Selective inhibition of Trypanosoma brucei GAPDH by 1,3-bisphospho-D-glyceric acid (1,3-diPG) analogues

Article abstract of DOI:10.1016/S0968-0896(00)00295-9

Various phosphono-phosphates and diphophonates were synthesized as 1,3-diphosphoglycerate 1,3-diPG) analogues by using a β-ketophosphonate, an α-fluoro,β-ketophosphonate or a β-ketophosphoramidate to mimic the unstable carboxyphosphate part of the natural

Full text of DOI:10.1016/S0968-0896(00)00295-9

Bioorganic & Medicinal Chemistry 9 (2001) 773±783
Selective Inhibition of Trypanosoma Brucei GAPDH by
1,3-Bisphospho-D-glyceric Acid (1,3-diPG) Analogues
Sylvain Ladame, Michel Bardet, Jacques Perie and Michele WillsonÃ
Â
Groupe de Chimie Organique Biologique, LSPCMIB, UMR CNRS 5068, Universite Paul Sabatier,
118 route de Narbonne, 31062 Toulouse cedex 4, France
Received 10 August 2000; accepted 2 November 2000
AbstractÐVarious phosphono-phosphates and diphosphonates were synthesized as 1,3-diphosphoglycerate (1,3-diPG) analogues by
usinga b-ketophosphonate, an a-¯uoro,b-ketophosphonate or a b-ketophosphoramidate to mimic the unstable carboxyphosphate
part of the natural substrate. The inhibitory e€ect of these analogues on glyceraldehyde-3-phosphate dehydrogenases (GAPDH) from
Trypanosoma brucei (Tb) and rabbit muscle were measured with respect to both substrates, glyceraldehyde-3-phosphate (GAP) and
1,3-diPG. Interestingly, all 1,5-diphosphono,2-oxopentanes without substitution at the C-3 position selectively inhibit the Tb
GAPDH with respect to 1,3-diPG and are without e€ect on Rm GAPDH. All 1-phospho,3-oxo,4-phosphonobutanes show them-
selves to be non-selective inhibitors either with regard to substrates or organisms, but they will be of a great interest as 1,3-diPG
stable models for structural studies of co-crystals with GAPDHs. # 2001 Elsevier Science Ltd. All rights reserved.
Introduction
potent and selective inhibitors of the GAPDH from para-
sites described to date are adenosine analogues.7 How-
ever, as the GAPDH natural substrate 1,3-diPG is able to
cross the glycosomal membrane,9 selective inhibitor
analogues of this substrate will probably o€er the
advantage of an easy uptake by this parasite's micro-
body. Their ®rst uptake into cytosol across the plasma
membrane should probably become possible from lipo-
Trypanosomes are ¯agellated protozoa responsible for
serious parasitic diseases in humans (sleepingsickness),
domestic animals, ®sh and plants in subtropical and
temperate regions. Today, the trypanosomiases repre-
sent a formidable medicinal and economic obstacle to
development in African and South American countries
and rank amongthe ®rst tropical diseases selected by
the World Health Organization to develop new or more
e€ective treatments.1 Owingto toxicity and lack of e-
cacy, most of the chemotherapies currently used remain
unsatisfactory and the design of novel classes of anti-
trypanosomatid drugs has become urgent.
10
philic prodrugphosphate analogues.
In order to design selective and speci®c inhibitors for T.
brucei GAPDH, we developed two families of 1,3-diPG
substrate analogues: 1,5-diphosphonate-2-oxopentane
and 1,4-phosphonophosphate-2-oxobutane (Schemes 2±
5).
Opperdoes et al.2,3 have shown that, due to the obligatory
dependence of the bloodstream form of Trypanosoma
brucei on glycolysis for ATP production, the glycolytic
enzymes are good targets for drug design. The glycosomal
glyceraldehyde-3-phosphate-dehydrogenase (GAPDH)4,5
has previously been considered as an important target
for the design of trypanocidal drugs.6À8 This enzyme
reversibly catalyses the oxidative phosphorylation of
d-glyceraldehyde-3-phosphate (GAP) into 1,3-dipho-
spho-d-glyceric acid (1,3-diPG) in the presence of
NAD+ and inorganic phosphate (Scheme 1). The most
In a ®rst step, inhibitor structures were selected to pro-
vide the closest homology with the substrate 1,3-diphos-
phoglycerate, keeping the overall size, the diphosphoryl
moieties, the carbonyl at the C-3 position and the (R)
con®guration for carbon C-2 bearing the hydroxyl group.
Indeed, owingto the low stability of the mixed anhydride
part in 1,3-diPG (t1=2 ˆ 30 s),11 this moiety was replaced
by a b-ketophosphonate structure.
In a second step, from results obtained in inhibition stud-
ies of GAPDHs with these compounds, all structural
modi®cations were carried out on the b-ketophosphonate
moiety to lead to a set of selective inhibitors having
ÃCorrespondingauthor. Fax: +33-560-55-60-11; e-mail: wilson@iris.
ups-tlse.fr
0968-0896/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(00)00295-9

774
S. Ladame et al. / Bioorg. Med. Chem. 9 (2001) 773±783
improved anities for the enzyme from T. brucei. We
report here their synthesis and evaluation as GAPDH
inhibitors from Trypanosoma brucei and from rabbit
muscle (chosen as the mammalian model).
We will distinguish (i) non-selective inhibitors (such as
the original phosphono-phosphate 1e (Scheme 2), 1,3-
diPG isosteric analogue) that may, by co-crystallization
with GAPDH, provide valuable information on these
Scheme 1. Reversible reaction catalyzed by GAPDH.
Scheme. 2. Reagents and conditions: (i) NaIO4, H2O, CH2Cl2, rt (95%); (ii) Br2, NaHCO3, MeOH/H2O (9/1), rt (80%); (iii) CH3P(O)(OBn)2, BuLi, THF,
À80 ꢀC (90%); (iv) HCl 1 N, THF/H2O, rt (40%); (v) P(OBn)3, I2, pyridine, CH2Cl2, À30 ꢀC (50%); (vi) Pd/C, H2, MeOH, cyclohexylamine, rt (100%).
Scheme 3. Reagents and conditions: (i) NaOH, MeOH (80%); (ii) Me3SiCl, pyridine, CH2Cl2 (90%); (iii) (BnO)2P(O)CH3, BuLi, THF, À80 ꢀC;
(iv) HCl, pH=1 (75%); (v) P(OBn)3, I2, pyridine, CH2Cl2, 0 ꢀC (75%); (vi) H2, Pd/C, MeOH (100%); (vii) cyclohexylamine.

S. Ladame et al. / Bioorg. Med. Chem. 9 (2001) 773±783
775
Scheme 4. Reagents and conditions: (i) NH2OBn, EtOH, 60 ꢀC, 4 days (55%); (ii) NaOH, H2O (98%); (iii) N-(3-dimethylaminopropyl)-N0-ethyl-
carbodiimide hydrochloride, THF/H2O, pH=5 (86%); (iv) H2, Pd/C, MeOH (98%); (v) Me3SiBr, NaOH (97%).
substrate's bindingsites, and (ii) the ®rst 1,5-diphos-
phonate-2-ketopentanes which selectively inhibit para-
site GAPDH by competition with the natural substrate
1,3-diPG.
approximately 20% and the products were then sepa-
rated. The next step was the monophosphorylation of
diol 1c. This was done without protection of the sec-
ondary alcohol group, using the tribenzylphosphite-
iodine couple as reagent.
Compound 2e (Scheme 3) is derived from methyl-3-
hydroxypropanoate resultingfrom the base-catalysed
ring-opening of b-propiolactone that proceeds mainly
by nucleophilic attack of methanol on the carbonyl
group.15 After a trimethylsilyl protection of the primary
hydroxyl group, the phosphonylation was obtained by
nucleophilic attack on the ester group of a carbanion
species, as described for 1b. The phosphorylation of 2c was
performed usingthe same procedure as for compound 1d,
namely P(OBn)3/I2. As the obtained phosphono-phos-
phate 2d was very sensitive to acidic conditions (puri-
®cation on silica gel led, by phosphate elimination, to 3,4-
ene-2-oxobutanephosphonate), the puri®cation was per-
formed by HPLC on a C18 silica gel column. Deprotec-
tion, in neutral conditions, of the four benzyl groups by
catalytic hydrogenation led to 2e as a tetracyclohexyl-
ammonium salt.
Results and Discussion
Phosphono-phosphates 1e and 2e (Schemes 2 and 3)
have as precursors R-methylglycerate and methyl-4-
hydroxypropanoate, respectively. Phosphonate and
phosphate groups were successively introduced on the
triose frame usingdibenzylmethylphosphonate, and
subsequently tribenzylphosphite with iodine. These
phosphorylatingaegnts were prepared by previously
described synthetic procedures.12,13
Concerningcompound 1 (Scheme 2), oxidative cleavage
of 1,2,5,6,-di-O-isopropylidene-d-mannitol by sodium
periodate gave two equivalents of 2,3-di-O-isopropyli-
dene-d-glyceraldehyde14 with an R con®guration at the
C-2 position, as in the GAPDH natural substrate. The
unstable aldehyde was immediately transformed into a
carboxylate group by oxidative reaction of the couple
Br2/H2O, followed by an in situ esteri®cation with
methanol in neutral conditions. Compound 1a was ®rstly
phosphorylated by condensation of the dibenzylmethyl-
phosphonate carbanion on the methyl ester. Deprotec-
tion of the diol by acidic hydrolysis of the isopropylidene
bridge without hydrolyzing the di-O-benzyl groups was
carried out in successive steps: the reaction, followed by
31P NMR analysis, was stopped as soon as the con-
centration of monobenzylphosphonate increased to
In order to maintain an OH group at the C-2 position as
in the natural substrate, the hydroxylamide 3d was syn-
thesized (Scheme 4). It was obtained by formation of an
amidic bond between dimethyl phosphono-acetic acid 3b
and diethylphosphonoethyl-(O-benzyl)amidoxime. The
phosphorylated intermediate 3a, arose from an addition
reaction of benzylhydroxylamine on diethylvinylphos-
phonate.16 Total deprotection of the di€erent hydroxyl
groups was successively carried out in two steps by (i)

776
S. Ladame et al. / Bioorg. Med. Chem. 9 (2001) 773±783
Scheme 5. Reagents and conditions: (i) (EtO)2P(O)CH3, BuLi, THF, À80 ꢀC (77%); (ii) (EtO)2P(O)CF2H, LDA, THF, À80 ꢀC (72%); (iii)
NaOH (95%); (iv) SOCl2, CH2Cl2 (100%); (v) (EtO)2P(O)CHFCOOH, BuLi, THF, À80 ꢀC (20%); (vi) (EtO)2P(O)NH2, Et3N, CH2Cl2 (25%);
(vii) (BnO)2P(O)CH3, BuLi, THF, À80 ꢀC (85%); (viii) H2, Pd/C, MeOH, cyclohexylamine (98%); (ix) Me3SiBr, NaOH (95±100%).
catalytic hydrogenation of N-O-benzyl, and then (ii)
acidic hydrolysis of the phosphoric tetrasilylester inter-
mediates resultingfrom the reaction with trimethyl-
bromosilane.17
di€erently protected phosphoryl groups. Catalytic
hydrogenation allowed the selective deprotection of the
benzyl groups borne by the phosphonate group at carbon
1; the anity of this compound 9b will be compared to
that of compound 5b. This will provide valuable infor-
mation on the importance of each phosphonate group
on the inhibitory e€ects.
For the synthesis of other compounds, the b-ketophos-
phonate moiety was modi®ed by introducing¯uorine or
nitrogen atoms to restore, to the maximum possible, the
pK and geometry of the mixed anhydride (O)C±O±P(O)
moiety of 1,3-diPG.
Finally, a-mono¯uoro-b-ketophosphonate 7b and keto-
phosphoramide 8b were synthesized from diethyl-phos-
phonobutyryl chloride: for 7b, by an acylation reaction
of a-¯uoro phosphonoacetic acid in the presence of
n-BuLi, followed by a decarboxylation reaction19 and for
8b, by an amidic couplingwith the commercially avail-
able diethyl phosphoramidate.
The diphosphonate analogues 5±9 (Scheme 5) were
synthesized from commercially available triethyl-4-
phosphonobutyrate by di€erent couplingreactions
dependingon the nature of the substituent at carbon
C-1. Compounds 5 and 6 were synthesized as previously
described18 by nucleophilic attack of the ester group by
the carbanion obtained from diethylmethyl (or di¯uo-
romethyl) phosphonate. For compound 9, the dibenzyl-
methylphosphonate carbanion was condensed on triethyl-
4-phosphonobutyrate leadingto a diphosphonate with
Inhibition Studies and Discussion
Activities of GAPDHs from trypanosome and rabbit
muscle were measured usingforward and reverse

S. Ladame et al. / Bioorg. Med. Chem. 9 (2001) 773±783
777
reactions of the enzymatic process (Scheme 1), as descri-
bed in the Experimental. The IC50 values, shown in Tables
1 and 2, were measured for both organisms, with respect
to both substrates in saturatingconcentrations as
described by Misset et al.:4,5
inhibitors only with respect to the 1,3-diPG. Conse-
quently, these di€erences indicate:
1. that phosphono-phosphates and diphosphonates
do not interact at the same bindingsites, and
2. that only phosphono-phosphates 1e and 2e interact
at the GAP bindingsite.
1. In the forward reaction, the concentration in GAP
was 0.8 mM, largely over the Km values (150 and
70 mM for Tb and Rm GAPDHs respectively);
2. In the reverse reaction, the concentration in 1,3-
diPG may appear lower than the Km values (100
and 130 mM for Tb and Rm GAPDHs, respec-
In addition, comparison of IC50 values between com-
pounds 1e, 3d, 4 and 5b clearly shows that the substitu-
tion at the C-3 position decreases anity and supresses
selectivity. Chloro-2-oxopentane-1,5-biphosphonic acid
20
tively), owingto the low equilibrium constant in
4
hindrance of the chlorine atom, bulkier than a hydroxyl.
22 had one of the lowest anities due probably to steric
the 3±5Â10À4 range. However, as shown by
Schmidt et al.21 substrates and products of the PGK
catalyzed reaction are strongly associated with the
enzyme, the correspondingequilibrium constant
being0.1. In the coupled PGK-GAPDH assay,
therefore, the corresponding1,3-diPG concentra-
tion is ten times higher than the part free in solution
and thus signi®cantly higher than the Km value (
8 Km). This bound 1,3-diPG is then transfered to
GAPDH either through a bienzyme complex, or
through a fast ko€ reaction from PGK.21
Since a stronghomology with the substrate (phosphate
group and C-2 substitution) seemed unfavourable to
selective activity, all 1,3-diPG analogues designed in the
second part (Table 2) were diphosphonates without
substitution at the C-3 position.
2-oxo-Diphosphonopentane 5b, the simplest of these 1,3-
diPG analogues, is a selective inhibitor of Tb GAPDH
with respect to 1,3-diPG but not the other substrate
GAP. No inhibition was observed on Rm GAPDH at a
5 mM inhibitor concentration; the same result was
reported in yeast.23 Surprisingly, 9b, an analogue of 5b
but bearinga phosphonic acid and a phosphonic ester,
has a very close IC50 value. It seems that the b-keto-
phosphonate moiety is the essential ionic frame respon-
sible for the selective inhibitory e€ect. To improve both
anity and selectivity, a series of chemical modi®ca-
tions of this b-ketophosphonate frame were performed
and the inhibitory e€ects are shown in Table 2.
Compound 1e, having the highest structural homology
with 1,3-diPG, proved to be a poor GAPDH inhibitor
(IC50=2 mM), totally non-selective either with regard
to substrates or with regard to organisms.
Compound 2e, without a hydroxyl group, was no more
selective with respect to substrates, but exhibited a
5-fold higher anity for the mammalian enzyme than
for the parasite enzyme.
Whereas phosphono-phosphates 1e and 2e are compe-
titive inhibitors with respect to both substrates, diphos-
phonates 3d and 4, substituted at the C-3 position, are
By introducingone or two ¯uorine atoms or one nitrogen
atom on the b-ketophosphonate moiety, the inhibitory
e€ect on Tb GAPDH was improved without altering
selectivity. The IC50 values for ¯uorinated derivatives 6b
and 7b are in the range of, or lower than, the Km of 1,3-
diPG (100±120 mM); moreover their active concentra-
tions are over-estimated since 6b and 7b (racemic com-
pound) exist under their hydrated forms at, respectively
10 and 20%, as evidenced by NMR spectroscopy.
Table 1. Inhibition studies of Tb and Rm GAPDH with respect to
GAP and with respect to 1,3-diPG: IC50 values (mM)
GAPDH Tb
/GAP /1,3-diPG
GAPDH Rm
/GAP /1,3-diPG
Compound
1e
2e
3d
4
2
2
1
0.5
1
0.35
2
2
The potential of a-mono¯uoro or a-di¯uorophosphonate
as stable phosphate mimics has been reported.18,24 They
could therefore act as alternative substrates of the gly-
colytic enzyme glycerol phosphate dehydrogenase.25
The introduction of one or two ¯uorine atoms onto the
methylene group increases the acidity of the phospho-
nate, from 7.6 to 6.5 g iving a pKa value identical to that of
the phosphate group. These substitutions also increase the
P±C±C angle value from 112 to 116ꢀ, restoringthe sub-
strate's P±O±C angle26 (Cambridge data bank). Concern-
ingthe b-ketophosphonate, these structural features are
less signi®cant since the P±X±C(O) ꢀangle is 113ꢀ for the
mixed anhydride (O)POC(O),27 116 for (O)PCC(O)28,29
and 125ꢀ for the amino analogue (O)PNC(O).30 X-ray 3D
structures are not available for a-halogenated-b-keto-
phosphonate. Since compounds 5b and 8b, despite their
di€erent P±X±C(O) angles (116 and 125ꢀ, respectively),
0.7
Ða
Ða
Ða
0.1
Ða
Ða
Ða
0.2
0.5
1.5
Ða
5b
aNo inhibition at 5 mM.
Table 2. Selective inhibition of Tb GAPDH by 1,3-diPG analoguesa
Compound
IC50 (mM)/1,3-diPGb
5b
9b
6b
7b
8b
350
300
65
150
200
aNo e€ect on the Rm enzyme at 5 mM concentration.
bNo inhibition with respect to GAP at 5 mM.

778
S. Ladame et al. / Bioorg. Med. Chem. 9 (2001) 773±783
have similar anities for the enzyme, it would seem that
the inhibitory e€ect is not directly linked to geometric
parameters.
benzophenone, dichloromethane (CH2Cl2) from P2O5
1
and pyridine or triethylamine from KOH. H, 13C and
31P NMR spectra were obtained on Bruker Instruments
AC-250, AC-50 and AC-81 usinginternal TMS ( 1H and
13C) and H3PO4 (31P) as references. Chemical shifts (d),
expressed in ppm, are reported for signal centers and
couplingconstants ( J) are given in Hz. IR spectra were
recorded on a Perkin-Elmer 1600 series FTIR and ele-
mental analyses were performed by the Ecole Nationale
Superieure de Chimie de ToulouseÐFrance. Optical
rotations were determined with a Perkin-Elmer 241
polarimeter. Preparative column chromatographies
were performed using70±230 mesh Merck silica egl.
HPLC was performed usinga hyperprep HS C18 col-
umn from Hypersil. The synthesis of compounds 1b, 1d,
2b, 2c, 2d, and of all diphosphonates from 5a to 9a were
performed under an argon atmosphere in dried glass-
ware usingfreshly distilled solvents.
However, the pKa values of these molecules18 show an
important di€erence between their ionisation state in
physiological conditions at pH 7.5. From pKa values of
6.09 for the b-keto-methylphosphonate 5b and 4.5 for the
di¯uoro-methylphosphonate 6b, an intermediate pKa
value for the mono¯uoro analogue 7b can be assumed. In
such cases all compounds are in poly-anionic form inside
the active site and the highest anity is observed for the
di¯uorinated compound 6b, almost entirely deproto-
nated. Since the percentage of mono-anionic form at
physiological pH is a little higher for compounds 5b and
7b than for 6b, it can be further assumed that the dif-
ference in pKa values could explain, at least partly, the
di€erent IC50 values for diphosphonates 5b to 7b. It is
very likely that an additive e€ect can be ascribed to the
restoration of an H-bondingcapability between the
¯uorine atoms and residues in the GAPDH active site.
These results clearly demonstrate:
General synthetic procedure
Methyl ester (2(R),3-O-isopropylidene) glyceric acid (1a).
To a solution of 1,2,5,6-di-O-isopropylidene-d-manni-
tol (10 g, 38.0 mmol) in CH2Cl2 (100 mL) sodium peri-
odate (16.2 g, 76 mmol) and water (4 mL) were slowly
added. After stirring for 3 h, magnesium sulphate (20 g)
was added and the mixture was ®ltered and evaporated.
To the colourless oil a solution of MeOH/H2O, 9:1
(100 mL), sodium hydrogenocarbonate (24 g, 285 mmol)
and bromine (7.7mL, 154 mmol) were slowly added. After
stirringovernight at rt, bromine excess was neutralized by
sodium thiosulphate and the mixture was extracted with
CH2Cl2 (2Â150 mL). The combined organic layers were
dried (MgSO4) and evaporated to give ester 1a as a col-
ourless oil (9.25 g, 76%).
1. for both enzymes, the bindingat the GAP site is
governed by the presence of the phosphate group
rather than the C-3 hydroxy group (compounds 1e
and 2e); replacement of the latter by an N±OH g roup
leads only to an anity at the 1,3-diPG bindingsite;
2. anities of the 3-(deoxy) compounds at the 1,3-
diPG bindingsite prove to be selective since only
observed for the Tb enzyme (compounds 5b to 8b
Table 2);
3. the anity for T. brucei GAPDH is optimized in
compounds where the carboxyphosphonate part
has a low pK (under its dianionic form in the active
site) and is able to make H-bonds with the protein.
Further rationalization and improvements of a-
nities will require the results of X-ray co-crystal-
lisation, now in progress with Tb GAPDH and
compounds 1e, 2e and 6b.
1H NMR (250 MHz, CDCl3) d 1.35 (s, 3H), 1.44 (s, 3H),
3.72 (s, 3H), 4.05 (m, 1H), 4.19 (m, 1H), 4.55 (m, 1H).
13C NMR (50 MHz, CDCl3) d 25.8, 26.3, 52.4, 67.2,
74.0, 111.3, 171.6.
Conclusion
IR (®lm): nCˆO=1755 cmÀ1
.
Diversely functionalized 1-phospho,3-oxo,4-phosphono-
butanes and 1,5-diphosphono,2-oxopentanes, 1,3-diPG
analogues, were synthesized and evaluated as GAPDH
inhibitors. Whereas phosphono-phosphates (1e and 2e)
and diphosphonates substituted at the C-3 position (3d
and 4) proved to inhibit Rm GAPDH as well as Tb
GAPDH, all non-substituted diphosphonates (from 5b
to 9b) are selective inhibitors of Tb GAPDH with
respect to 1,3-diPG. The introduction of ¯uorine or
nitrogen atoms on the a b-ketophosphonate improves
the anity of these inhibitors for the parasite enzyme
without alteringselectivity.
Dibenzyl (3(R),4-O-isopropylidene, 2-oxobutyl) phospho-
nate (1b). To an Ar-purged ¯ask containing a solution
of dibenzylmethylphosphonate (2.46 g, 8.9 mmol) in
THF (15 mL) a 1.6 M solution of butyl-lithium (6 mL,
9.8 mmol) was added at À78 ꢀC. After stirringfor
30 min at À78 ꢀC, this mixture was added, via a cannula,
to a solution of ester 1a (1.3 g, 8.1 mmol) in anhydrous
THF (15 mL). After stirringfor 1 h at À78ꢀC, the mixture
was slowly warmed overnight up to 20ꢀC. The organic
layer was subsequently neutralized with a saturated aqu-
eous solution of NH4Cl and after THF evaporation, the
residue was extracted with AcOEt (3Â100 mL). The
combined organic layers were dried (MgSO4) and ®l-
tered to give phosphonate 1b (2.95 g, 90%) used without
further puri®cation.
Experimental
Chemistry
General directions. Prior to use, tetrahydrofuran (THF)
and diethyl ether (Et2O) were distilled from sodium
1H NMR (250 MHz, CDCl3) d 1.21 (s, 3H), 1.28 (s, 3H),
3.97 (m, 2H), 4.40 (m, 1H), 5.08 (m, 4H), 7.36 (m, 10H).

S. Ladame et al. / Bioorg. Med. Chem. 9 (2001) 773±783
779
13C NMR (50 MHz, CDCl3) d 25.0, 26.0, 37.8 (JCÀP
=
0.56 mmol) in MeOH (70 mL), Pd/C 10% (230 mg) was
added. After stirringunder a H atmosphere for 30 min
at rt, cyclohexylamine (250 mL, 0.56 mmol) was further
added and the Pd/C ®ltered o€. Evaporation of MeOH
gave tetracyclohexylammonium salt 1e (370 mg, 100%)
as a white solid; [a]D=À4.6ꢀ (c 10, MeOH).
132.2), 65.9, 68.1 (JCÀP=6.1), 80.1 (JCÀP=2.4), 111.2,
127.8, 135.9 (JCÀP=6.2), 202 (JCÀP=6.6).
2
31P NMR (81 MHz, CDCl3) d 20.8.
IR (®lm): nCˆO=1724 cmÀ1; nPˆO=1013, 1214 cmÀ1
Anal. calcd for C28H62N4O9P2,4H2O (732): C, 45.95; H,
9.63; N, 7.65. Found C, 46.01; H, 9.46; N, 7.52.
Dibenzyl (3(R),4-dihydroxy, 2-oxobutyl) phosphonate (1c).
1 M HCl (5 mL) was added to a solution of phospho-
nate 1b (2.7 g, 6.7 mmol) in a mixture of THF/H2O 1:1
(150mL). After stirringfor 3 days at rt, THF was
evaporated o€ and the aqueous layer extracted with
CH2Cl2. Evaporation of the solvent and chromatography
(CH2 Cl2/MeOH, 95:5) yielded 1c (1 g, 40%).
1H NMR (250 MHz, D2O) d 1.7 (m, 40H), 3.1 (m, 4H),
4.1 (m, 2H), 4.5 (m, 1H).
13C NMR (50 MHz, D2O) d 26.4, 26.9, 33.0, 47.5 (JCÀP
=
100.5), 52.8, 67.9 (JCÀP=4.4), 79.7 (JCÀP=7.3), 213.4
(JCÀP=5.4 Hz).
1H NMR (250 MHz, CDCl3) d 3.35 (d, 2H, JHÀP=22.4),
3.85 (m, 2H), 4.23 (m, 1H), 5.0 (m, 4H), 7.32 (m, 10H).
31P NMR (81 MHz, D2O) d 4.6, 10.4.
13C NMR (50 MHz, CDCl3) d 38.6 (JCÀP=130.2), 63.5,
68.5 (JCÀP=6.3), 78.6 (JCÀP=1.8), 128.2, 128.8, 135.6
(JCÀP=6.1), 204 (JCÀP=6.4 Hz).
IR (H2O): nCˆO=1701 cmÀ1; nPˆO=1110 cmÀ1
IR (KBr dispersion):
1057 cmÀ1
n
CˆO=1701 cmÀ1
;
nPˆO=
.
31P NMR (81 MHz, CDCl3) d 21.6.
Methyl ester (3-hydroxy) propionic acid (2a). To a solu-
tion of NaOH (140 mg, 3.47 mmol) in anhydrous
MeOH (17 mL) cooled in an ice bath b-propiolactone
(5 g, 69.4 mmol) was slowly added. After stirring for 1 h
at rt, the solution was neutralized with concentrated
chlorhydric acid and the solvent evaporated. The resi-
due was diluted in CH2Cl2 and washed with water. The
organic layer was dried (MgSO4), ®ltered and con-
centrated to give a colourless oil (5.77 g, 80%).
IR (®lm): nCˆO=1723 cmÀ1; nPˆO=1019, 1216 cmÀ1
.
Anal calcd for C18H21O6P (364): C, 59.34; H, 5.81.
Found: C, 59.40; H, 5.75.
Dibenzyl (4-dibenzylphosphono, 3-oxo, 2(R)-hydroxybutyl)
phosphate (1d). Iodine (390 mg, 1.5 mmol) was added at
0 ꢀC to a solution of tribenzylphosphite (560 mg, 1.6
mmol) in anhydrous CH2Cl2 (15 mL). After stirringfor
10 min at 0 ꢀC and for 15 min at rt, this mixture was
added, via a cannula at À30 ꢀC, to a solution of diol 1c
(370 mg, 1.0 mmol) in anhydrous CH2Cl2 (20 mL) and
pyridine (0.13 mL, 1.5 mmol). After 30 min at rt, the
pyridinium salts were ®ltered o€ and the organic solution
washed with water, dried (MgSO4) and ®ltered. Evap-
oration of the solvent and chromatography (CH2Cl2/
MeOH, 98:2) yielded 1d (315 mg, 50%) as a colourless
oil; [a]D=+1.8ꢀ (c 10, MeOH).
1H NMR (250 MHz, CDCl3) d 2.54 (t, 2H), 2.85 (s, 1H,
OH), 3.68 (s, 3H), 3.84 (t, 2H).
13C NMR (50 MHz, CDCl3) d 36.7, 51.8, 58.1, 173.3.
IR (®lm): nCˆO=1730 cmÀ1
.
Methyl ester (3-hydroxy-trimethylsilyl) propionic acid
(2b). To a solution of 2a (2 g, 19.3 mmol) in anhydrous
CH2Cl2, anhydrous pyridine (1.55 mL, 19.3 mmol) and
chlorotrimethylsilane (2.45 mL, 19.3 mmol) were succes-
sively added at 0 ꢀC. After stirringfor 1 h at 0 ꢀC, the
solution was allowed to warm at rt and the solvent evap-
orated o€. Anhydrous Et2O was then added and the pyr-
idinium salts ®ltered o€. After evaporation, compound 2b
was obtained as a colourless liquid. (3.05 g, 90%).
1H NMR (250 MHz, CDCl3) d 3.34 (d, 2H, J=23.3),
4.26 (m, 3H), 5.02 (dd, 8H, JHÀH=1.6, JHÀP=9.7), 7.32
(m, 20H).
13C NMR (50 MHz, CDCl3) d 39.1 (JCÀP=128.8), 68.4
(JCÀP=6.2), 69.6 (JCÀP=6.0), 76.5 (JCÀP=6.3), 128.5,
135.6 (JCÀP=6.2), 202.4 (JCÀP=6.3).
1H NMR (250 MHz, CDCl3) d 0.05 (s, 9H), 2.47 (t, 2H),
3.61 (s, 3H), 3.80 (t, 2H).
31P NMR (81 MHz, CDCl3) d À0.6, 21.0.
IR (®lm): nCˆO=1725 cmÀ1; nOH=3410 cmÀ1; nPˆO
=
13C NMR (50 MHz, CDCl3) d À0.66, 37.6, 51.5, 58.3,
1015, 1256 cmÀ1
.
172.1.
Anal. calcd for C32H34O9P2 (624): C, 61.54; H, 5.49.
Found: C, 61.43; H, 5.55.
Dibenzyl (4-hydroxy, 2-oxobutyl) phosphonate (2c). To a
solution of dibenzylmethylphosphonate (3 g, 10.87 mmol),
in anhydrous THF (15 mL) at À80 ꢀC, a 1.6 M solution
of BuLi (7.5 mL, 12 mmol) was added. After stirringfor
half an hour at À78 ꢀC, this solution was added, via a
cannula, to a solution of ester 2b (1.9 g, 10.87 mmol) in
Dicyclohexylammonium (4-dicyclohexylammoniumphos-
phono, 3-oxo, 2(R)-hydroxybutyl) phosphate (1e). To a
H2-purged ¯ask containing a solution of 1d (350 mg,

780
S. Ladame et al. / Bioorg. Med. Chem. 9 (2001) 773±783
anhydrous THF (20 mL) at À78 ꢀC. After 1 h at À78 ꢀC,
the mixture was slowly warmed overnight up to 20 ꢀC.
The organic layer was then neutralized with a saturated
aqueous solution of NH4Cl and acidi®ed until pH=1
with a 1 M HCl solution. After THF evaporation, the
mixture was extracted with AcOEt (3Â100 mL). The
combined organic layers were then dried on MgSO4,
®ltered, and evaporated to give 3.9 g of a pale yellow oil.
Puri®cation by chromatography (CH2Cl2/MeOH, 98:2)
yielded 2c as a colourless oil (2.84 g, 75%).
Diethyl (ethyl, 2-O-benzylhydroxylamino) phosphonate
(3a). To a solution of O-benzyl-hydroxylamine (2.6 g,
21.1 mmol) in anhydrous methanol (25 mL), diethyl-
vinylphosphonate (1.1 mL, 7.1 mmol) was added. After
stirringfor 4 days at 60 ꢀC, CH2Cl2 (200 mL) was
added, and the mixture washed with water (40 mL). The
organic layer was dried (MgSO4), ®ltered and evapor-
ated. After puri®cation by chromatography (AcOEt), 3a
was obtained as a colourless oil (1.12 g, 55%)
1H NMR (250 MHz, CDCl3) d 1.26 (td, 6H), 1.98 (td,
2H, JHÀP=18.3), 3.14 (m, 2H), 4.04 (qd, 4H), 4.65 (s,
2H), 7.29 (m, 5H).
1H NMR (250 MHz, CDCl3) d 2.73 (t, 2H), 3.37 (d, 2H,
JHÀP=22.3), 3.51 (s, 1H, OH), 3.77 (t, 2H), 4.94±5.06
(m, 4H), 7.29 (m, 10H).
13C NMR (50 MHz, CDCl3) d 16.4 (d, JCÀP=6), 24.0 (d,
JCÀP=139.6), 45.8 (d, JCÀP=2.8), 61.6 (d, JCÀP=6.3),
76.1, 128.1, 137.8.
13C NMR (50 MHz, CDCl3) d 42.9 (d, JPÀC=128.3), 46.6,
57.5, 68.1 (d, JPÀC=6.3), 128.4, 135.6, 202.1 (d, JPÀC=6).
31P NMR (81 MHz, CDCl3) d 21.2.
31P NMR (81 MHz, CDCl3) d 30.3.
Dibenzyl (4 - dibenzylphosphono, 3 - oxobutyl) phosphate
(2d). A quantity of iodine (1.08 g, 4.2 mmol) was added
at 0 ꢀC to a solution of tribenzylphosphite (1.66 g,
4.71 mmol) in anhydrous CH2Cl2 (15 mL). After stirring
for 10 min at 0 ꢀC and for 5 min at rt, this solution was
added, via a cannula, to a solution of alcohol 2c (820 mg,
2.35 mmol) and anhydrous pyridine (0.36 mL, 4.2 mmol)
in anhydrous CH2Cl2 (15 mL) at 0 ꢀC.
Dimethyl-phosphono-acetic acid (3b). To a solution of
NaOH (444 mg, 11.1 mmol) in water (20 mL), trimethyl-
phosphono-acetate (2.01 g, 11.1 mmol) was added. After
stirringfor 1 h, the aqueous layer was washed with
AcOEt (20 mL) and evaporated to give 3b as a white
powder. (2.07 g, 98%)
1H NMR (250 MHz, D2O) d 2.78 (dd, 2H, JHÀP=21.5),
3.75 (d, 3H, JHÀP=11.2), 3.76 (d, 3H, JHÀP=11.2).
ꢀ
After stirringfor 30 min at 0 C and for a further 30 min at
rt, the pyridinium salts were ®ltered o€, the organic layer
was washed with water, dried (MgSO4) and ®ltered. Pur-
i®cation by HPLC on C18 silica gel (CH3CN±H2O 60:40)
led to 2d as a colourless oil (2.14 g, 75%).
13C NMR (50 MHz, D2O) d 40.2 (d, JCÀP=121.4), 56.1
(d, JCÀP=6.5).
31P NMR (81 MHz, CDCl3) d 31.1.
1H NMR (250 MHz, CDCl3) d 2.83 (t, 2H), 3.04 (d, 2H,
JHÀP=22.7), 4.18 (td, 2H), 5.01 (m, 8H), 7.32 (m, 20H).
(N-ethyldiethylphosphonate, N-1-oxoethyldimethylphos-
phonate) O-benzylhydroxylamine (3c). To a solution of
3b (532mg, 2.8mmol) in a mixture THF±H2O 1/1 (60 mL),
3a (803 mg, 2.8 mmol) was added. The solution was
acidi®ed to pH=4 with 1 M HCl and N-(3-dimethyla-
minopropyl)-N0-ethylcarbodiimide hydrochloride (537
mg, 2.8 mmol) was added. After stirring overnight at rt,
THF was evaporated. Compound 3c was extracted with
AcOEt (3Â100 mL), washed with saturated aqueous
NaHCO3, dried on MgSO4, ®ltered and evaporated.
Compound 3c was obtained in a pure state, no further
puri®cation beingnecessary (1.05 g, 86%).
13C NMR (50 MHz, CDCl3) d 42.9 (d, JCÀP=117.6),
44.0 (d, JCÀP=2.5), 62.0 (d, JCÀP=5.4), 68.2 (d, JCÀP
=
6.4), 69.4 (d, JCÀP=5.5), 127.5±128.7, 135.7 (m), 198.3
(d, JCÀP=6.35).
31P NMR (81 MHz, CDCl3) d À1.13, 20.41.
Dicyclohexylammonium (4-dicyclohexylammoniumphos-
phono, 3-oxo) phosphate (2e). To an H2-purged ¯ask
containinga solution of 1d (114 mg, 0.19 mmol) in
MeOH (20 mL), Pd/C 10% (100 mg) was added. After
1H NMR (250 MHz, CDCl3) d 1.05 (td, 6H), 1.86 (m,
2H), 2.84 (d, 2H, JHÀP=21.8), 3.51 (d, 6H, JHÀP=7.4),
3.59 (m, 2H), 3.82 (qd, 4H), 4.68 (s, 2H), 7.13 (m,
5H).
stirringunder a H atmosphere for 30min at rt, cyclo-
2
hexylamine (81 mL, 0.75mmol) was added and Pd/C ®l-
tered o€. Evaporation of MeOH gave tetracyclohexyl-
ammonium salt 2e as a white amorphous powder.
(116mg, 97%).
13C NMR (50 MHz, CDCl3) d 16.1, 22.7 (d, JCÀP
=
139.1), 30.9 (d, JCÀP=136.6), 40.0, 52.9, 61.6, 76.5,
128.9, 133.8, 166.6.
13C NMR (50 MHz, D2O) d 26.6 (m), 31.4 (m), 33.0 (m),
41.7, 52.3(d, JCÀP=74.7), 52.8, 60.2 (d, JCÀP=12.4),
211.0 (d, JCÀP=5.3).
31P NMR (81 MHz, CDCl3) d 23.6, 27.6.
31P NMR (81 MHz, D2O) d 2.28, 10.17.
(N-ethyldiethylphosphonate, N-1-oxoethyldimethylphos-
phonate) hydroxylamine (3d). To an H2-purged ¯ask
containinga solution of 3c (270 mg, 0.62 mmol) in
MeOH (70 mL), 10% Pd/C (40 mg) was added. After
Anal. calcd for C28H62N4O8P2,5H2O (734): C, 45.76; H,
9.87; N, 7.62. Found C, 45.98; H, 8.36; N, 7.72.

S. Ladame et al. / Bioorg. Med. Chem. 9 (2001) 773±783
781
stirringfor 2 h at rt, Pd/C was ®ltered o€ and MeOH
evaporated to give 3d as a colourless oil (210 mg, 98%).
5a (534 mg, 1.5 mmol) under argon. After stirring for 1
night at rt, the volatiles were evaporated in vacuo and
the residue dissolved in Et2O. The bisphosphonic acid
was then extracted in water (2Â10 mL) and the aqueous
layer titrated to pH 7.1 with 1 M NaOH. Lyophilization
provided the title compound as a hygroscopic white
solid (477 mg, 95%).
1H NMR (250 MHz, CDCl3) d 1.20 (t, 6H), 2.00 (m,
2H), 3.10 (d, 2H, JHÀP=21.4), 3.67 (d, 6H, JHÀP=11.0),
3.79 (m, 2H), 3.98 (m, 4H).
13C NMR (50 MHz, CDCl3) d 16.2 (d, JCÀP=5.0), 21.5 (d,
JCÀP=93.7), 31.2 (d, JCÀP=134.9), 42.4, 53.2 (d, JCÀP
=
1H NMR (250 MHz, D2O) d 1.25±1.65 (m, 4H), 2.60 (t,
2H), 2.75 (2H, d, JHÀP=22.0).
5.7), 62.1 (d, JCÀP=5.7), 164.9.(d, JCÀP=6.5)
31P NMR (81 MHz, CDCl3) d 25.5, 29.6.
13C NMR (50 MHz, D2O) d 18.2, 27.7 (d, JCÀP=131.9),
44.1 (d, JCÀP=12.5), 47.1 (d, JCÀP=104.1), 210.0 (d,
JCÀP=15.9).
(N-ethyldihydroxyphosphonate, N-1-oxoethyldihydroxy-
phosphonate) hydroxylamine, tetra-sodium salt (3e). To
an Ar-purged ¯ask containing 3d (210 mg, 0.61 mmol),
bromotrimethylsilane (0.66 mL, 5 mmol) was added.
After stirringovernight at rt, bromotrimethylsilane and
ethyl-bromide excesses were evaporated, the residue
31P NMR (81 MHz, D2O) d 11.5, 25.7.
Tetraethyl 1,1-di¯uoro-2-oxopentane-1,5-bisphosphonate
(6a). To a solution of LDA (3.9mL, 2 M solution in hex-
ane) in anhydrous THF (2 mL) at À78ꢀC, a solution of
diethyl-di¯uoromethylphosphonate (1.45 g, 7.73mmol) in
THF (2 mL) was added. After stirringfor 30 min at
À78 ꢀC, a solution of triethyl 4-phosphonobutyrate
(1.5 g, 5.95 mmol) in THF (2 mL) was slowly added, via
a cannula, and the reaction mixture was maintained for
1 h at À78ꢀC. Subsequently, glacial AcOH (0.78 mL) and
a saturated solution of NH4Cl (15mL) were successively
added. The bisphophonate was extracted with AcOEt
(3Â100 mL), the combined organic layers were dried
(MgSO4), ®ltered and the solvent evaporated. Puri®cation
by ¯ash chromatography (CH2Cl2/MeOH, 97:3) led to
compound 6a as a colourless oil (1.69 g, 72%).
beingdiluted in Et O (15 mL) and the tetra-hydroxy-
2
diphosphonate extracted with water (2Â25 mL). After
neutralization until pH=7.5 with 1 M NaOH and lyo-
philization, 3e was obtained as a white powder (207 mg,
97%).
1H NMR (250 MHz, D2O) d 1.79±1.92 (m, 2H), 2.78 (d,
2H, JHÀP=19.4), 3.69±3.78 (m, 2H).
13C NMR (50 MHz, D2O) d 28.7 (d, JCÀP=128.8), 39.2
(d, JCÀP=110.2), 46.4, 172.5 (d, JCÀP=7.3).
31P NMR (81 MHz, D2O) d 15.3, 21.1.
Tetraethyl 2-oxopentane-1,5-bisphosphonate (5a). To a
solution of diethylmethylphosphonate (304 mg, 2 mmol),
in anhydrous THF (15 mL) at À78 ꢀC, a 1.6 M solution
of BuLi (1.25 mL) was added. After stirringfor half an
hour at À78 ꢀC, this solution was added, via a cannula,
to a solution of triethyl 4-phosphonobutyrate (505 mg,
2 mmol) in anhydrous THF (20 mL) at À78ꢀC. After 1 h
at À78ꢀC, the mixture was slowly warmed overnight up to
20ꢀC. The organic layer was then neutralized with a
saturated aqueous solution of NH4Cl and the mixture
extracted with AcOEt (3Â100 mL). The combined organic
layers were subsequently dried on MgSO4, ®ltered, and
the solvent evaporated o€ to give a yellow oil. The
bisphosphonate 5a was isolated by ¯ash chromatography
(CH2Cl2/MeOH, 97:3) as a colourless oil (548mg, 77%).
1H NMR (250 MHz, CDCl3) d 1.25 (t, 12H), 1.56±1.90
(m, 4H), 2.81 (t, 2H), 3.97 (m, 4H), 4.16 (m, 4H).
13C NMR (50 MHz, CDCl3) d 15.8, 16.3, 24.5 (d, JCÀP
=
141.8), 37.7 (d, JCÀP=15.3), 61.5 (d, JCÀP=6.4), 65.5 (d,
JCÀP=6.6), 112.9 (td, JCÀP=195.2, JCÀF=273.4), 198.4
(dt, JCÀP=14.1, JCÀF=23.4).
31P NMR (81 MHz, CDCl3) d 3.05 (t, JPÀF=97), 30.7.
1,1-Di¯uoro-2-oxopentane-1,5-bisphosphonic acid (6b).18
Bromotrimethylsilane (2 mL, 16 mmol) was added
dropwise to ester 6a (886 mg, 2.25 mmol) under argon.
After stirringfor 3 days at rt, the volatiles were evapor-
ated in vacuo and the residue dissolved in Et2O. The
bisphosphonic acid was then extracted in water
(2Â10 mL) and the aqueous layer titrated to pH 7.1 with
1 M NaOH. Lyophilization gave the title compound as
a hygroscopic white solid (816 mg, 98%).
1H NMR (250MHz, CDCl3) d 1.98 (t, 12H), 1.55±1.90 (m,
4H), 2.85 (t, 2H), 2.95 (d, 2H, JHÀP=22.8), 3.95 (q, 8H).
13C NMR (50 MHz, CDCl3) d 16.3, 16.5, 24.4 (d, JCÀP
=
141.3), 42.5 (d, JCÀP=142.1), 43.7, 61.5 (d, JCÀP=6.5),
62.5 (d, JCÀP=6.3), 201.0 (d, JCÀP=6.0).
1H NMR (250 MHz, D2O) d 1.45±1.85 (m, 4H), 2.90 (m,
2H)
31P NMR (81 MHz, CDCl3) d 19.8, 31.3.
13C NMR (50 MHz, D2O) d 19.3, 29.6 (d, JCÀP=132.1),
38.5 (m), 120.7 (td, JCÀP=158.7, JCÀF=269.5), 206.8 (m).
Mass spectrometry (DCI, NH3) m/z 395 [MH+], 412
[MNH+4 ].
31P NMR (81 MHz, D2O) d 2.7 (t, JPÀF=98), 25.7.
2-Oxopentane-1,5-bisphonic acid (5b).18 Bromotrimethyl-
silane (1.41 mL, 11.2 mmol) was added dropwise to ester
Tetraethyl 1,¯uoro-2-oxopentane-1,5-bisphosphonate (7a).
To a solution of diethyl,4-phosphonobutyric acid (2 g,

782
S. Ladame et al. / Bioorg. Med. Chem. 9 (2001) 773±783
8.9 mmol) in anhydrous CH2Cl2 (20 mL), SOCl2 (0.72
mL, 9.7 mmol) was added. After stirringfor 4 h at rt, the
solvent was evaporated o€ and the oily residue dis-
solved in anhydrous THF (20 mL).
ammonium salts were subsequently ®ltered o€, the
organic layer washed with water (10 mL), dried (MgSO4),
®ltered o€ and evaporated in vacuo. Puri®cation by ¯ash
chromatography (CH2Cl2/MeOH, 95:5) led to 8a (206
mg, 25%).
To a solution of diethyl,1-¯uoro-phosphonoacetic acid
(1.63 g, 7.6 mmol) in anhydrous THF (25 mL), a 1.6 M
solution of BuLi (10.4 mL, 16.7 mmol) was added at
À78 ꢀC. After stirringfor 30 min at À78 ꢀC, the solution
of acid chloride was added, via a cannula, to the carb-
anion. After 1 h at À78 ꢀC, the solution was allowed to
warm overnight up to rt, neutralized with a saturated
solution of NH4Cl and extracted with CH2Cl2
(2Â100 mL). The organic layers were dried (MgSO4),
®ltered and evaporated o€ to give a yellow oil. Puri-
®cation by chromatography (CH2Cl2/MeOH, 97:3) led
to 7a as a colourless oil (570 mg, 20%).
1H NMR (250 MHz, CDCl3) d 1.21±1.35 (m, 12H),
1.71±1.94 (m, 4H), 2.45 (t, 2H), 3.98±4.22 (m, 8H).
13C NMR (50 MHz, CDCl3) d 16.1 (d, JCÀP=6.8), 16.5
(d, JCÀP=6.0), 17.9 (d, JCÀP=4.5), 24.8 (d, JCÀP=141.2),
34.1 (d, JCÀP=16.1), 61.6 (d, JCÀP=6.5), 64.1 (d, JCÀP
=
5.8), 174 (d, JCÀP=3.3).
31P NMR (81 MHz, CDCl3) d À2.2, 31.3.
1-oxo,4-phosphonobutane phosphoramidate (8b). Bromo-
trimethylsilane (0.65 mL, 5.2 mmol) was added dropwise
to ester 8a (310 mg, 0.86 mmol) under argon. After stir-
ringfor 3 h at rt ( 31P NMR: À20.9, 21.9), the volatiles
were evaporated in vacuo and the residue dissolved in
acetone. Water (0.1 mL) was added and the solution
stirred for half an hour. Evaporation in vacuo of sol-
vents led to pure 8b (198 mg, 93%).
1H NMR (250 MHz, CDCl3) d 1.28 (t, 12H), 1.63±1.92
(m, 4H), 2.74 (t, 2H), 3.96±4.22 (m, 8H), 5.09 (dd, 1H,
JHÀP=14.1, JHÀF=47.8).
13C NMR (50MHz, CDCl3) d 15.8 (d, JCÀP=2.7), 16.3,
24.5 (d, JCÀP=141.6), 38.9 (d, JCÀP=15.1), 61.5 (d, JCÀP
6.5), 64.2 (d, JCÀP=5.2), 91.5 (dd, JCÀP=152.7, JCÀF
197.6), 202.0 (d, JCÀP=19.6).
=
=
1H NMR (250 MHz, D2O) d 1.48±1.79 (m, 4H), 2.31 (t,
2H).
31P NMR (81 MHz, CDCl3) d 10.2 (d, JPÀF=71), 31.1.
13C NMR (50 MHz, D2O) d 20.1 (d, JCÀP=3.9), 27.9 (d,
JCÀP=135.3), 36.4 (d, JCÀP=17.5), 180.1.
1,Fluoro-2-oxopentane-1,5-bisphosphonic acid (7b). Bro-
motrimethylsilane (1 mL, 8 mmol) was added dropwise
to ester 7a (300 mg, 0.80 mmol) under argon. After stir-
ringfor 3 days at rt, the volatiles were evaporated in
vacuo and the residue dissolved in Et2O. The bispho-
sphonic acid was then extracted in water (2Â15 mL) and
the aqueous layer titrated to pH 7.1 with 1 M NaOH.
Lyophilization supplied the title compound as a hygro-
scopic white solid (273 mg, 97%).
31P NMR (81 MHz, D2O) d 0.6, 31.6.
IR (®lm): nCˆO=1816.
Mass spectrometry (DCI, CH4) m/z 248 [MH+].
Dibenzyl(2-oxo,5-diethylphosphono) phosphonate 9a. To
a solution of dibenzylmethylphosphonate (1.1g, 4 mmol),
in anhydrous THF (25 mL) at À78ꢀC, a 1.6 M solution of
n-BuLi (2.5mL) was added. After stirringfor 30 min at
À78ꢀC, this solution was added, via a cannula, to a
solution of triethyl 4-phosphonobutyrate (1.0 g, 4 mmol)
in anhydrous THF (30 mL) at À78 ꢀC. After 1 h at
À78 ꢀC, the mixture was slowly warmed overnight up to
20 ꢀC. The organic layer was then neutralized with a
saturated aqueous solution of NH4Cl and the mixture
extracted with AcOEt (3Â150 mL). The combined organic
layers were then dried on MgSO4, ®ltered, and the solvent
evaporated o€ to give a yellow oil. The bisphosphonate 9a
was isolated by ¯ash chromatography (CH2Cl2/MeOH,
97:3) as a colourless oil (1.6g, 85%).
1H NMR (250 MHz, D2O) d 1.45±1.60 (m, 2H), 1.64±
1.80 (m, 2H), 2.62±2.85 (m, 2H), 5.18 (dd, JHÀP=13.9,
JHÀF=48.2).
13C NMR (50 MHz, D2O) d 19.8 (d, JCÀP=16.0), 29.8
(d, JCÀP=132.6), 42.1 (d, JCÀP=17.2), 65.3 (d, JCÀP
=
6.2), 99.5 (dd, JCÀP=125.8, JCÀF=188.0), 213.0 (d,
JCÀP=14.7).
31P NMR (81 MHz, D2O) d 5.9 (d, JPÀF=58.8), 25.9;
7.4 (d, JPÀF=63.6), 25.7.
19F NMR (188 MHz, D2O) d À123.2 (dd, JFÀP=60.1,
JFÀH=48.5), À126.5 (dd, JFÀP=64.0, JFÀH=48.3).
1H NMR (250 MHz, CDCl3) d 1.23 (t, 6H), 1.52±1.78
(m, 4H), 2.60 (t, 2H), 3.00 (d, 2H, JHÀP=22.7), 3.94±
4.06 (m, 4H), 4.88±5.05 (m, 4H), 7.28 (m, 10H).
Diethyl, 1-oxo,4-diethylphosphonobutane phosphoramidate
(8a). To a solution of diethyl, 4-phosphonobutyric acid
(530 mg, 2.36 mmol) in anhydrous Et2O (20 mL), oxalyl
chloride (0.4 mL, 3.06 mmol) was added. After re¯uxing
for 1 night, the solution was neutralized with anhydrous
triethylamine and a solution of diethylphosphoramidate
(350 mg, 2.30 mmol) in anhydrous Et2O (20 mL) was
added. Anhydrous Et3N (0.32 mL, 2.3 mmol) was then
added, the solution re¯uxingfor 4 days. The triethyl-
13C NMR (50 MHz, CDCl3) d 16.4 (d, JCÀP=6.9), 16.5,
24.4 (d, JCÀP=141.1), 42.6 (d, JCÀP=128.2), 43.9 (d,
JCÀP=14.8), 61.4 (d, JCÀP=6.5), 68.0 (d, JCÀP=6.3),
128.1, 128.6, 135.7 (d, JCÀP=6.2), 200.6 (d, JCÀP=5.6).
31P NMR (81 MHz, CDCl3) d 20.9, 31.3.

S. Ladame et al. / Bioorg. Med. Chem. 9 (2001) 773±783
783
Acknowledgements
Dicyclohexylammonium (2-oxo,5-diethylphosphono) phos-
phonate 9b. To an H2-purged ¯ask containing a solu-
tion of 9a (1.0 g, 2.1 mmol) in MeOH (30 mL), Pd/C
10% (100 mg) was added. After stirring under an H2
atmosphere for 30 min at rt, cyclohexylamine (450 mL,
4.2 mmol) was added and Pd/C ®ltered o€. Evaporation
of MeOH gave tetracyclohexylammonium salt 2e as a
white amorphous powder (1.0 g, 98%).
We are thankful to P. A. M. Michels and V. Hannaert
for providingthe Trypanosoma brucei GAPDH over-
expression system. We also acknowledge the European
Science Research and Development Commission (IC 18
CT 970220) for its ®nancial support.
13C NMR (50 MHz, CD3OD) d 16.8 (d, JCÀP=6.0), 17.6
(d, JCÀP=4.5), 25.2 (d, JCÀP=140.1), 25.6, 26.1, 32.6,
51.2, 63.1 (d, JCÀP=6.4), 209.3.
References and Notes
1. W.H.O. The World Health Report 2000: Health Systems,
Improving Performance, 2000.
2. Lambeir, A. M.; Loiseau, A. M.; Kuntz, D. A.; Vellieux,
F. M.; Michels, P. A. M.; Opperdoes, F. R. Eur., J. Biochem.
1991, 198, 429.
31P NMR (81 MHz, CD3OD) d 11.3, 34.7.
3. Fairlamb, A. H.; Opperdoes, F. R.; Borst, P. Nature 1977,
265, 270.
Inhibition studies
Rabbit
muscle
glyceraldehyde-3-phosphate-dehy-
4. Misset, O.; Opperdoes, F. R. Eur. J. Biochem. 1987, 162, 493.
5. Misset, O.; Bos, O. J. M.; Opperdoes, F. R. Eur. J. Bio-
chem. 1986, 157, 441.
6. Willson, M.; Lauth, N.; Perie, J.; Callens, M.; Opperdoes,
F. R. Biochemistry 1994, 33, 214.
drogenase (EC 1.2.1.12), yeast phosphoglycerate kinase
(EC 2.7.2.3), and all substrates and cofactors were pur-
chased from Sigma.
7. Aronov, A. M.; Suresh, S.; Buckner, F. S.; Van Voorhis,
W. C.; Verlinde, C. L. M. J.; Opperdoes, F. R.; Hol, W. G. J.;
Gelb, M. H. Proc. Natl. Acad. Sci. USA 1999, 96, 4273.
8. Barnard, J. P.; Reynafarje, B.; Petersen, P. L. J. Biol.
Chem. 1993, 288, 3654.
9. Opperdoes, F. R. Ann. Rev. Microbiol. 1987, 41, 127.
10. Srivastva, D. N.; Farquhar, D. Bioorg. Chem. 1984, 12, 118.
11. Trentham, D. R. Biochem, J. 1971, 122, 71.
12. Mak, C. P.; Mayerl, C.; Fliri, H. Tetrahedron Lett. 1983,
24, 347.
Inactivation studies
In the di€erent coupled assays shown in the following
inactivation study, the enzyme activities were measured
usingeither the forward or the reverse direction. The
reactions were monitored by absorbance decrease or
increase (at 340 nm) of NADH; its concentration was
calculated from this absorbance usingthe value
e
340=6.22 mMÀ1 cmÀ1. All inhibitors were tested by
13. Stowell, J. K.; Widlanski, T. S. Tetrahedron Lett. 1995, 36,
1825.
preincubation for 5 min with the enzyme in 0.1 M trie-
thanolamine bu€er (pH 7.5 and ionic strength 0.15) at
25 ꢀC. The enzymatic reaction was initiated by the
addition of the reaction mixture in accordance with the
direction of the reaction followed.
14. Jackson, D. Y. Synth. Commun. 1988, 18, 337.
15. Gresham, T. L.; Jansen, J. E.; Shaver, F. W.; Gregory,
J. T.; Beears, W. L. J. Am. Chem. Soc. 1948, 70, 7041.
16. Rengaraju, S.; Berlin, K. D. J. Org. Chem. 1972, 37, 3304.
17. Salomon, C. J.; Breuer, E. Tetrahedron Lett. 1995, 36, 6759.
18. Jackeman, D. L.; Ivory, A. J.; Williamson, M. P.; Black-
burn, G. M. J. Med. Chem. 1998, 41, 4439.
GAPDH inactivation
19. Kim, D. Y.; Choi, Y. J. Synth. Commun. 1998, 28, 1491.
20. Nageswara Rao, B. D.; Cohn, M.; Scopes, R. K. J. Biol.
Chem. 1978, 253, 8056.
21. Schmidt, P. P.; Travers, F.; Barman, T. Biochemistry
1995, 34, 824.
22. De Macedo Puyau, P.; Perie, J. J. Phosphorus, Sulfur and
Silicon 1997, 129, 13.
23. Li, Y. K.; Byers, L. D. Biochim. Biophys. Acta 1993, 1164,
17.
24. (a) Blackburn, G. M.; Kent, D. E. Chem. Commun. 1981,
511. (b) Blackburn, G. M.; Kent, D. E. J. Chem. Soc., Perkin
Trans. I 1984, 1119. (c) Blackburn, G. M. Chem. Ind. (London)
1981, 134.
25. Nieschalk, J.; Batsanov, A. S.; O'Hagan, D.; Howard,
J. A. K. Tetrahedron 1996, 52, 165.
1. With respect to glyceraldehyde-3-phosphate
(GAP) the reaction mixture (1 mL) contained buf-
fer, 1 mM EDTA, 2 mM NAD+, 0.8 mM GAP,
0.1 M KCl and 10 mM potassium phosphate.
2. With respect to 1,3-diPG, a linked assay system
was used in which the reaction of GAPDH is the
reverse of that of the glycolytic pathway. The assay
mixture (1 mL) contained bu€er, 1 mM EDTA,
5.6 mM 3-PG, 1 mM ATP, 5 mM MgSO4, 0.42 mM
NADH and a saturatinglevel of yeast PGK
(11units).
26. Adams, P. R.; Harrison, R. Biochem. J. 1974, 141, 729.
27. Smith, G. D.; Caughlan, C. N.; Ramirez, F.; Glaser, S. L.;
Stern, P. J. Am. Chem. Soc. 1974, 1, 2698.
28. Rudisill, D. E.; Whitten, J. P. Synthesis 1994, 351.
29. Scwalbe, C. H.; Freeman, S. Chem. Commun. 1990, 251.
30. Mizrahi, V.; Modro, T. A. Cryst. Struct. Commun. 1982,
11, 627.
The inhibitor concentration required for 50% inhibition
(IC50) was calculated from at least ®ve inhibitor con-
centrations which were tested at substrate and cofactor
saturatingconditions. The percentaeg of remaining
activity was calculated by comparison with an inhibitor-
free control experiment.