Effects of Bisphosphonates on the Growth of Entamoeba histolytica and Plasmodium Species in Vitro and in Vivo

Article abstract of DOI:10.1021/jm030084x

The effects of a series of 102 bisphosphonates on the inhibition of growth of Entamoeba histolytica and Plasmodium falciparum in vitro have been determined, and selected compounds were further investigated for their in vivo activity. Forty-seven compounds tested were active (IC₅₀ < 200 μM) versus E. histolytica growth in vitro. The most active compounds (IC ₅₀ ～ 4-9 μM) were nitrogen-containing bisphosphonates with relatively large aromatic side chains. Simple n-alkyl-1-hydroxy-1,1-bisphosphonates, known inhibitors of the enzyme farnesylpyrophosphate (FPP) synthase, were also active, with optimal activity being found with C9-C10 side chains. However, numerous other nitrogen-containing bisphosphonates known to be potent FPP synthase inhibitors, such as risedronate or pamidronate, had little or no activity. Several pyridine-derived bisphosphonates were quite active (IC₅₀ ～ 10-20 μM), and this activity was shown to correlate with the basicity of the aromatic group, with activity decreasing with increasing pK_a values. The activities of all compounds were tested versus a human nasopharyngeal carcinoma (KB) cell line to enable an estimate of the therapeutic index (TI). Five bisphosphonates were selected and then screened for their ability to delay the development of amebic liver abscess formation in an E. histolytica infected hamster model. Two compounds were found to decrease liver abscess formation at 10 mg/kg ip with little or no effect on normal liver mass. With P. falciparum, 35 compounds had IC₅₀ values <200 μM in an in vitro assay. The most active compounds were also simple n-alkyl-1-hydroxy-1,1-bisphosphonates, having IC₅₀ values around 1 μM. Five compounds were again selected for in vivo investigation in a Plasmodium berghei ANKA BALB/c mouse suppressive test. The most active compound, a C9 n-alkyl side chain containing bisphosphonate, caused an 80% reduction in parasitemia with no overt toxicity. Taken together, these results show that bisphosphonates appear to be useful lead compounds for the development of novel antiamebic and antimalarial drugs.

Full text of DOI:10.1021/jm030084x

J . Med. Chem. 2004, 47, 175-187
175
Effects of Bisp h osp h on a tes on th e Gr ow th of En ta m oeba h istolytica a n d
P la sm od iu m Sp ecies in Vitr o a n d in Vivo
Subhash Ghosh,† J ulian M. W. Chan,† Christopher R. Lea,† Gary A. Meints,† J ared C. Lewis,† Zev S. Tovian,†
Ryan M. Flessner,† Timothy C. Loftus,† Iris Bruchhaus,‡ Howard Kendrick,§ Simon L. Croft,§ Robert G. Kemp,|
Seiki Kobayashi, Tomoyoshi Nozaki,# and Eric Oldfield*,†,$
Department of Chemistry, University of Illinois at Urbana-Champaign, 600 South Mathews Avenue, Urbana, Illinois 61801;
Bernhard Nocht Institute for Tropical Medicine, Bernhard-Nocht Strasse 74, 20359 Hamburg, Federal Republic of Germany;
Department of Infectious and Tropical Diseases, London School of Hygiene and Tropical Medicine, Keppel Street,
London WC1E 7HT, United Kingdom; Department of Biochemistry and Molecular Biology, The Chicago Medical School,
North Chicago, Illinois 60064; Department of Tropical Medicine and Parasitology, Keio University School of Medicine, J apan;
Department of Parasitology, National Institute of Infectious Diseases, 1-23-1 Toyama, Shinjuku-ku, Tokyo 162-8640, J apan;
Precursory Research for Embryonic Science and Technology, J apan Science and Technology Corporation, J apan; Center for
Biophysics and Computational Biology, University of Illinois at Urbana-Champaign, 600 South Mathews Avenue,
Urbana, Illinois 61801; and Center for Zoonoses Research, University of Illinois at Urbana-Champaign,
2001 South Lincoln Avenue, Urbana, Illinois 61802
Received February 19, 2003
The effects of a series of 102 bisphosphonates on the inhibition of growth of Entamoeba
histolytica and Plasmodium falciparum in vitro have been determined, and selected compounds
were further investigated for their in vivo activity. Forty-seven compounds tested were active
(IC50 < 200 µM) versus E. histolytica growth in vitro. The most active compounds (IC50 ∼ 4-9
µM) were nitrogen-containing bisphosphonates with relatively large aromatic side chains.
Simple n-alkyl-1-hydroxy-1,1-bisphosphonates, known inhibitors of the enzyme farnesylpyro-
phosphate (FPP) synthase, were also active, with optimal activity being found with C9-C10
side chains. However, numerous other nitrogen-containing bisphosphonates known to be potent
FPP synthase inhibitors, such as risedronate or pamidronate, had little or no activity. Several
pyridine-derived bisphosphonates were quite active (IC50 ∼ 10-20 µM), and this activity was
shown to correlate with the basicity of the aromatic group, with activity decreasing with
increasing pKa values. The activities of all compounds were tested versus a human nasopha-
ryngeal carcinoma (KB) cell line to enable an estimate of the therapeutic index (TI). Five
bisphosphonates were selected and then screened for their ability to delay the development of
amebic liver abscess formation in an E. histolytica infected hamster model. Two compounds
were found to decrease liver abscess formation at 10 mg/kg ip with little or no effect on normal
liver mass. With P. falciparum, 35 compounds had IC50 values <200 µM in an in vitro assay.
The most active compounds were also simple n-alkyl-1-hydroxy-1,1-bisphosphonates, having
IC50 values around 1 µM. Five compounds were again selected for in vivo investigation in a
Plasmodium berghei ANKA BALB/c mouse suppressive test. The most active compound, a C9
n-alkyl side chain containing bisphosphonate, caused an 80% reduction in parasitemia with
no overt toxicity. Taken together, these results show that bisphosphonates appear to be useful
lead compounds for the development of novel antiamebic and antimalarial drugs.
In tr od u ction
protozoa.1,2 Conventional therapy for P. falciparum
involves use of a range of compounds including meflo-
quine, artesunate, chloroquine, and quinine, while
treatment for E. histolytica involves use of the drugs
metronidazole or iodoquinol. However, resistance to
many antimalarials is widespread and resistance to
metronidazole is known in many pathogenic bacteria
and protozoa,3 so there is interest in the development
of new and inexpensive drugs. In early work, Eubank
and Reeves4 found that hydrolytically stable analogues
of pyrophosphate, bisphosphonates, had activity against
E. histolytica, and they proposed that these compounds
inhibited the parasite’s pyrophosphate-dependent phos-
phofructokinase (PFK).4 More recently, other results on
E. histolytica were reported using several nitrogen-
containing bisphosphonates.5 These compounds are now
known to be potent, nanomolar inhibitors of the enzyme
Plasmodium falciparum is the major cause of malaria
in humans, with 300-500 million individuals affected
annually, causing ∼2-3 million deaths, while Entam-
oeba histolytica is the cause of tens of millions of cases
of amebic dysentery and is the third leading cause of
morbidity and mortality in humans, due to parasitic
* To whom correspondence should be addressed. Telephone: (217)
333-3374. Fax: (217) 244-0997. E-mail: eo@chad.scs.uiuc.edu.
† Department of Chemistry, University of Illinois at Urbana-
Champaign.
‡ Bernhard Nocht Institute for Tropical Medicine.
§ London School of Hygiene and Tropical Medicine.
| The Chicago Medical School.
Keio University School of Medicine.
National Institute of Infectious Diseases, and J apan Science and
#
Technology Corp.
$
Center for Biophysics and Computational Biology, and Center for
Zoonoses Research, University of Illinois at Urbana-Champaign.
10.1021/jm030084x CCC: $27.50 © 2004 American Chemical Society
Published on Web 12/09/2003

176 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 1
Ghosh et al.
farnesyl pyrophosphate synthase6-12 (FPPS) and are
used clinically to treat osteoporosis, Paget’s disease, and
hypercalcemia due to malignancy.13 In addition, nitrogen-
containing bisphosphonates have activity as herbicides,6
a direct effect on some tumor cells,14-17 as well as
activity against several apicomplexan and trypanoso-
matid parasites. These include Trypanosoma cruzi,18
Leishmania donovani,19 Leishmania mexicana,20 Toxo-
plasma gondii,19,21 and Cryptosporidum parvum.22 In
some cases, bisphosphonate effects appear to be rela-
tively specific against the protozoan, a phenomenon that
is thought to be related to the presence of a pyrophos-
phate-rich metabolism in these cells and the presence
of acidocalcisomes, electron-rich dense granules contain-
ing condensed phosphates.18
In our laboratories, we recently investigated the
effects of 19 bisphosphonates on the intraerythrocytic
growth of P. falciparum.21 The results obtained were
interesting in that only five of the bisphosphonates
tested had measurable activity (IC50 < 200 µM), and
the most active compounds were not the nitrogen-
containing bisphosphonates used clinically for treating
bone resorption disorders. Rather, our results showed
that the more hydrophobic de-aza analogue of risedro-
nate (2-phenyl-1-hydroxyethane-1,1-bisphosphonic acid),
as well as simple alkyl bisphosphonates, were the most
active species.21 This observation is reminiscent of the
earlier observation of Eubank and Reeves that 1-hy-
droxynonane-1,1-bisphosphonic acid is an inhibitor of
the growth of E. histolytica4 and the observation that
other n-alkyl bisphosphonates have activity versus T.
cruzi,23 mediated most likely by inhibition of the para-
sites’ farnesylpyrophosphate synthase.24 Since it is
known that small, very hydrophilic bisphosphonates,
such as 1-hydroxyethane-1,1-bisphosphonate (etidro-
nate, Didronel) and dichloromethane-1,1-bisphospho-
nate (clodronate), do not readily move across cell
membranes,25 we reasoned that the poor activity of the
nitrogen-containing bisphosphonates in P. falciparum
might arise, at least in part, due to poor transport
properties because of the presence of a charged nitrogen
atom in the alkyl/aryl side chain. That is, the more
lipophilic bisphosphonates (containing n-alkyl side chains)
might have improved uptake across the erythrocyte
membrane, resulting in improved IC50 values, and
perhaps, the activity of the alkylbisphosphonates in
some systems might be due to enhanced uptake.
F igu r e 1. Structures of the 1-hydroxy-1,1-bisphosphonates
and 1,2-bisphosphonates investigated.
In this paper, we report on the activity of a broad
variety of nitrogen-containing and non-nitrogen-con-
taining bisphosphonates against both E. histolytica and
P. falciparum. A pattern of activity emerges in which
simple n-alkyl bisphosphonates and nitrogen-containing
bisphosphonates containing large and/or uncharged side
chains have in general much greater activity in inhibit-
ing cell growth than do the nitrogen-containing bispho-
sphonates that are currently used in bone resorption
therapy, such as risedronate, alendronate, and pamid-
ronate. The toxicity of the compounds investigated was
estimated using a human cell line, and based in part
on these results, several bisphosphonates were evalu-
ated for their ability to reduce amebic liver abscess
formation in an E. histolytica infected hamster model
and for their activity in a Plasmodium berghei BALB/c
mouse suppressive test.
F igu r e 2. Structures of the aminomethylene bisphosphonates
investigated.
Resu lts a n d Discu ssion
E. h istolytica in Vitr o Testin g. We show in Figures
1-3 the structures of the compounds tested. A wide
range of structural motifs were investigated, and the
basic structures are shown in bold in the figures. Figure
1 is composed primarily of 1-hydroxy-1,1-bisphospho-
nates, where the side chain extends from a backbone
carbon that is also bonded to two phosphonate groups
and one hydroxyl group. We also made three novel 1,2-
bisphosphonates, in which the phosphonate groups are

Bisphosphonates as Antiparasitics
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 1 177
F igu r e 4. Graph showing the effect of n-alkylbisphosphonate
side chain length on IC50 values against E. histolytica. The
line is to guide the eye.
commercially available bisphosphonates zoledronate
(17) and risedronate (20) were active, with IC50s of 12.0
and 73.5 µM, respectively, but the chemically similar
bisphosphonates 16, 19, and 21, were all inactive. Of
the alkyl, nitrogen-containing 1-hydroxy bisphospho-
nates, only the long tertiary amine compound 31 (iban-
dronate) had measurable activity, at 53.6 µM. All of the
aminopropylidene bisphosphonates containing a phenyl
ring, 32-36, were active, with IC50s of 143, 45.1, 20.5,
23.4, and 6.49 µM, respectively. None of the 1,2-
bisphosphonates (all with pyridyl side chains), 37-39,
had any measurable activity.
Among the n-alkyl secondary amine side chain bis-
phosphonates, only the two longest compounds, 44 (31.1
µM) and 45 (136 µM), had activity against E. histolytica.
The branched alkane secondary amines, 47-49, were
active compounds with respective IC50s of 36.9, 135, and
65.3 µM, yet the structurally similar compound 46 was
not. Why is 46 inactive? Clearly, 46 has the smallest
alkyl side chain in the monoalkyl species 46-49, so this
could be a factor, and indeed, as shown in Figure 4 (in
another homologous series), large effects can be seen
on increasing chain size. Interestingly, we found com-
pound 47 to be the most active of compounds 46-49,
and it certainly seems possible that these compounds
might be FPP synthase inhibitors in E. histolytica
since, as shown in Table 2, on average their IC50 values
versus Leishmania major FPPS are only 780 nM. Plus,
the most active species found here would appear to
most closely resemble the dimethylallyl pyrophosphate
(DMAPP) transition state/reactive intermediate, since
it has a branched C5-side chain:
F igu r e 3. Structures of the arylaminomethylene bisphospho-
nates, aminoethylene bisphosphonates, and bisphosphonate
tetraethyl-esters investigated.
attached to adjacent carbons. Figure 2 consists of
aminomethylene 1,1-bisphosphonates, in which the
backbone carbon is bonded to a nitrogen in the side
chain and two phosphonate groups. Figure 3 shows
additional aminomethylene 1,1-bisphosphonates tested,
derived from anilines and aminopyridines, as well as
three aminoethylene 1,1-bisphosphonates and two tet-
raethyl bisphosphonate esters, which were investigated
to see if they had enhanced activity due to enhanced
lipophilicity.
Table 1 shows the IC50 values for all compounds (1-
102) tested. The experimental concentrations required
to reduce parasite proliferation by 50%, the IC50 values,
were extracted from experimental growth-versus-inhibi-
tor concentration assays by fitting the experimental
data to the rectangular hyperbolic function
I
maxC
I )
(1)
IC50 + C
where I is the percent inhibition, Imax ) 100%, and C is
the concentration of the inhibitor (µM). Representative
graphical results (for compounds 5, 8, 31, and 44) are
shown in Figure S1 of the Supporting Information.
The 47 bisphosphonates that showed activity (IC50
<
200 µM) against E. histolytica arise from several dif-
ferent classes of compound. Among the n-alkyl side
chain containing bisphosphonates, eight of the 11
compounds tested showed activity (Table 1 and Figure
4). The C5 compound (3) was the least active species
(IC50 ) 157 µM). Activity increased as the chain length
increased, to a maximum activity at C10 (7, 11.0 µM),
before dropping off once again for the longer chain
compounds (Figure 4). The branched alkane side chain,
12, also showed good activity, at 12.4 µM. The three
alicyclic compounds (13-15, Table 1) were inactive. The
The tertiary amine side chain compounds, 50, 51, and
53, all had measurable activity (62.6, 47.3, and 194 µM,
respectively), but the closely related compound 52 was
inactive (IC50 > 200 µM). For these dialkylated com-
pounds, the presence of two C3 side chains (52, 53)
results in low activity, presumably due to a steric effect,
but species containing at least one smaller ethyl group

178 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 1
Ghosh et al.
Ta ble 1. Growth Inhibition (IC50) Data for P. falciparum and E. histolytica, Toxicity Data (LD50) for a KB Cell Line, and Computed
Therapeutic Index (TI) Results for Bisphosphonates
IC50 (µM)
TI
IC50 (µM)
TI
compd
E. his.a
P. falb
LD50 (µM)c
E. his.d
P. fale
compd
E. his.a
P. falb
LD50 (µM)c
E. his.d
P. fale
1
2
3
4
5
6
7
8
>200
>200
157
29.5
22.3
13.3
11.0
63.7
74.0
130
56.7
5.53
4.34
26.7
0.83
5.23
1.07
>967
>1041
>993
109
>7.44
>18.4
>180
25.1
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
>200
194
>200
>200
>200
18.4
>200
>200
>200
ndg
>200
>200
>200
>200
>200
>200
>200
>200
>200
>200
>200
>200
31.3
322
852
>1090
964
>1119
>1009
>1098
>922
99.8
>970
246
260
629
272
>844
166
354
150
<0.93
>1016
278
210
373
211
495
749
314
634
>976
710
>779
465
782
18.6
127
187
>991
>963
>992
86.7
466
257
>764
633
520
673
144
386
>568
127
>790
4.39
>6.33
3.69
13.8
>58.3
>73.3
7.72
>200
>200
>200
>200
>200
71.0
>200
15.6
113
>200
35.9f
16.0
>200
>200
15.5
>200
nd
>200
172f
3.98
52.4
308
11.5
>775
>806
492
>933
>154
460
17.4
>25.3
>13.0
9
2.12
28.8
36.8
0.50
>16.6
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
44.0
>200
12.4
>729
<0.65
228.1
>1244
>1219
1086
41.0
63.7
>931
13.9
249
171
218
>804
>746
>977
156
145
91.0
>62.2
2.18
>200
>200
>200
>200
>200
>200
167
7.7
>200
123
>200
>200
>200
158
>200
>200
>200
>200
>200
>200
59.1
19.1
>200
>200
>200
>200
12.0f
>200
>200
73.5
>200
>200
>200
>200
>200
>200
>200
>200
>200
>200
53.6
17.5
17.0
5.31
3.46
0.38
>121
22.9
2.02
1.62
52.7
2.83
5.83
56.5
113
8.89
>200
46.3
103
76.6
132
36.2
8.76
6.60
>200
19.2
8.06
2.05
6.46
>4.72
>200
>200
>200
>200
104
>200
>200
>200
>200
>200
>200
10.6
>200
>200
>200
>200
197
>200
>200
>200
50.4
33.0
680
>1052
10.1
>200
>200
>200
>200
60.0
6.83
0.19
>5.20
8.40
>35.5
>29.5
48.1
0.17
>22.0
16.6
143
45.1
20.5
23.4
33.55
22.8
>200
2.90
>739
379
>727
>690
312
<1.1
1021
>1123
499
13.0
0.29
64.0f
>200
>200
87.0
>238
9.94
6.49
31.4
>200
>200
>200
>200
>200
>200
>200
31.1
>200
>200
>200
>200
>200
4.8
56.5
13.3
112
>200
>200
>200
103
90.1
>200
>11.4
>93.5
>200
21.7
44.2
39.5
96.1
>200
177
>45.7
1.96
11.8
212
>1148
>966
66.8
>919
50.5
480
394
>989
809
>240
>17.1
5.02
2.67
1.30
2.15
>6.76
3.58
18.2
136f
>200
36.9
>8.18
28.5f
>200
>200
>200
>200
>200
>200
13.3
5.12
13.0
2.92
>15.1
12.9
17.0
135
>200
>200
>200
>200
65.3f
62.6f
47.3
>9.64
8.98
>1055
>22.3
a
b
IC50 (µM) for E. histolytica (metronidazole IC50 ) 3.2 µM). IC50 (µM) for P. falciparum (chloroquine IC50 ) 0.012 µM). c LD50 (µM)
for a KB (human nasopharyngeal carcinoma) cell line. Therapeutic index for E. histolytica. e Therapeutic index for P. falciparum. f IC50
determined analytically, utilizing eq 1, from the average inhibition of multiple trials (g3) at 100 µM. nd denotes value not determined.
d
g
(50, 51) are more active. The only fused ring tertiary
amine active against E. histolytica was the tetrahy-
droisoquinoline derivative 59 (71.0 µM). The arylalkyl
side chain of compound 61 confered relatively good
activity, at 15.6 µM. Among the cycloalkane-containing
aminomethylene bisphosphonates, the propyl (62, 113
µM), pentyl (64, 35.9 µM), hexyl (65, 16.0 µM), and octyl
(68, 15.5 µM) species were all active, in sharp contrast
to the lack of activity of the nitrogen-free alicyclic
compounds, 13-15.
increased, i.e., as the molecules became more hydro-
phobic. Compound 73, which contains a butoxy subtitu-
ent on the ring, was the most active compound screened,
at 3.98 µM. The biphenyl side chain containing bisphos-
phonates (76, 77) were also very active, with the para
derivative (76, 8.70 µM) being slightly less active than
the meta species (77, 6.60 µM). Compound 79, the meta
di-tert-butyl aniline, also showed activity, at 19.2 µM.
Of the 14 aminopyridine derivatives, nine had activity
(IC50 < 200 µM), including the 3-pyridine-derived
compound 84 (60.0 µM), but not the 2-pyridine isomer,
83. Conversely, the two active alkyl-substituted ami-
nopyridines were both 2-pyridine derivatives. Com-
pound 85 (64.0 µM) has a methyl substitution in the 3
Seven of the 12 aniline derivatives were active, three
of which had alkyl substitutions on the ring. These
compounds, 72 (172 µM), 74 (132 µM), and 75 (36.2 µM),
have increasing activity as the substitutent chain length

Bisphosphonates as Antiparasitics
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 1 179
Ta ble 2. Activity against E. histolytica and L. major FPPS
FPPS enzyme, while others that show potent FPPS
activity (such as 23) are inactive in other systems. This
is demonstrated in Table 2, which shows the IC50 results
for E. histolytica in rank order with, for comparison, IC50
results for a Leishmania major FPPS enzyme.27 Twenty-
five of the 47 compounds active against E. histolytica
have only moderate to low activity (IC50 > 1 µM) against
the L. major FPPS.27 Moreover, overall, the activities
are uncorrelated (R2 ) 0.027, P-value ) 0.373, n ) 32).
This could indicate that either the E. histolytica FPPS
is very different than the L. major enzyme or that there
is more than one target or, in principle, both could be
true.
IC50 (µM)
L. major
IC50 (µM)
L. major
FPPS
1.4
no. E. histolytica
FPPS
no. E. histolytica
73
36
77
76
7
17
12
6
68
61
65
79
34
90
5
35
96
4
3.98
6.49
6.60
8.76
>100
0.45
10-100
nda
91
33
51
31
84
50
8
85
49
59
20
9
88
93
62
74
48
45
32
3
44.2
45.1
47.3
53.6
60.0
62.6
63.7
64.0
65.3
71.0
73.5
74.0
87.0
96.1
0.50
nd
0.48
0.16
18.2
7.75
1.29
1.14
nd
0.17
10-100
0.24
nd
40.5
nd
0.36
0.90
0.31
9.6
11.0
3.42
0.11
6.75
2.37
2.44
10.74
0.16
>100
0.25
>100
19.94
0.49
nd
5.88
1.06
0.54
10-100
0.83
nd
12.0
12.4
13.3
15.5
15.6
16.0
19.2
20.5
21.7
22.3
23.4
28.5
29.5
31.1
35.9
36.2
36.9
39.5
44.0
The isoprene biosynthesis pathway in E. histolytica
has not been extensively studied. However, it is ex-
tremely unusual, since there is no evidence for the
presence of either classical mevalonate or nonmeva-
lonate pathway enzymes.28 There is, however, a se-
quence that corresponds to an archaeal multifunctional
enzyme capable of synthesizing farnesyl pyrophosphate,
so FPPS is certainly a possible target. When considering
the top three active compounds (73, 36, 77), 73 and 77
are aniline derivatives that, because of their relatively
low pKa values, are expected to be poor FPPS inhibitors.
However, 36 is a known potent FPPS inhibitor29 and
likely to inhibit E. histolytica FPPS. We conclude,
therefore, that there is more than one target for the
bisphosphonates in E. histolytica.
113
132
135
136
143
157
172
177
194
44
64
75
47
92
10
72
95
53
10-100
>100
>100
6.5
a
nd denotes not determined.
position on the ring and compound 95 (177 µM) has
methyl groups at positions 3 and 5 on the ring, yet
compounds 86 (methyl substitution at the 5 position)
and 87 (methyl substitution at position 6) had no
discernible activity, although all four were derived from
2-pyridine. Compound 93, the 4-hydroxy-3-pyridine
derivative, had an IC50 of 96.1 µM. The remaining five
active aminopyridines all contained halogen substitu-
ents. The activity of compounds 88 (87.0 µM), 91 (44.2
µM), and 92 (39.5 µM), which are all derived from
2-pyridine with a halogen (chlorine, bromine, and iodine,
respectively) substitution at the para position, increased
with the atomic weight of the halogen. Compound 96
(28.5 µM) is also a 2-pyridine derivative, but it has a
dibromo substitution, at positions 4 and 6 on the ring.
Of the halogenated 3-pyridine derivatives, 89 (chlorine
at position 2) was inactive, while 90 (chlorine at position
4) was active, with an IC50 of 21.7 µM.
The aminoethyl bisphosphonates (98-100) were all
inactive, including compound 100, which is derived from
the antibiotic ciprofloxacin. The tetraethyl ester bisphos-
phosphonates (101, 102) were also inactive, as were
their parent bisphosphonates (86, 87 respectively).
The results shown in Table 1 and as outlined above
were quite surprising, since in previous work it has been
shown that the enzyme farnesyl pyrophosphate syn-
thase (FPPS) is the primary target of bisphosphonates
in the (trypanosomatid) parasites T. cruzi and T.
brucei,24,26 as well as in bone resorption therapy.12 Our
hypothesis was, therefore, that FPPS would be the
target for most of the bisphosphonates tested in these
systems as well. However, from the results we have
obtained here, this hypothesis may be incorrect, espe-
cially for the most active compounds.
The structures of the most active, non-FPPS inhibi-
tors are of some interest. Compounds 73, 77, and 79
are all aniline derivatives, containing phenoxylalkyl,
biphenyl, and phenyl-di-tert-butyl side chains, respec-
tively, but none of them are potent FPPS inhibitors
(Table 2). Other actives include 7, a simple n-alkyl; 12,
a branched alkyl; and 61, a phenylalkyl bisphosphonate.
All of these species have in common large, hydrophobic
side chains, suggesting that compounds with more
lipophilic side chains generally appear to have relatively
good activity against E. histolytica.
Lipophilicity also clearly plays a significant role in
the activity of the n-alkylbisphosphonates, compounds
that are also known to be moderate inhibitors of FPP
synthase. While it might be thought that alkylbis-
phosphonates would not be FPP synthase inhibitors,
since they lack the nitrogen positive charge feature
responsible for carbocation transition state/reactive
state mimicry, this is an oversimplification, since both
electrostatic (positive charge, negative ionizable) as well
as hydrophobic (van der Waals or London dispersion
forces, i.e., attractive) interactions contribute to the
overall potential energy of interaction of these drug
molecules with their target. And for FPP synthase, the
hydrophobic/steric interactions account for ∼60% of the
total interaction energy, while electrostatic interactions
(from both the bisphosphonate and positive N charge
features) account for only ∼4% of the interaction
energy.21 Thus, removing just one part of the electro-
static potential (due to the positive charge feature) can
be expected to have only a relatively small effect,
especially if another interaction (the hydrophobic or van
der Waals dispersion interaction) increases, as with an
increase in chain length.
For example, several of the compounds with the
highest activity (<20 µM) against E. histolytica (7, 12,
61, 73, 77, 79) have little activity (>3 µM) against an
The activity of the alkylbisphosphonates is, in fact,
very strongly dependent on overall side chain length.

180 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 1
Ghosh et al.
As shown in Table 1 and Figure 4, activity rapidly
increases from 1 and 2 (C3, C4: not active) to 3 and 4
(C5, C6: 157 and 29.5 µM, respectively). Optimal
activity is seen with the C9 and C10 alkyl side chains
(6, 7: 13.3 and 11 µM, respectively), then activity begins
to fall off. At C11, C13 and C15, the IC50 increases to
∼60 µM and for the very long (C17) species (11), the
IC50 is >200 µM. This pattern of activity is not unex-
pected, however, since the short alkyl species will only
have a weak interaction with FPP synthase, and it is
well-known that chain elongation of FPP (which has an
overall chain length of 12 carbons, excluding the methyl
substituents) to GGPP does not occur with FPP syn-
thase, due to the presence of phenylalanine, tyrosine,
or histidine groups in the active site, which are thought
to block chain elongation.27 The effect of chain elonga-
tion on activity is also seen in enzyme inhibition data
for FPP synthase (from T. brucei), where the IC50
increases from 3.5 µM (at C10) to 4.9 µM (at C11), and
from 2.4 µM (C9) to 3.5 µM (C10) to 7.8 µM (C11) for
the FPP synthase from L. major (see Table 2).27 These
results are all consistent with the early work of Reeves
and Eubank, who found that the C8 alkylbisphospho-
nate was a good inhibitor of E. histolytica growth,4 and
based on our chain length dependence results (Figure
4), it appears that this length is close to an optimum
one for the n-alkyl bisphosphonates. This length is also
approximately the overall chain length of the GPP
substrate (eight carbons) of FPP synthase. We propose,
therefore, that alkylbisphosphonates are good inhibitors
of E. histolytica growth, due at least in part to the
lipophilic nature of their alkyl side chains in enhancing
membrane transport, and targeting FPPS.
Another class of bisphosphonates exhibiting interest-
ing growth inhibiton results are the pyridyl amino-
methylene bisphosphonates, shown in Figure 3 (com-
pounds 83-97). These bisphosphonates were first
developed by Nissan30 and later by Zeneca6 as herbi-
cides, and they can be potent FPP synthase inhibitors,
having low nanomolar Ki values.6 However, several of
these species (Table 1) appear to be inactive versus E.
histolytica, while others (such as 85 and 88), which are
good FPP synthase inhibitors,27 have only low activity.
Are there some underlying principles that one might
apply to give a general description of the activity of
this class of compounds in E. histolytica growth inhibi-
tion?
F igu r e 5. Graph showing correlation between the pIC50
[)-log IC50 (M)] for E. histolytica growth inhibition by
bisphosphonates and the computed pKa value of the parent
base. The line is to guide the eye.
effect can operate in the case of p-aminopyridines as
well:
But in the case of the m-aminopyridines, there is no
such resonance stabilization possible. As a result, the
m-aminopyridine is the weakest base (computed31 ortho
pKa ) 6.67; meta pKa ) 6.16; para pKa ) 9.25). On
halogen substitution, the electron-withdrawing effect of
the halogens is pronounced, and results in even lower
basicity. For example, in the case of 96, addition of the
two bromine atoms reduces the (computed) pKa to
∼1.90, an extremely weak base, while addition of the
single chloro substituent in 90 (which already has a
m-nitrogen) again results is extremely weak basicity,
with a computed pKa of 2.09. But what does this mean
in terms of E. histolytica growth inhibition?
As shown in Table 1, 90 and 96 are active, and they
will have very low pKa values. However, five other
pyridyl aminomethylene bisphosphonates are less ac-
tive, and as shown in Figure 5, activity can be correlated
with pKa. In particular, the most active compounds have
pKa values of ∼2. and at physiological pH values, their
rings will be uncharged, while those of the less active
species will be charged (to a greater or lesser extent).
Of course, the trend shown in Figure 5 does not consider
any steric or electrostatic interactions with the target
protein. Rather, it simply indicates that the most active
species will have uncharged side chains, just as found
with the alkylbisphosphonates (and most likely, the
aniline derivatives, 73, 76 and 77).
Upon inspection, it can be seen that the two most
active aminopyridine derivatives, 90 and 96, are rather
unusual (when compared with known FPP synthase
inhibitors) in that 90 has a m-nitrogen as well as a
p-chloro substituent, while 96 has a dibromo substitu-
tion. These substitutions can be expected to have very
major effects on the pKas of these compounds and,
hence, on their protonation state. In the case of o-
aminopyridines, there is expected to be extensive charge
delocalization due to the presence of amidinium-like
structures:
In conclusion, of the four most active compounds
against E. histolytica, only one is active against farnesyl
pyrophosphate synthase, and overall there is no cor-
relation between E. histolytica growth and FPPS en-
zyme inhibition activity. If FPPS is not the universal
target, then what are other likely targets? There are,
of course, numerous intermediates in the isoprene and
mevalonate pathways that are phosphorylated, where
bisphosphonates could act as substrate mimics, and
A similar resonance stabilization/charge delocalization

Bisphosphonates as Antiparasitics
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 1 181
points (open circles) represent only lower limits for the
LD50. Nevertheless, there is no correlation either in
cases where the LD50 values (closed circles) are known.
Next we estimated a therapeutic index ratio (TI) for
each of the bisphosphonates investigated, using the
following definition
LD50
TI )
(2)
IC50
in which LD50 is the concentration of drug that killed
50% of a human nasopharyngeal carcinoma (KB) cell
line and IC50 is that for E. histolytica growth inhibition.
This approach enables an estimate of which compounds
might be efficacious in vivo. The numerical results for
each compound are given in Table 1. We then plotted
these therapeutic indices versus the IC50 values, obtain-
ing the results shown in Figure 6B. Here, the solid
circles again indicate data points for which discrete LD50
values were obtained, while the open circles indicate
only lower limits for the therapeutic index, since no
toxicity was observed at the highest bisphosphonate
levels tested (300 µg/mL).
We selected a series of five compounds having good
IC50 and TI values for in vivo testing in a hamster model
of E. histolytica induced liver abscess formation. We
chose the alkyl bisphosphonate 7 (IC50 ) 11 µM, TI )
>73.3), a known29 potent bone resorption drug (N-
[methyl(4-phenylbutyl)]-3-aminopropyl-1-hydroxy-1,1-
bisphosphonate (35, IC50 ) 23.4 µM; TI ) >29.5), the
phenoxyethyl analogue of N-Me pamidronate (36, IC50
) 6.49 µM; TI ) 48.1), a biphenyl aminomethylene
bisphosphonate (77, IC50 ) 6.6 µM; TI ) 113), and the
dibromopyridyl aminomethylene bisphosphonate (96,
IC50 ) 28.5 µM; TI ) 18.2). We tested each of these five
compounds in pairs of hamsters, together with an
infected but otherwise nontreated pair. Drugs were
administrered 1 day after infection at 10 mg/kg ip for 5
days. Results for liver abscess weights and the weights
of the normal liver fractions (equal to total liver weight
minus liver abscess weight, in mg) are given in Table
3. Surprisingly, 36, the most active compound tested in
vitro, was ineffective in vivo. The largest decrease in
liver abscess formation was seen with compound 35,
N-[methyl(4-phenylbutyl)]-3-aminopropyl-1-hydroxy-
1,1-bisphosphonate. However, there was a statistically
significant decrease in normal liver weight (26%),
together with ascites formation and diarrhea. In con-
trast, the alkyl bisphosphonate 7 showed a 68% reduc-
tion in liver abscess formation but only a 9% reduction
in normal liver weight (about the experimental uncer-
tainty), while the biphenyl compound 77 showed a 36%
abscess reduction with an insignificant change in nor-
mal liver weight. These are very promising initial in vivo
results for reduction in liver abscess formation and
suggest that 7 and 77 may represent useful new leads
for the development of antiamebic drugs. Also of interest
is the fact that these two species clearly cluster to the
upper left of the TI/IC50 plot (Figure 6B), supporting the
use of this representation to select the most promising
drug leads.
F igu r e 6. Scatter plots of (A) IC50 versus LD50 and (B) IC50
versus TI for E. histolytica growth inhibition by bisphospho-
nates. The open circles represent data points where only a
lower limit to the LD50 or TI was determined. Compounds 7,
35, 36, 77, and 96 were investigated in vivo for their ability
to delay the development of liver abscesses.
many more outside this pathway. One system that has
been proposed previously is the E. histolytica phospho-
fructokinase, and we describe below additional experi-
ments to test the hypothesis that this enzyme is the
target for some of the more active inhibitors. Fortu-
nately, however, even in the absence of a confirmed
target for many of the most active species, the observa-
tion of quite potent inhibitory activity is, in and of itself,
of considerable interest and encouraged us to test a
selection of these compounds in vivo.
E. h istolytica in Vivo Testin g. The patterns of E.
histolytica growth inhibition by bisphosphonates are of
interest since they are different than those seen with
mammalian cells and this might, therefore, lead to the
development of compounds that have selective activity
against E. histolytica. For example, while the simple
alkylbisphosphonates are active against E. histolytica
at low micromolar concentrations, their effects on, for
example, rat calvaria cells have been reported to be
negligible.30 Likewise, the lack of FPPS inhibition
activity of two of the top three growth inhibitors could
lead to antiamebic drugs with low toxicity. To further
explore this topic in detail, we investigated if there was
any correlation between the IC50 for parasite growth
inhibition (the IC50s) and the LD50 for growth of mam-
malian cells, using a human nasopharyngeal carcinoma
(KB) cell line.21 If the IC50 results are correlated, there
is unlikely to be good parasite specificity. There is no
correlation, however, as may be seen in Table 1 and
Figure 6A, although we hasten to add that some data
P . fa lcipa r u m Testin g in Vitr o. We have also
investigated the activity of the compounds shown in
Figures 1-3 against P. falciparum, in vitro. Table 1

182 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 1
Ghosh et al.
Ta ble 3. Effects of Bisphosphonates on Liver Abscess Formation in E. histolytica Infected Hamstersa
liver abscess (mg)
mean
normal liver weight (mg)b
drug
1c
2c
% decrease
1c
2c
mean
% decrease
control
7
35
36
77
816
516
10
1145
561
853
960
53
178
970
588
680
888
285
94
1058
570
767
0
2552
2564
1667
2314
2320
2977
2178
1761
1845
2275
2597
2788
2365
2163
1756
2295
2459
2883
0
9
26
3
-4
-22
68
89
-19
36
14
96
a
b
Two hamsters were treated at 10 mg/kg of the corresponding drug ip once a day for 5 days. The normal liver weight is the difference
between the total liver weight and the liver abscess weight. c The numbers refer to the two hamsters used in each experiment.
shows the IC50 values for all compounds tested. IC50
values were determined as described above using eq 1,
and TI values using eq 2. Representative graphical
results for compounds 9, 33, 42, and 88 are shown in
Figure S2 of the Supporting Information.
The 35 bisphosphonates that had measurable activity
(IC50 < 200 µM) against P. falciparum again consist of
a diverse collection of structures. Of the 11 n-alkyl side
chain compounds, all but the longest (11, C17) showed
activity. The span of activity for these compounds covers
a range from 0.83 µM (6, C9) to 130 µM (1, C3), with
six of the compounds having IC50s below 10 µM. Neither
the branched alkyl side chain compound (12) nor the
cycloalkyl side chain compounds (13-15) had any
discernible activity.
F igu r e 7. Graph showing the effect of alkyl side chain length
of n-alkylbisphosphonates on IC50 values against P. falci-
parum. The line is to guide the eye.
Both of the commercially available bisphosphonates,
zoledronate (17) and risedronate (20), had modest
activities (167 and 123 µM, respectively). However, the
de-aza analogue of risedronate (18, 7.7 µM) was much
more active. Isomers of zoledronate (16) and risedronate
(19 and 21) were all inactive. A homologue of risedro-
nate containing an ethyl-3-pyridyl side chain (23) was
also inactive. Compound 24, which has a 4-ethylmer-
captopyridine side chain, also had measurable activity,
with an IC50 of 158 µM.
The nitrogen-containing alkyl side chain compounds
(26-31) were all inactive except for 31 (ibandronate),
the longest of this set, which had an IC50 of 59.1 µM.
Of the aminopropylidene bisphosphonates containing a
phenyl ring, those that had a methyl group on the
nitrogen, 32, 33, 35, 36, had activities of 33.6, 22.8, 2.90,
and 31.4 µM, respectively. Compound 34, which has a
hydrogen instead of a methyl substituent on the nitro-
gen, was inactive. None of the pyridine-derived 1,2-
bisphosphonates (37-39) showed any measurable ac-
tivity.
Of the n-alkyl aminomethylene bisphosphonates,
compounds 42-45 were active with IC50s of 4.8, 56.5,
13.3, and 112 µM, respectively, but the only branched
alkyl side chain to display activity was compound 49
(103 µM), and of the tertiary amine side chain contain-
ing species, the only active compound was the diethyl-
amino species 50, at 90.1 µM. The morpholino compound
(55, 18.4 µM) was the only fused-ring tertiary amine to
possess activity. None of the alkyl aromatic (60, 61),
cycloalkane (62-68), or thiazole (69, 70) derivatives had
any discernible activity.
Five of the 12 aniline derivatives were active, three
of which had alkyl substitutions on the ring. Unlike the
trend with E. histolytica, compounds 72 (31.3 µM), 74
(46.3 µM), and 75 (103 µM) had decreasing activity as
the substituted chain length increased, although the
effects (in both cases) were small. Of the biphenyl side
chain bisphosphonates, the p-biphenylamine (76, 76.6
µM) was active while the meta isomer (77) was not.
These activities were also much lower than those found
with E. histolytica. Compound 81, the N-ethylcarbazolyl
derivative, also showed some activity, with an IC50 of
104 µM.
Only three of the 15 aminopyridine derivatives showed
activity. Compound 88 (10.6 µM) has a nitrogen at the
2 position and a chlorine at the 4 position. Compound
93, the 4-hydroxy-3-pyridine derivative, had an IC50 of
197 µM and is essentially inactive (existing primarily
as the zwitterionic or pyridone species) The isoquinolyl
species, 97, had an IC50 of 50.4 µM. Of the aminoeth-
ylene bisphosphonates (98-100), the 2-pyridine deriva-
tive was active (98, 17 µM), but the 4-pyidine and the
ciprofloxacin derivatives (99 and 100, respectively) were
not. As with E. histolytica, the tetraethyl bisphospho-
nate esters (101, 102) were inactive, as were their
respective parent bisphosphonates (86, 87).
These results, while albeit of some complexity, are
nevertheless again of considerable interest, since many
of the most active compounds (four of the top six most
active) are simple n-alkylbisphosphonates (3-6). For all
of the n-alkyl compounds (1-11), we again find that
there is a general pattern of increasing activity with
increasing chain length for the short chain compounds,
as shown in Figure 7, with good activity for the C5-
C11 side chains, then generally decreased activity for
the very long side chains. Similar results were found
for E. histolytica, where two (6, 7) of the top eight active
compounds were n-alkylbisphosphonates, and the same
general trend in side chain length vs activity was
observed in Figure 4.
There are several similarities between the results
obtained with the two organisms. Indeed, for the 35

Bisphosphonates as Antiparasitics
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 1 183
Ta ble 4. Activity against P. falciparum and L. major FPPS
IC50 (µM)
L. major
IC50 (µM)
L. major
no. P. falciparum
FPPS
no. P. falciparum
FPPS
6
8
9
35
4
42
7
3
18
88
44
98
55
33
5
10
72
36
0.83
1.07
2.12
2.90
4.34
4.80
5.23
5.53
7.70
2.37
7.75
10-100
0.49
5.88
0.35
3.42
9.80
16.6
0.24
1.06
0.17
>100
0.50
19.9
6.50
32
74
97
43
2
31
76
50
49
75
81
45
20
1
33.6
46.3
50.4
56.5
56.7
59.1
76.6
90.1
0.31
nda
0.43
1.72
>100
0.48
nd
18.2
1.14
10-100
nd
0.90
0.17
>100
>100
0.11
nd
103
10.6
103
104
112
123
130
158
167
197
13.3
17.0
18.4
22.8
26.7
28.8
31.3
31.4
24
17
93
10-100
0.45
a
nd denotes not determined.
Ta ble 5. In Vivo Results for P. berghei ANKA Suppressive
Test in BALB/c Mice
dose
(mg/kg)
schedule
(days)
mean reduction
in parasitemia (%)
drug
3
4
6
18
26
25
10
10
25
25
10
4
4
2
4
4
4
80
48
75
64
5
F igu r e 8. Scatter plots of (A) IC50 versus LD50 and (B) IC50
versus TI, for P. falciparum. The open circles represent data
points where only a lower limit to the LD50 or TI was
determined. Compounds 3, 4, 6, and 18 were investigated in
the P. berghei mouse suppressive test.
chloroquine
100
that compounds 3, 4, 6, and 18 have therapeutic indices
>25 and low IC50 values (<10 µM). The IC50 values for
4 and 6 are particularly low, in the 0.5-5 µM range [4,
IC50 ) 500 nM (W2 strain), 500 nM (Ghana strain), 4.34
µM (3D7 strain); 6, IC50 ) 1 µM (W2), 2 µM (Ghana)
and 830 nM (3D7)].
We selected compounds 3, 4, 6, and 18 for in vivo
testing in a P. berghei BALB/c mouse suppressive test.
In addition, we tested pamidronate (26, Aredia), a
second-generation bisphosphonate, to see if it had any
effects on parasitemia reduction. On day 1, five female
BALB/c mice per group were inoculated with 0.2 mL of
1% parasitemia (P. berghei ANKA, 1 × 107 infected
RBCs). Drug administration commenced 2 h postinfec-
tion, ip. On day 5 (day 3 with 6), tail smears were taken,
fixed with methanol, and stained with 10% Giemsa’s
stain, and parasitemia was determined under oil im-
mersion at 1000× magnification.
compounds active against P. falciparum, 25 also have
activity versus E. histolytica. There are also similarities
in the structures of the most active species, with the
long, hydrophobic side chain containing species having,
in general, higher activity in both organisms. This may
suggest that membrane transport plays an important
role, so development of additional lipophilic compounds
appears to be a worthwhile goal. Indeed, McIntosh and
Vaidya have suggested that there might be difficulty
with cellular uptake of bisphosphonates in vitro with
Plasmodium species,33 citing previous work which con-
cluded that endocytosis (specifically, fluid phase pinocy-
tosis34) is the primary cellular uptake method for
bisphosphonates. Since erythrocytes are incapable of
endocytosis, some bisphosphonates may have difficulty
in penetrating the erythrocyte membrane, unlike the
situation with the trypanosomatids or macrophages or
macrophage-related systems. Nevertheless, the observa-
tion that some bisphosphonates have ∼1 µM IC50 values
versus P. falciparum in vitro prompted us to investigate
the activities of these species in vivo, with promising
results.
In a first series of experiments, we tested 3, 18, and
26 (pamidronate) at 25 mg/kg ip for 4 days. The mean
reductions in parasitemia are shown in Table 5, where
it can be seen that there is a 64% reduction in para-
sitemia for the benzyl bisphosphonate 18 and an even
larger 80% reduction in parasitemia with 3, the C5
bisphosphonate. This pattern of activity parallels their
in vitro activity, as seen in Table 1. Next, we investi-
gated the more active bisphosphonates 4 and 6, this
time using a reduced dosing scheme (Table 5). With 4,
there was a 48% reduction in parasitemia at 10 mg/kg
× 4 ip and with 6 there was a 75% reduction at 10 mg/
kg × 2 ip, although toxicity (as evidenced by panting
and horripilation) was noted with 6. Nevertheless, the
64% reduction seen with 18 and the 80% reduction in
P . ber gh ei Testin g in Vivo. To assess which types
of compounds might have potential for further develop-
ment as antimalarial agents, we derived the therapeutic
index values shown in Table 1, and IC50 vs LD50 and
IC50 vs TI values are shown graphically in parts A and
B of Figure 8, respectively. As may be seen in Figure
8A, there is no correlation between the IC50 (P. falci-
parum) and LD50 (KB cell line) results. From the
therapeutic index/IC50 results (Figure 8B), it is clear

184 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 1
Ghosh et al.
PFK inhibition for 4 is ∼3 mM, a factor of 100 times
greater. And in another example, the IC50 for E. his-
tolytica PFK inhibition by risedronate (20) is >3 mM,
although 20 is a ∼70 µM E. histolytica growth inhibitor
and a known potent FPP synthase inhibitor (of both
human and trypanosomatid enzymes). The IC50 values
for each of these compounds in inhibiting E. histolytica
growth (from Table 1) are also shown in Table 6 and
are essentially uncorrelated with PFK inhibition (R2 )
0.207, P-value ) 0.160, n ) 11). Some caution must be
used, however, in trying to directly compare IC50s for
growth and PFK inhibition, since several compounds
tested were found to be competitive inhibitors with
respect to pyrophosphate (PPi), so the actual inhibitory
potency in vivo will depend on the intracellular concen-
tration of PPi, which is not known. Nevertheless, the
observation that the relative potency of the bisphospho-
nates as PFK inhibitors was uncorrelated with their
ability to inhibit growth makes it unlikely that phos-
phofructokinase is the major bisphosphonate target in
E. histolytica, although it may be a secondary target for
the alkyl bisphosphonates.
E. histolytica also employs a pyrophosphate-depend-
ent pyruvate phosphate dikinase (PPDK) that catalyzes
the reversible conversion of pyruvate to phospho-
enolpyruvate. Compounds 17, 36, 73, 75-77, 79, 84, 93,
and 94 were tested against a recombinant E. histolytica
PPDK. The lowest resultant IC50s were greater than 60
µM, meaning that this enzyme is unlikely to be the
target for the bisphosphosphonates (Ruy Pe´rez-Mont-
fort, private communication).
F igu r e 9. Dose-response curves for selected bisphosphonates
against an expressed E. histolytica ATP PFK: 8, 9; 6, b; 12,
0; 4, O. IC50 values deduced from these and similar plots for
other compounds are given in Table 4.
Ta ble 6. Effects of Bisphosphonates on the Activity of an
Expressed E. histolytica Phosphofructokinase
IC50, µM
IC50, µM
PFK
drug activitya
E. histolyticab
PFK
E. histolyticab
growth
drug activitya
growth
7
8
6
20
40
70
11
12 1000
12.4
29.5
31.1
53.6
73.5
63.7
13.3
6.6
156.8
>200
4
2500-3000
44 >3000
31 >3000
20 >3000
77 90
3
18 900-1000
600-1200
Likewise, 4, 7, 20, and 37 were tested for activity
against another ATP-producing enzyme, a recombinant
Giardia lamblia acetyl-CoA synthetase (ADP-forming).
This enzyme is known in only two species, G. lamblia
and E. histolytica (NCBI accession numbers AAD44021
and AAF88064, respectively), and there is moderate
homology between the two enzymes (38% identity and
58% similarity; BLAST, http://workbench.sdsc.edu). The
enzyme converts acetyl-CoA plus ADP to acetate plus
ATP. The IC50 values were all >1 mM (Lidya´ Sanchez,
private communication), making this enzyme an un-
likely drug target. Thus, while we have not yet fully
established the bisphosphonate targets in these organ-
isms, several of the active compounds studied are known
to be inhibitors of either bone resorption or human or
parasite (trypanosomatid) FPP synthases, while activity
against phosphofructokinase, PPDK, and acetyl-CoA
synthetase (ADP-forming) are very weak.
a
Assay conditions: 50 M Ktes (pH 7.0), 100 mM KC1, 0.1 mM
EDTA, 1 mM DTT, 2 mM MgCl2, 0.5 mM Fru 6-P, 0.1 mM PPi,
b
30 °C. E. histolytica growth inhibition, from Table 1.
parasitemia observed with 3 are, again, very promising
initial in vivo results.
Oth er P ossible Bisp h osp h on a te Ta r gets. Finally,
we briefly consider the nature of some other possible
non-FPP synthase targets for the bisphosphonates. In
earlier studies, it was proposed that a pyrophosphate-
dependent phosphofructokinase was the bisphosphonate
target in E. histolytica, with a 40 µM Ki being reported
for 1-hydroxynonane-1,1-bisphosphonate.4 In more re-
cent work, a much more active PFK has been isolated
from E. histolytica,35,36 so we investigated the inhibition
of this enzyme using 11 different bisphosphonates.
Phosphofructokinase is a potentially attractive target
for growth inhibition in both E. histolytica and P.
falciparum, since both organisms are reliant on glyco-
lysis for ATP production, and it appears that the
recently discovered enzyme is the one that is primarily
responsible for PFK activity, at least in E. histolytica.36
We show in Figure 9 enzyme inhibition as a function
of bisphosphonate concentration for a selection of com-
pounds, and in Table 6, we show all the IC50 values
derived from the experimental data. As may be seen
from the results in Table 6, the most active PFK
inhibitors are the C9-C11 n-alkyl bisphosphonates,
which have IC50 values in the range 20-70 µM. The two
most active species are 7 and 8, but these IC50s are J2
times higher than their IC50s against E. histolytica.
Likewise, the IC50 for E. histolytica growth inhibition
for e.g. 4 is ∼30 µM, while the IC50 for E. histolytica
Con clu sion s
The results we have described above are of interest
for several reasons. First, we have investigated the
effects of 102 bisphosphonates on the growth of E.
histolytica. Several of the most promising compounds
in in vitro inhibition appear to be those with biphenyl
(76, 77), n-alkyl (6, 7), arylalkyl (35, 36), and phenoxy-
alkyl (73) side chains. A common characteristic of these
compounds is the presence of long side chains (eight or
more heavy atoms in a continuous chain). We also found
an unusual pattern of reactivity with the bisphospho-
nates in that many potent nitrogen-containing species
(such as pamidronate and risedronate) used as FPPS
inhibitors in bone resorption therapy had relatively little

Bisphosphonates as Antiparasitics
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 1 185
activity against E. histolytica proliferation in vitro, while
most of the active compounds had low activity versus
FPPS. Second, we found that n-alkylbisphosphonates
have much more pronounced activity, as low as 11 µM
in vitro. The results indicate that activity is chain length
dependent, with optimal activity being found for a ∼C9-
C10 chain length. Third, we found a correlation between
the activities of a series of pyridyl aminomethylene
bisphosphonates and the pKa values of the arylamine
bases. These results suggest that low pKa/deprotonated
side chains are important for activity, consistent with
the activity of the alkylbisphosphonates (which likewise
do not contain a positive charge feature). Fourth, we
investigated if there was any correlation between the
activity of a given bisphosphonate in inhibiting the
growth of a mammalian cell line and the activity of the
given bisphosphonate in inhibiting E. histolytica cell
growth. No correlation was evident. Fifth, we then used
these results to estimate therapeutic indices for each
compound. Five compounds found to have relatively low
IC50 values and relatively high therapeutic indices were
tested for their ability to delay the onset of liver abscess
formation in a hamster model. At 10 mg/kg per day ip,
we obtained a 68% reduction in liver abscess formation
for compound 7 with no overt toxicity. For in vivo
inhibition, 7 appears to be the best compound we have
investigated to date; however, compound 77 also looks
promising, since although it resulted in a smaller
decrease in abscess formation, there was no effect on
the host liver. Sixth, we investigated the activity of all
compounds against the growth of P. falciparum in vitro.
The most active compounds were the n-alkylbisphos-
phonates and we found a broadly similar dependence
of activity on chain length to that seen with E. his-
tolytica growth inhibition. Seventh, we made therapeu-
tic index/activity plots for P. falciparum growth inhibi-
tion and used these to choose a small set of compounds
for testing in vivo in a P. berghei BALB/c mouse
suppressive test. The best inhibitor was a simple
n-alkylbisphosphonate with a C5 side chain (3), which
gave an 80% reduction in parasitemia in vivo with no
overt toxicity. Eighth, we tested the idea that phospho-
fructokinase might be the principal bisphosphonate
drug target in E. histolytica. For the eleven compounds
studied, the potency of the bisphosphonates as phos-
phofructokinase inhibitors was uncorrelated with their
ability to inhibit growth. The IC50 values for phospho-
fructokinase inhibition were also higher than the IC50s
for E. histolytica growth inhibition. Similar (negative)
results were obtained for PPDK and acetylCoA synthase
(ADP-forming). Taken together, these results are in
contrast to many previous results obtained on T. cruzi,
T. brucei, T. gondii, L. donovani, and L. mexicana,
where nitrogen-containing bisphosphonates, such as
pamidronate and risedronate, have been found to have
the most potent antiparasitic activity, and suggest the
importance of neutral side chains and extended hydro-
phobic interactions for optimal activity. The observation
of considerable in vivo activity with both E. histolytica
(liver abscess formation) and P. berghei (reduction in
parasitemia) with some bisphosphonates suggests that
these compounds represent potentially interesting leads
for the development of antiamebic and antimalarial
drugs.
Exp er im en ta l Section
Bisp h osp h on a tes. We used the Merck method37 for pro-
duction of compounds 1-3, 5, 7, 8, 16, and 18-30.
Compounds 4, 6, and 9-15 were synthesized following the
method of Lecouvey et al.:38
Compounds 17 and 32-36 were produced using the method
of Mikhalin et al.:39
The synthesis of 31 has been described in detail elsewhere.40
The 1,2-bisphosphonates 37-39 were produced using the
method described by Allen et al.:41
The aminomethylene bisphosphonates 40-97, 101, and 102
were made following the method of Soloducho et al.:42
Compounds 98-100 were produced via the method of
Hutchinson and Thornton:43
An exception to the synthesis procedures above is compound
93, which was orignally intended to be synthesized as the
methoxy pyridine derivative. During the hydrolysis step, the
methoxy group was removed by HCl, leaving the hydroxy
compound 93, which can also exist as the tautomers A and B,
below:
The purity of compounds 1-8, 16, 18-31, 42-44, 50, 65,
86-88, 91, and 102 was verified by H/C/N microanalysis and/
or 1H, 13C and 31P NMR spectroscopy, as reported previ-
ously.21,24,40 For compounds 9-15, 17, 32-41, 45-49, 51-64,
66-85, 89, 90, and 92-101 we used H/C/N microanalysis and
1H and 31P NMR. For convenience, a detailed exemplary

186 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 1
Ghosh et al.
synthesis for each method is described in the Supporting
Information.
Sciences Foundation to I.B. and T.N., and by a grant
for Precursory Research for Embryonic Science and
Technology, J apan Science and Technology Corporation
to T.N. G.M. is a USPHS NRSA Postdoctoral Fellow
(NIH Grant GM-65782). We thank B. Weseloh and I.
Hennings for skillful technical assistance. We also thank
L. B. Sa´nchez, R. Pe´rez-Montfort, I. Anderson, R.
Docampo, G. Cameron, and R. Pink (WHO/Tibotec) for
helpful comments and advice and for providing their
unpublished results.
In Vitr o Testin g. E. h istolytica . E. histolytica trophozoites
of the isolate HM-1:IMSS were cultured axenically in TYI-33
medium.5 Inocula for experimental tubes and their controls
were taken from stocks still in the exponential growth phase.
The amoebae were cultured in 96-well plates under anaerobic
conditions. Various concentrations of bisphosphonates were
added to the growth medium at time t ) 0 (0, 5, 10, 25, 50,
100, and 150 µM) and the cells incubated for 90 h. At that
time, [3H]hypoxanthine (0.1 µCi) was added, and the cells were
cultured for a further 6 h. The cells were then lysed by
submitting them to two freeze (-70 °C) and thaw (37 °C) cycles
and harvested using a Micro Cell Harvester (Skatron Instru-
ments). During harvesting, the 3H-labeled DNA was spotted
onto a filter and radioactivity was counted by using a 1205
Betaplate counter (Walac).
Su p p or tin g In for m a tion Ava ila ble: Typical growth
inhibition curves for E. histolytica and P. falciparum and
exemplary syntheses. This material is available free of charge
via the Internet at http://pubs.acs.org.
P . fa lcipa r u m . P. falciparum (chloroquine-sensitive strain
3D7) was maintained in human A+ erythrocytes in RPMI1640
medium supplemented with Albumax II at 37 °C in a 5% CO2-
air mixture. P. falciparum intraerythrocytic cultures were set
up as above, with 1% ring stage parasitemia, 2.5% hematocrit,
in triplicate in 100 µL of medium in 96-well, flat-bottomed
Microtest III tissue culture plates. Drugs were added in 3-fold
dilution series and cultures incubated for a total of 48 h at 37
°C in a 5% CO2-air mixture. After 24 h, [3H]hypoxanthine (0.2
µCi) was added to each well. At the end of the assay, plates
were rapidly freeze-thawed (3×), cells were harvested using
a Tomtec Mach III cell harvester onto a 96-well format filter-
mat, and Meltilex solid scintillant (both Wallac, Finland) was
added prior to reading in a Microbeta 1450 scintillation counter
(Wallac, Finland) at 1 min per well.
Ha m ster Exp er im en ts. Hamsters were challenged with
the direct injection of 5 × 105 E. histolytica trophozoites into
the left lobe of the liver (day 0). Animals were treated with ip
administration of compounds (10 mg/kg) dissolved in 0.1 mL
of 100% DMSO once a day for 5 days starting on day 1.
Animals were sacrificed on day 6 and amebic liver abscess and
normal liver weights determined.
Refer en ces
(1) Phillips, R. S. Current status of malaria and potential for control.
Clin. Microbio. Rev. 2001, 14, 208-226.
(2) Schuster, H.; Chiodini, P. L. Parasitic infections of the intestine.
Curr. Opin. Infect. Dis. 2001, 14, 587-591.
(3) Adagu, I. S.; Nolder, D.; Warhurst, D. C.; Rossignol, J .-F. In vitro
activity of nitazoxanide and related compounds against isolates
of Giardia intestinalis, Entamoeba histolytica and Trichomonas
vaginalis. J . Antimicrob. Chemother. 2002, 49, 103-111.
(4) Eubank, W. B.; Reeves, R. E. Analogue inhibitors for the
pyrophosphate-dependent phosphofructokinase of Entamoeba
histolytica and their effect on culture growth. J . Parasitol. 1982,
68, 599-602.
(5) Bruchhaus, I.; J acobs, T.; Denart, M.; Tannich, E. Pyrophos-
phate-dependent phosphofructokinase of Entamoeba histolyti-
ca: Molecular cloning, recombinant expression and inhibition
by pyrophosphate analogues. Biochem. J . 1996, 316, 57-63.
(6) Cromartie, T. H.; Fisher, K. J .; Grossman, J . N. The discovery
of a novel site of action of herbicidal bisphosphonates. Pestic.
Biochem. Physiol. 1999, 63, 114-126.
(7) Martin, M. B.; Arnold, W.; Heath, H. T. I.; Urbina, J . A.; Oldfield,
E. Nitrogen-containing bisphosphonates as carbocation transi-
tion state analogues for isoprenoid biosynthesis. Biochem. Bio-
phys. Res. Commun. 1999, 263, 754-758.
P . ber gh ei Exp er im en ts. Female BALB/c mice (18-20 g)
infected with P. berghei (ANKA strain) were used. Blood was
taken from donor mice; the serum was diluted in heat-
inactivated fetal calf serum to a parasitemia of 1% (the
equivalent of 1 × 107 infected erythrocytes), and 0.2 mL was
administered iv to each mouse. Mice were randomly sorted
into groups of five and dosing commenced 2 h postinoculation.
Compounds were prepared as stock solutions at 10 mg/mL in
10% DMSO/PBS. The control drug, chloroquine, was given ip
every day for 4 days. Bisphosphonate drugs were given in a
0.2 mL bolus every day for the number of days indicated and
at the concentrations indicated (Table 5).
P KF Exp r ession a n d In h ibition . Isolation of E. his-
tolytica PPi-PFK expressed in E. coli bearing a vector consist-
ing of the PPi-PFK gene cloned into the prokaryotic expression
plasmid palter-Ex1 (Promega) has been described previously.44
Homogeneity of the preparation was demonstrated by the
presence of a single band upon SDS-PAGE.
Enzyme activity of PPi-PFK was assayed spectrophoto-
metrically at 30 °C in an assay solution that contained 50 mM
Tes/KOH (pH 7.0), 2 mM MgCl2, 1 mM EDTA, 2 mM DTT,
0.5 mM Fru 6-P, 0.1 mM PPi, 0.2 mM NADH, and 2-6 units
each of aldolase, triosephosphate isomerase, and glycerol
3-phosphate dehydrogenase. These auxiliary enzymes were
dialyzed against 50 mM Tes/KOH (pH 7.0), 1 mM EDTA prior
to use. Varying concentrations of potential inhibitors were
added to determine IC50s. With several inhibitors, the mode
of inhibition was examined by using a fixed concentration of
inhibitor with varying concentrations of substrates.
(8) van Beek, E.; Pieterman, E.; Cohen, L.; Lowik, C.; Papapoulos,
S. Farnesyl pyrophosphate synthase is the molecular target of
nitrogen-containing bisphosphonates. Biochem. Biophys. Res.
Commun. 1999, 264, 108-111.
(9) Keller, R. K.; Fliesler, S. J . Mechanism of aminobisphosphonate
action: Characterization of alendronate inhibition of the iso-
prenoid pathway. Biochem. Biophys. Res. Commun. 1999, 266,
560-563.
(10) Grove, J . E.; Brown, R. J .; Watts, D. J . The intracellular target
for the antiresorptive aminobisphosphonate drugs in Dictyoste-
lium discoideum is the enzyme farnesyl diphosphate synthase.
J . Bone Miner. Res. 2000, 15, 971-981.
(11) Bergstrom, J . D.; Bostedor, R. G.; Masarachia, P. J .; Reszka,
A.; Rodan, G. Alendronate is a specific nanomolar inhibitor of
farnesyl diphosphate synthase. Arch. Biochem. Biophys. 2000,
373, 231-241.
(12) Dunford, J . E.; Thompson, K.; Coxon, F. P.; Luckman, S. P.;
Hahn, F. M.; et al. Structure-activity relationships for inhibition
of farnesyl diphosphate synthase in vitro and inhibition of bone
resorption in vivo by nitrogen-containing bisphosphonates. J .
Pharmacol. Exp. Ther. 2001, 296, 235-242.
(13) Rodan, G. A.; Martin, T. J . Therapeutic approaches to bone
diseases. Science 2000, 289, 1508-1514.
(14) Derenne, S.; Amiot, M.; Barille´, S.; Collette, M.; Robillard, N.;
et al. Zoledronate is a potent inhibitor of myeloma cell growth
and secretion of IL-6 and MMP-1 by the tumoral environment.
J . Bone Miner. Res. 1999, 14, 2048-2056.
(15) Senaratne, S. G.; Pirlanov, G.; Mansi, J . L.; Arnett, T. R.;
Colston, K. W. Bisphosphonates induce apoptosis in human
breast cancer cell lines. Br. J Cancer 2000, 82, 1459-1468.
(16) Tassone, P.; Forciniti, S.; Galea, E.; Morrone, G.; Turco, M. C.;
et al. Growth inhibition and synergistic induction of apoptosis
by zoledronate and dexamethasone in human myeloma cell lines.
Leukemia 2000, 14, 841-844.
(17) Lee, M. V.; Fong, E. M.; Singer, F. R.; Guenette, R. S. Bisphos-
phonate treatment inhibits the growth of prostate cancer cells.
Cancer Res. 2001, 61, 2602-2608.
(18) Urbina, J . A.; Moreno, B.; Vierkotter, S.; Oldfield, E.; Payares,
G.; et al. Trypanosoma cruzi contains major pyrophosphate
stores, and its growth in vitro and in vivo is blocked by
pyrophosphate analogues. J . Biol. Chem. 1999, 274, 33609-
33615.
Ack n ow led gm en t. This work was supported by the
United States Public Health Service (NIH Grant GM-
65307 to E.O.), by the UNDP/World Bank/WHO Special
Program for Research and Training in Tropical Diseases
(TDR) (E.O., S.L.C.), by a grant from the J apan Health

Bisphosphonates as Antiparasitics
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 1 187
(19) Yardley, V.; Khan, A. A.; Martin, M. B.; Slifer, T. R.; Araujo, F.
G.; et al. In vivo activity of the farnesyl pyrophosphate synthase
inhibitors alendronate, pamidronate and risedronate against
Leishmanai donovani and Toxoplasma gondii. Antimicrob.
Agents Chemother. 2001, 46, 929-931.
(20) Rodriguez, N.; Bailey, B. N.; Martin, M. B.; Oldfield, E.; Urbina,
J . A.; et al. Radical cure of experimental cutaneous leishmaniasis
using pamidronate. J . Infect. Diseases 2002, 186, 138-140.
(21) Martin, M. B.; Grimley, J . S.; Lewis, J . C.; Heath, H. T., III.;
Bailey, B. N.; et al. Bisphosphonates inhibit the growth of
Trypanosoma brucei, Trypanosoma cruzi, Leishmania donovani,
Toxoplasma gondii, and Plasmodium falciparum: A potential
route to chemotherapy. J . Med. Chem. 2001, 44, 909-916.
(22) Moreno, B.; Bailey, B. N.; Luo, S.; Martin, M. B.; Kuhlenschmidt,
M.; et al. 31P NMR of apicomplexans and the effects of risedr-
onate on Cryptosporidium parvum growth. Biochem. Biophys.
Res. Commun. 2001, 284, 632-637.
(32) Kissinger, J . C.; et al. PlasmoDB Nature 2002, 419, 490-492.
(33) McIntosh, M. T.; Vaidya, A. B. Vacuolar type H+ pumping
pyrophosphatases of parasitic protozoa. Int. J . Parasit. 2002, 32,
1-14.
(34) Rogers, M. J .; Xiong, X.; J i, X.; Mo¨nkko¨nen, J .; Russel, R. G. G.;
Williamson, M. P.; Ebetino, F. H.; Watts, D. J . Inhibition of
growth of Dictyostelium discoideum amoebae by bisphosphonate
drugs is dependent on cellular uptake. Pharm. Res. 1997, 14,
625-630.
(35) Chi, A.; Kemp, R. G. The primordial high energy compound: ATP
or inorganic pyrophosphate? J . Biol. Chem. 2000, 275, 35677-
35679.
(36) Chi, A. S.; Deng, Z.; Albach, R. A.; Kemp, R. G. The two
phosphofructokinase gene products of Entamoeba histolytica. J .
Biol. Chem. 2001, 276, 19974-19981.
(37) Kieczykowski, G. R.; J obson, R. B.; Melillo, D. G.; Reinhold, D.
F.; Grenda, V. J .; et al. Preparation of (4-amino-1-hydroxybu-
tylidene) bisphosphonic acid sodium salt, MK-217 (alendronate
sodium). An improved procedure for the preparation of 1-hy-
droxy-1,1-bisphosphonic acids. J . Org. Chem. 1995, 60, 8310-
8312.
(23) Szajnman, S. H.; Bailey, B. N.; Docampo, R.; Rodriguez, J . B.
Bisphosphonates derived from fatty acids are potent growth
inhibitors of Trypanosoma cruzi. Bioorg. Med. Chem. Lett. 2001,
11, 789-792.
(24) Martin, M. B.; Sanders, J . M.; Kendrick, H.; de Luca-Fradley,
K.; Lewis, J . C.; Grimley, J . S.; et al. Activity of bisphosphonates
against Trypanosoma brucei rhodesiense. J . Med. Chem. 2002,
45, 2904-2914.
(38) Lecouvey, M.; Mallard, I.; Bailly, T.; Burgada, R.; Leroux, Y. A
mild and efficient one-pot synthesis of 1-hydroxymethylene-1,1-
bisphosphonic acids. Preparation of new tripod ligands. Tetra-
hedron Lett. 2001, 42, 8475-8478.
(39) Mikhalin, N. V.; Alferiev, I. S.; Kotlyarevskii, I. L.; Krasnukhina,
A. V. Method of preparation of higher 1-hydroxyalkylidene-1,1-
diphosphonic acids or their mixtures or salts. Russian Patent
SU 1,719,405; Chem. Abstr. 1992, 117, 234254.
(40) Szabo, C. M.; Matsumura, Y.; Fukura, S.; Martin, M. B.; Sanders,
J . M.; et al. Inhibition of geranylgeranyl diphosphate synthase
by bisphosphonates and diphosphates: A potential route to new
bone anti-resorption and anti-parasitic agents. J . Med. Chem.
2002, 45, 2185-2196.
(41) Allen J r, A.; Manke, D. R.; Lin, W. Synthesis of functional
bisphosphonates via new palladium-catalyzed bis-hydrophos-
phorylation reactions. Tetrahedron Lett. 2000, 41, 151-154.
(42) Soloducho, J .; Gancarz, R.; Wieczorek, P.; Korf, J .; Hafner, J .;
et al. Preparation of novel derivatives of (aminomethylene)bis-
(phosphonic acid) as herbicides. Patent PL93-298436, 1997.
(43) Hutchinson, D. W.; Thornton, D. M. Michael addition reactions
of ethenylidenebisphosphonates. J . Organomet. Chem. 1988, 346,
341-348.
(25) Felix, R.; Fleisch, H. The effect of pyrophosphate and diphos-
phonates on calcium transport in red cells. Experientia 1977,
33, 1003-1005.
(26) Montalvetti, A.; Fernandez, A.; Sanders, J . M.; Ghosh, S.; Van
Brussel, E.; Oldfield, E. Docampo, R. Farnesyl pyrophosphate
synthase is an essential enzyme in Trypanosoma brucei: In vitro
RNA interference and in vivo inhibition studies. J . Biol. Chem.
2003, 278, 17075-17083.
(27) Sanders, J . M.; Go´mez, A. O.; Mao, J .; Meints, G. A.; Van Brussel,
E. M.; Burzynska, A.; Kafarski, P.; Gonza´lez-Pacanowska, D.;
Oldfield, E. 3D-QSAR investigations of the inhibition of Leish-
mania major farnesyl pyrophosphate synthase by bisphospho-
nates. J . Med. Chem. 2003, 46, 5171-5183.
(28) The TIGR Entamoeba histolytica Genome Project: http://ww-
w.tigr.org/tdb/e2k1/eha1/index.shtml.
(29) Widler, L.; J aeggi, K. A.; Glatt, M.; Muller, K.; Bachmann, R.;
et al. Highly potent geminal bisphosphonates. From pamidronate
disodium (Aredia) to zoledronic acid (Zometa). J . Med. Chem.
2002, 45, 3721-3738.
(30) Suzuki, F.; Yoshihiro, F.; Mizutani, H.; Funabashi, C.; Ikai, T.;
et al. N-Pyridylaminomethylenediphosphonsaureverbindungen.
Ger Offen. 1979, 2.831.578.
(44) Deng, Z.; Huang, M.; Singh, K.; Albach, R. A.; Latshaw, S. P.;
Chang, K.-P.; Kemp, R. G. Cloning and expression of the gene
for the active PPi-dependent phosphofructokinase of Entamoeba
histolytica. Biochem. J . 1998, 329, 659-664.
(31) Advanced Chemistry Development (ACD) Software Solaris
V4.67.
J M030084X