Synthesis and evaluation of new 4-phospho-D-erythronic acid derivatives as competitive inhibitors of spinach ribose-5-phosphate isomerase

Article abstract of DOI:10.1016/j.tetlet.2003.11.038

This paper reports the synthesis of new 4-phospho-D-erythronic acid derivatives, namely 4-phospho-D-erythronohydroxamic acid (1), 4-phospho-D-erythronohydrazide (2), and 4-phospho-D-erythronamide (3), and their kinetic evaluation as new competitive inhibi

Full text of DOI:10.1016/j.tetlet.2003.11.038

Tetrahedron
Letters
Tetrahedron Letters 45 (2004) 753–756
Synthesis and evaluation of new 4-phospho-D-erythronic acid
derivatives as competitive inhibitors of spinach
ribose-5-phosphate isomerase
Emmanuel Burgos and Laurent Salmon*
Laboratoire de Chimie Bioorganique et Bioinorganique, CNRS-UMR 8124, Institut de Chimie Molꢀeculaire et des Matꢀeriaux
d’Orsay (ICMMO), B^atiment 420, Universitꢀe Paris-Sud XI, F-91405 Orsay, France
Received 5 September 2003; revised 5 November 2003; accepted 11 November 2003
Abstract—This paper reports the synthesis of new 4-phospho-
oxamic acid (1), 4-phospho- -erythronohydrazide (2), and 4-phospho-
competitive inhibitors of the isomerization reaction between -ribose 5-phosphate and
ribose-5-phosphate isomerase (RPI). By comparison to the only known RPI inhibitor, 4-phospho-
-arabinose, appears as a new potent high-energy
D
-erythronic acid derivatives, namely 4-phospho-
-erythronamide (3), and their kinetic evaluation as new
-ribulose 5-phosphate catalyzed by spinach
-erythronate (4, Ki ¼ 28 lM,
D-erythronohydr-
D
D
D
D
D
Km=Ki ¼ 270), the hydroxamic acid 1, obtained by an eight-step synthesis from
D
intermediate analogue inhibitor of the isomerization reaction (Ki ¼ 29 lM, Km=Ki ¼ 260).
Ó 2003 Elsevier Ltd. All rights reserved.
Ribose-5-phosphate isomerase (RPI, E.C. 5.3.1.6), an
aldose–ketose isomerase involved in the pentose phos-
phate pathway, catalyzes the isomerization reaction of
specific RPI inhibitors of therapeutic interest, e.g.
against the pathogenic human African trypanosomiasis
where the pentose phosphate pathway has been shown
to play a crucial role in the survival and development of
the parasite Trypanosoma brucei.9 As part of our pro-
gram dedicated to the study of aldose–ketose isomerases
and by analogy to our previous finding of strong GPI
inhibitors,10;11 we report in this study the synthesis of
D
-ribose 5-phosphate (R5P) and D-ribulose 5-phosphate
(Ru5P) (Fig. 1).1 The reaction is thought to proceed
through a proton transfer mechanism and to involve a
high-energy 1,2-cis-enediolate intermediate analogous to
that reported previously for triosephosphate isomerase
(TIM)2 and glucose-6-phosphate isomerase (GPI).3;4
During the last 3 years, site-directed mutagenesis5;6 and
X-ray diffraction studies6;7 gave new insights into the
nature and possible role of some of the active site resi-
dues. Nevertheless, unlike for TIM and GPI, little is
known about RPI. For example, only a limited number
of inhibitors have been reported in the literature,8 with
only one reaction intermediate analogue inhibitor
available to date.1 Consequently, it appeared important
to us to provide new, strong competitive inhibitors of
RPI to the scientific community, so that detailed kinetic
and crystallographic studies could be pursued. This
could then lead to the design of strong and species-
three original 4-phospho-
namely 4-phospho- -erythronohydroxamic acid 1,
4-phospho- -erythronohydrazide 2, and 4-phospho-D-
D-erythronic acid derivatives,
D
D
erythronamide 3 (Fig. 2), as new reaction intermediate
analogues of the R5P to Ru5P isomerization reaction
catalyzed by spinach RPI. A comparison of their
OH
CHO
OH
CH2OH
O
-O
RPI
RPI
OH
OH
OH
OH
OH
OH
CH2OPO3
2-
2-
2-
CH2OPO3
CH2OPO3
1,2-cis-enediolate
high-energy intermediate
Keywords: Phosphate sugars; Enzyme inhibitors; Hydroxamic acids;
Hydroxamates.
R5P
Ru5P
Figure 1. Isomerization reaction catalyzed by
isomerases.
D-ribose-5-phosphate
* Corresponding author. Tel.: +33-169-156311; fax: +33-169-157281;
e-mail: lasalmon@icmo.u-psud.fr
0040-4039/$ - see front matter Ó 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2003.11.038

754
E. Burgos, L. Salmon / Tetrahedron Letters 45 (2004) 753–756
OH
NH
NH
NH
2
O
O
O
NH
-O
O
2
OH
OH
OH
OH
OH
OH
OH
OH
2-
2-
2-
2-
CH OPO
CH OPO
CH OPO
CH OPO
2
3
2
3
2
3
2
3
1
2
3
4
Figure 2. Synthesized models of the 1,2-cis-enediolate high-energy intermediate species of the RPI-catalyzed isomerization reaction.
Scheme 1. Reagents and conditions: (a) i. Dowexâ 50X4-400 (Hþ form), MeOH, )20 °C, 15 min; ii. CH2N2, Et2O,)20 °C, 25%; (b) TrCl (1.1 equiv),
pyridine, DMAP (0.03 equiv), 45 °C, 12 h, 94%; (c) PhCOCl (2.1 equiv), pyridine, 24 h, 25 °C, 84%; (d) H2, 25 bar, Pd/C 10%, CH2Cl2/MeOH (4/1),
12 h, 100%; (e) (PhO)2POCl (1 equiv), pyridine, 12 h, 25 °C, 94%; (f) i. H2, 25 bar, PtO2, 3 days; ii. cyclohexylamine, recrystallization (MeOH/
n-hexane, 1/50), 70%; (g) 1 (X ¼ OH): i. solid NH2OH, MeOH, 12 h, 4 °C; ii. Dowexâ 50X4-400 (Naþ form); freeze-drying, 95%; 2 (X ¼ NH2):
i. NH2NH2ÆH2O, MeOH, 12 h, 4 °C; ii. Dowexâ 50X4-400 (Naþ form); freeze-drying, 95%; 3 (X ¼ H): i. NH3, MeOH, 12 h, 4 °C; ii. NH3–H2O,
MeOH, 12 h, 4 °C; iii. Ba(OH)2; iv. Dowexâ 50X4-400 (Hþ form); v. Dowexâ 50X4-400 (Naþ form); freeze-drying, 95%.
inhibitory properties with that of 4-phospho-
D
-erythro-
nate 4 (Fig. 2), the only known efficient inhibitor of the
ated to give compound 8 in 94% yield. Protection of the
secondary hydroxyl groups of 8 was achieved with
PhCOCl in pyridine. The resulting product 9 (84% yield)
was then hydrogenolyzed (25 bar) over 10% Pd/C to give
compound 10 (quantitative yield), which was next
phosphorylated with diphenyl phosphorochloridate in
pyridine to yield 11 (94%). Direct synthesis of the
inhibitors from 11 (or its dibenzylphosphate analogue)
led to complex, unexploitable mixtures, probably
because of the phosphoester reactivity with nucleo-
philes. The product was finally converted into the pre-
cursor 614 by both hydrogenation and hydrogenolysis
(25 bar) over PtO2 for 3 days. Following neutralization
with cyclohexylamine and crystallization (MeOH/n-
hexane, 1/50), pure 6 was obtained in 70% yield.
enzyme, is also reported.
Our synthetic strategy to obtain compounds 1, 2, and 3
started from potassium
steps to methyl 4-phospho-(bis-cyclohexylammonium)-
2,3-di-O-cyclohexanecarbonyl- -erythronate 6, the pre-
D
-erythronate 5,12 and led in six
D
cursor of the targeted compounds (Scheme 1). Reaction
of diazomethane with the acid form of 5 (Dowexâ 50X4-
400, Hþ form) in methanol at )20 °C led to a mixture13
of methyl
D-erythronate 7, D-erythronolactone, and
other methylated derivatives. Pure 7 was obtained by
silica gel chromatography (acetone/toluene 1:1,
Rf ¼ 0:16) and recrystallization (AcOEt) in 25% yield.
We did not succeed in purifying 7 by distillation because
it easily equilibrates to erythronolactone upon heating.
In order to prevent the lactone formation, 7 was trityl-
An interesting feature of this strategy lies in the fact that
nucleophilic substitution at C-1 of the precursor 6 and

E. Burgos, L. Salmon / Tetrahedron Letters 45 (2004) 753–756
755
Table 1. Inhibitory effect of 4-phospho-
D
-erythronohydroxamic acid 1, 4-phospho-
D
-erythronohydrazide 2, 4-phospho-
D
-erythronamide 3,
4-phospho-D-erythronate 4, D-erythronate, and D
-erythronohydroxamic acid on spinach ribose-5-phosphate isomerase (RPI)
Inhibitor
IC50 (mM)
Ki (mM)
Km=Kia
1
2
3
4
0.018 0.003
1.8 0.2
1.8 0.2
0.029 0.003
1.8 0.2
2.5 0.5
0.028 0.005
––
260
4
3
0.010 0.002
270
––
––
D
-Erythronate
-Erythronohydroxamic acid
24
32
2
3
D
––
a Km(R5P) ¼ 7.5 0.8 mM.
deprotection of the hydroxyl groups on C-2 and C-3 are
achieved in one-step. The three new molecules 1
(X ¼ OH), 2 (X ¼ NH2), and 3 (X ¼ H) depicted in
Scheme 1 were obtained by reaction in dry methanol
with, respectively, solid NH2OH15 (CAUTION, this very
volatile and toxic product must be handled in a fume
hood, wearing glovesꢀ), NH2NH2ꢀH2O and NH3 at 4 °C.
The initial fraction of 1, bis(hydroxylammonium) salt
was insoluble in the reaction mixture and could be col-
lected (30%). A second fraction (65%) further precipi-
tated from the mother liquor upon addition of diethyl
U/mL of RPI (and replots of apparent Km=Vmax values vs
inhibitor concentration). IC50 determinations (Table 1)
were achieved using a 3.2 mM R5P concentration.
Compounds 1–4 all behave as competitive inhibitors of
the isomerization reaction of R5P to Ru5P catalyzed by
spinach RPI. Comparison of the IC50 values of 1 and 4
with those of the respective nonphosphorylated ana-
logues confirms the primary importance of the phos-
phate group for achieving effective inhibition. Both the
hydrazide 2 and amide 3 neutral derivatives behave as
weak competitive inhibitors with Ki values close to the
Km value of the substrate R5P. The Ki value we obtained
for the anionic compound 4 (28 lM, Table 1), the
known strong competitive RPI inhibitor, is about 6-fold
higher than the reported value for the same enzyme
(4.4 lM),1 a difference likely due to the different condi-
tions used for the kinetic study (lit.1: 40 mM phosphate
buffer, pH 6.4). With a Ki value of 29 lM and a Km=Ki
ratio of 260, the hydroxamic acid 1, mostly neutral in
solution owing to its probable pKa of 9.5, appears to be
a new potent competitive inhibitor of the reaction cat-
alyzed by spinach RPI.21 However, although 4-phospho-
ether (attempted precipitation of
1 with barium
hydroxide led to precipitation of both barium salts of 1
and cyclohexanehydroxamate formed upon deprotec-
tion of the hydroxyl groups at C-2 and C-3, which could
not be separated). Following elution (H2O) on Dowexâ
50X4-400 (Naþ form) and freeze-drying, pure 1 (diso-
dium salt)16 was obtained in 95% overall yield from 6.
Compound 2, precipitated as the bis(hydrazonium) salt
from the reaction mixture and was rinsed with metha-
nol, eluted (H2O) on Dowexâ 50X4-400 (Naþ form) and
freeze dried to afford pure 2 as the disodium salt17 in
95% yield. Compound 3, as the bis(ammonium) salt, did
not precipitate from the reaction mixture. However,
upon addition of barium hydroxide, the barium salt of
3, could be collected upon precipitation. It was then
successively eluted (H2O) on Dowexâ 50X4-400 (Hþ
form) and Dowexâ 50X4-400 (Naþ form). Following
lyophilization, pure 3 as the disodium salt,18 was
recovered in 95% yield.
D
-erythronohydroxamic acid 1 is a much better struc-
tural mimic of the 1,2-cis-enediolate reaction interme-
diate than 4-phospho- -erythronate 4, both inhibitors
D
display about the same inhibition constant values. In
view of our results, we suggest the hypothesis that the
anionic character of the high-energy intermediate ana-
logue is a significant parameter in the design of strong
RPI inhibitors. In the case of the hydroxamic acid 1,
kinetic measurements at a higher pH value where the
enzyme is still active and the inhibitor is partially ionized
will have to be run in order to show whether or not the
ionized form of 1 is the true inhibitor of the enzyme. In
addition, a study of Ki as a function of the inhibitorꢁs
pKa (from a series of fluoro analogues of 1 for instance)
would be valuable for addressing this question. By
comparison to our reported inhibition studies on GPI
achieved in about the same conditions, the hydroxamic
The new compounds 1, 2, and 3, the only known potent
RPI inhibitor 4,1 as well as the nonphosphorylated
analogues of 1 and 4, respectively D-erythronohydrox-
amic acid19 and
D
-erythronate, were all evaluated
against spinach RPI (Aldrich) as potential inhibitors of
the R5P to Ru5P isomerization reaction. Enzymatic
activities were determined by following the change in
ultraviolet absorbance that accompanies conversion of
R5P to Ru5P,20 (k ¼ 282 nm, e ¼ 58:6 Mꢁ1 cmꢁ1) at
25 °C in 50 mM TRIS.HCl buffer (pH 7.5). Apparent
acid higher homologue of 1, 5-phospho-D-arabi-
nonohydroxamic acid, proved to be a stronger inhibitor
by an order of magnitude than the carboxylate higher
€
Michaelis constants (Km) and inhibition constants (Ki)
were determined (Table 1) from double reciprocal plots
of the initial reaction velocity versus R5P concentration
obtained at various concentrations of inhibitors
(Lineweaver–Burk graphical representation) with 0.5
homologue of 4, 5-phospho-
higher Km=Ki ratios were reported for the GPI inhibitors
5-phospho- -arabinonohydrazide and 5-phospho-D-
D
-arabinonate.10 Also,
D
arabinonamide than for the corresponding RPI inhibi-
tors 2 and 3.11 As previously reported in the case of TIM
and GPI, and in view of the recently reported 3D high-
resolved structures of RPI6;7 and GPI,22;23 our results are
in accordance with the fact that RPI and GPI do not
share, at least in part, an identical architecture of
ꢀ
Because hydroxamic acids are known to readily form stable metal
complexes, a glass spatula was used to handle solid NH2OH, and
thereafter 1.

756
E. Burgos, L. Salmon / Tetrahedron Letters 45 (2004) 753–756
2.39–1.20 (m, 50H). 13C NMR (50 MHz, CD3OD)
d ¼ 176:2 (C-1), 169.2 (CO, C0O), 73.2 (d, C-3,
J ¼ 9:2 Hz), 72.3 (C-2), 63.2 (d, C-4, J ¼ 4:6 Hz), 53.0
(OMe), 51.2 (Ca CHA), 44.2–43.9 (Ca CH, Ca0 CH), 32.9
(Cb CHA), 30.3–29.8 (Cb CH, Cb0 CH, Cc CH, Cc0 CH),
26.9–26.5 (Cd CH, Cd0 CH), 26.2 (Cd CHA), 25.6 (Cc
CHA). 31P NMR (CDCl3) d ¼ ꢁ0:85. MS (negative-ion
electrospray): m=z (%) 321 [H2PO4] (29), 449 [M+Hþ]
(100). Anal. Calcd for C31H57O10PN2ÆH2O (666.8): C,
55.84; H, 8.92; N, 4.20; P, 4.65. Found C, 56.11; H, 8.93;
N, 4.17; P, 4.60.
the active site. Thus, the synthesis of a new potent
competitive inhibitor of the RPI-catalyzed isomerization
reaction such as 4-phospho-D-erythronohydroxamic
acid 1 appears to be very promising in order to further
develop kinetic, mechanistic, and structural investiga-
tions on ribose-5-phosphate isomerases.
Acknowledgements
15. Brauer, G. Handbook of Preparative Inorganic Chemistry,
2nd ed.; Academic: NY, 1963; Vol. I, p 504.
ꢁ
E.B. is grateful to the Ministere de lꢁEducation Natio-
nale et de la Recherche for a grant (2001–2004).
28
16. Selected data for compound 1: ½aꢂ )14.1° (c 1.17, H2O).
D
IR mmax (cmꢁ1) (KBr): 3420 (OH), 1653 (CO), 1086 (OH,
PO). 1H NMR (250 MHz, D2O) d ¼ 4.17–4.14 (m, 1H),
3.90–3.87 (m, 3H). 13C NMR (50 MHz, D2O) d ¼ 170:7
(C-1), 72.5 (d, C-3, J ¼ 6:1 Hz), 71.1 (C-2), 64.8 (d, C-4,
J ¼ 4:7 Hz). 31P NMR (D2O) d ¼ 7:58. MS (negative-ion
electrospray): m=z (%) 97 [H2PO4] (5), 230 [M+Hþ] (100),
252 [M+Naþ] (40). HRMS (negative-ion electrospray):
References and notes
1. Woodruff, W. W.; Wolfenden, R. J. Biol. Chem. 1979, 254,
5866–5867.
2. Rieder, S. V.; Rose, I. A. J. Biol. Chem. 1959, 234, 1007–
1010.
calcd for C4H9NO8P (M+Hþ) 230.0065, found 230.0066.
28
17. Selected data for compound 2: ½aꢂ )14.2° (c 1.75, H2O).
D
3. Bruns, F.; Okada, S. Nature 1956, 177, 87–88.
4. Topper, Y. J. J. Biol. Chem. 1957, 225, 419–423.
5. Jung, C.-H.; Hartman, F. C.; Lu, T.-Y.S.; Larimer, F. W.
Arch. Biochem. Biophys. 2000, 373, 409–417.
6. Ishikawa, K.; Matsui, I.; Payan, F.; Cambillau, C.; Ishida,
H.; Kawarabayasi, Y.; Kibuchi, H.; Roussel, A. Structure
2002, 10, 877–886.
7. Zhang, R.-G.; Andersson, C. E.; Savchenko, A.; Skarina,
T.; Evdokimova, E.; Beasley, S.; Arrowsmith, C. H.;
Edwards, A. M.; Joachimiak, A.; Mowbray, S. H.
Structure 2003, 11, 31–42.
8. Burger, A.; Tritsch, D.; Biellmann, J.-F. Carbohydr. Res.
2001, 332, 141–149.
9. Dardonville, C.; Rinaldi, E.; Hanau, S.; Barrett, M. P.;
Brun, R.; Gilbert, I. H. Bioorg. Med. Chem. 2003, 11,
3205–3214.
IR mmax (cmꢁ1) (KBr): 3387 (OH), 1674 (CO), 1091 (OH,
PO). 1H NMR (250 MHz, D2O) d ¼ 4.22–4.20 (m, 1H),
3.94–3.85 (m, 3H). 13C NMR (50 MHz, D2O) d ¼ 173:4
(C-1), 72.8 (d, C-3, J ¼ 6:4 Hz), 71.9 (C-2), 64.9 (d, C-4,
J ¼ 4:6 Hz). 31P NMR (D2O) d ¼ 7:59; MS (negative-ion
electrospray): m=z (%) 97 [H2PO4] (5), 229 [M+Hþ] (100),
251 [M+Naþ] (40). HRMS (negative-ion electrospray):
calcd for C4H10N2O7P (M+Hþ) 229.0225, found 229.0225.
18. Selected data for compound 3: IR mmax (cmꢁ1) (KBr): 3421
(OH), 1675 (CO), 1092 (OH, PO). 1H NMR (250 MHz,
D2O) d ¼ 4.23–4.20 (m, 1H), 3.91–3.89 (m, 3H). 13C NMR
(50 MHz, D2O) d ¼ 177:9 (C-1), 72.8 (d, C-3, J ¼ 7:4 Hz),
72.5 (C-2), 65.3 (d, C-4, J ¼ 4:6 Hz). 31P NMR (D2O)
d ¼ 6:45; MS (negative-ion electrospray): m=z (%) 97
[H2PO4] (4), 214 [M+Hþ] (100), 236 [M+Naþ] (17).
HRMS (negative-ion electrospray): calcd for C4H9NO7P
(M+Hþ) 214.0116, found 214.0114.
ꢀ
10. Hardre, R.; Bonnette, C.; Salmon, L.; Gaudemer, A.
Bioorg. Med. Chem. Lett. 1998, 8, 3435–3438.
ꢀ
19. Salmon, L.; Prost, E.; Merienne, C.; Hardre, R.; Morgant,
ꢀ
11. Hardre, R.; Salmon, L. Carbohydr. Res. 1999, 318, 110–
115.
G. Carbohydr. Res. 2001, 335, 195–204.
20. Knowles, F. C.; Pon, M. K.; Pon, N. G. Anal. Biochem.
1969, 29, 40–47.
21. Qualitative observation on 1 with RPI showed that the
enzyme did not hydrolyze the hydroxamic acid function
(positive Fe3þ test after 12 h incubation time).
22. Davies, C.; Muirhead, H. Proteins 2002, 50, 577–579.
12. Hardegger, E.; Kreis, K.; El Khadem, H. Helv. Chim. Acta
1951, 34, 2343–2348.
13. Barker, R.; Wold, F. J. Org. Chem. 1963, 28, 1847–1851.
28
14. Selected data for compound 6: ½aꢂ )3.9° (c 1.17, CDCl3).
D
IR mmax (cmꢁ1) (KBr): 3447 (NH3þ), 2934, 2854 (CH2), 1740
(CO), 1086, 1066 (OH, PO). H NMR (250 MHz, CDCl3)
1
ꢀ
23. Arsenieva, D.; Hardre, R.; Salmon, L.; Jeffrey, C. J. Proc.
Natl. Acad. Sci. U.S.A. 2002, 99, 5872–5877.
d ¼ 5.36–5.30 (m, 2H), 3.90–3.80 (m, 2H), 3.71 (s, 3H),