity of the non termi-
protecting groups and the introduction of the terminal 5’- nally phosphorylated analog 17 and the methyl ester 18
phosphate. This was accomplished in the following man- indicates a stringent requirement by the enzyme for the
ner (Schemes 5 and 6). The benzoates were removed from charged functionalities of the linking and terminal phos-
the adenine base with methanolic ammonium hydroxide. phates.
Treatment of this product 14 with tetrabutylammonium
fluoride in THF removed all four silyl protecting groups
and brief exposure of the triol 15 to 80% aqueous trifluo-
roacetic acid cleaved the isopropylidine group to give the
completely deprotected phosphonate ester 16 as the triflu-
oroacetate salt in 69% overall yield from 13.
The phosphonate methyl ester 16 could be cleanly con-
verted to the salt 1715 by heating in t-butylamine (Scheme
6). This de-esterification was completely selective for the
methyl vs. the adenosyl ester. Treatment of 16 with phos-
phoryl chloride, partially poisoned with water, gave the
primary phosphate 1816 as the bis-ammonium salt after re-
Synlett 1999, S1, 897–900 ISSN 0936-5214 © Thieme Stuttgart · New York

900
E. J. Salaski, H. Maag
LETTER
(644 mg, 0.489 mmol, 87%). 1H NMR indicated a mixture of
two diastereomers.
The crystal structure of a catalytically incompetent form
of GMP synthetase from E. coli has been determined.19 In
this structure, the ATP binding site and a putative XMP
binding pocket are located on different, somewhat distant,
structural subunits. Stable binary reaction intermediates,
such as 2, might be useful in inducing the crystallization
of a catalytically competent form of the enzyme, in which
the XMP and ATP binding sites are presumed to be held
in closer proximity.
(15) Preparation of 17: The methyl ester 16 (32 mg, 0.099 mmol)
was placed in a Teflon capped pressure tube, t-butylamine (1.5
mL) was added, the tube sealed and placed in an oil bath main-
tained at 70 °C for 14 h. The mixture was cooled to r.t., and
the white precipitate of product was filtered off and rinsed
with a small amount of t-butylamine to give the first crop of
17 (11.2 mg). The filtrate was evaporated in vacuo and the re-
sidue was dried to give a second crop of 17 (24 mg), combined
yield: 35 mg (0.099 mmol, 100%). 1H NMR (DMSO, 300
MHz): 8.44 (s, 1H, I-H8), 8.22 (s, 1H) and 8.13 (s, 1H) A-H2
and A-H8), 7.26 (br.s. 2H, NH2), 5.89 (d, J=6Hz, 1H) and 5.84
(d, J=6Hz, 1H) I-1’ and A-1’; 4.60 (t, J=5Hz, 1H) and 4.50 (t,
J=5Hz, 1H) I-2’ and A-2’, 4.18 (t, J=5Hz, 1H) and 4.11 (t,
J=5Hz, 1H) I-3’ and A-3’, 2.89 (d, J=29Hz, 2H, CH2P), 1.18
(s, 9H, t-BuN). The signals of the 5’ and 4’ protons could not
be clearly discerned. MS (ES): 612 (M+H)+, 480, 361.
(16) Preparation of 18: Freshly distilled POCl3 (56 µL, 92 mg, 0.6
mmol) and 1.8 µL of water was added to triethyl phosphate
(3.0 mL) under argon at 0°C. The alcohol 16 (171 mg, 0.2
mmol) was added at once and the mixture was stirred at 0°C
for 6 h. The reaction was quenched by the addition of 1M
triethyl ammonium bicarbonate (2 mL) and allowed to warm
to r.t. The solvent was removed under vacuum and the crude
product was purified by HPLC (Vidac C-4 column, gradient of
7-25% MeOH in 0.1 M NH4OAc). The product fractions were
combined and lyophilized to give 18 (142 mg) containing
NH4OAc, yet suitable for use in the next step. A small sample
was re-lyophilized to remove the excess buffer and characte-
rized by NMR (2:3 mixture of diastereomers). 1H NMR (D2O,
300 MHz): 8.38 and 8.34 (2 s, 1H, I-H8), 8.27 and 8.24 (2 s,
1H, A-H2), 8.18 and 8.15 (2 s, 1H, A-H8), 6.15 – 6.05 (4 d,
2H, J=5 Hz, I-1’ and A-1’), 4.04 and 3.98 (2 d, 3H, J=11Hz,
OMe), 3.79 and 3.78 (2d, 2H, J=22Hz, CH2P).
(17) Preparation of 2: The crude bisphosphate methyl ester 18 (142
mg) was placed in a sealable tube under argon. t-Butylamine
(25 mL) was added, the tube was sealed and placed into an oil
bath at 70 °C for 3 days. The mixture was cooled to r.t., the
tube opened and most of the t-butylamine was removed with a
stream of nitrogen. The crude product was purified by re-
versed phase HPLC (Vidac C-4 column, gradient of 2.5 to
10% MeOH in 0.03 M NH4OAc). The main peak was collec-
ted and lyophilized three times to remove the buffer to provide
2 as a white solid (30mg, 33 µmol, 17% for two steps). 1H
NMR (DMSO, 300 MHz): 8.42 (s, 1H, I-H8), 8.22 (s, 1H) and
8.13 (s, 1H) A-H2 and A-H8, 7.26 (br.s, 2H, NH2), 5.89 (d,
J=4Hz, 1H) and 5.82 (d, J=4Hz, 1H) I-1’ and A-1’, 4.81 (t,
J=5Hz, 1H) and 4.58 (t, J=5Hz, 1H) A-2’ and I-2’), 4.25 (m,
1H) and 4.17 (m, 1H) A-3’ and I-3’, 2.96 (d, J=22Hz, 2H,
CH2P), 1.19 (s, 27H, t-BuN); MS: Negative ion FAB: 690 (M-
H)-1; daughter ions: 609 (-H2PO3), 478, 441, 423, 361, 343,
228, 211. MS Accurate mass calculated for C21H26N9O14P2
(M-H)-1: 690.1074; Found: 690.1056.
Acknowledgement
We thank the analytical department at Syntex for the measurement
of the physical properties, in particular Brenda Schwartz for the
high resolution mass determinations on the phosphate esters. We
also thank Lillian Lou and John Nakamura for the determination of
the inhibitory properties of the compounds reported in this paper.
References and Notes
(1) Current address: Wyeth-Ayerst Research, 401 N. Middletown
Rd., Pearl River, NY 10965, USA.
(2) Current address: Neurobiology Unit, Roche Bioscience, 3401
Hillview Ave., Palo Alto, CA 94304, USA.
(3) Christopherson, R.I.; Lyons, S.D. Med. Res. Rev. 1990, 10,
505.
(4) Eugui, E.M.; Allison, A.C. Ann. N.Y. Acad. Sci. 1993, 685,
309.
(5) Lagerkvist, U. J. Biol. Chem. 1958, 233, 143.
(6) Abrams, R., Bentley, M. Arch. Biochem. Biophys. 1959, 79,
91.
(7) Fukuyama, T.T. J. Biol. Chem. 1966, 241, 4745.
(8) von der Saal, W.; Crysler, C.S.; Villafranca, J.J. Biochemistry,
1985, 24, 5343.
(9) Ti, G.S., Gaffney, B.L., Jones, R.A. J. Am. Chem. Soc. 1982,
104, 1316.
(10) Mautz, D.S., Davisson, V.J., Poulter, C.D. Tetrahedron Lett.
1989, 30, 7333.
(11) Gray, M.D.M., Smith, D.J.H. Tetrahedron Lett. 1980, 21, 859.
(12) Vrudhula, V.M., Kappler, F.A.C., Ginell, S.L., Lessinger, L.,
Hampton, A. J. Med. Chem. 1989, 32, 885.
(13) Campbell, D.A., Bermak, J.C., Burkoth, T.S., Patel, D.V. J.
Am. Chem. Soc. 1995, 117, 5381.
(14) Preparation of 13: The t-butylammonium salt 10 (500 mg,
0.56 mmol) was dissolved in CH2Cl2 (10 mL) and the mixture
was cooled in an ice/water bath. A precooled aqueous solution
of KHSO3 (5%, 10 mL) was added and the mixture was stirred
vigorously for 2 min at 0°C before being partitioned between
CH2Cl2 (200 mL) and water (200 mL). The organic phase was
separated, washed with water (200 mL), dried over MgSO4
and evaporated to about 20 mL under reduced pressure. Ben-
zene (100 mL) was added and the mixture was evaporated to
dryness under vacuum. The residue was dissolved in THF
(8 mL); triphenylphosphine (279 mg, 1.06 mmol), followed
by alcohol 8 (366 mg, 0.71 mmol) and finally di-isopropyla-
zodicarboxylate (0.21 mL, 216 mg, 1.07 mmol) was added un-
der stirring. The mixture was stirred at r.t. for 1 h, diluted with
ether (300 mL) and washed with dilute HCl (150 mL) follo-
wed by half-saturated NaHCO3 (150 mL). The organic phase
was dried over MgSO4 and evaporated to dryness under vacu-
um. The crude product was chromatographed twice on silica-
gel (40% acetone/60% hexane) to give 13 as a colorless oil
(18) For experimental details of the biological assay see: Nakamu-
ra, J., Lou, L. J. Biol. Chem. 1995, 270, 7347.
(19) Tesmer, J.J.G., Klem, T.J., Deras, M.L., Davisson, V.J.,
Smith, J.L. Nat. Struct. Biol. 1996, 3, 74.
Article Identifier:
1437-2096,E;1999,0,S1,0897,0900,ftx,en;W05099ST.pdf
Synlett 1999, S1, 897–900 ISSN 0936-5214 © Thieme Stuttgart · New York