Fosmidomycin Uptake into Plasmodium and Babesia-
Infected Erythrocytes Is Facilitated by Parasite-Induced
New Permeability Pathways
Stefan Baumeister1., Jochen Wiesner2., Armin Reichenberg2, Martin Hintz2, Sven Bietz1, Omar S. Harb3,
David S. Roos3, Maximilian Kordes4, Johannes Friesen4, Kai Matuschewski4, Klaus Lingelbach1, Hassan
Jomaa2, Frank Seeber1,5*
1Parasitologie, Fachbereich Biologie, Philipps-Universita ¨t, Marburg, Germany, 2Institut fu ¨r Klinische Immunologie und Transfusionsmedizin, Universita ¨tsklinikum Giessen
und Marburg GmbH, Giessen, Germany, 3Department of Biology, University of Pennsylvania, Philadelphia, Pennsylvania, United States of America, 4Parasitology Unit,
Max-Planck-Institute for Infection Biology, Berlin, Germany, 5Fachgebiet 16 Parasitologie, Robert-Koch-Institut, Berlin, Germany
Background: Highly charged compounds typically suffer from low membrane permeability and thus are generally regarded
as sub-optimal drug candidates. Nonetheless, the highly charged drug fosmidomycin and its more active methyl-derivative
FR900098 have proven parasiticidal activity against erythrocytic stages of the malaria parasite Plasmodium falciparum. Both
compounds target the isoprenoid biosynthesis pathway present in bacteria and plastid-bearing organisms, like
apicomplexan parasites. Surprisingly, the compounds are inactive against a range of apicomplexans replicating in
nucleated cells, including Toxoplasma gondii.
Methodology/Principal Findings: Since non-infected erythrocytes are impermeable for FR90098, we hypothesized that
these drugs are taken up only by erythrocytes infected with Plasmodium. We provide evidence that radiolabeled FR900098
accumulates in theses cells as a consequence of parasite-induced new properties of the host cell, which coincide with an
increased permeability of the erythrocyte membrane. Babesia divergens, a related parasite that also infects human
erythrocytes and is also known to induce an increase in membrane permeability, displays a similar susceptibility and uptake
behavior with regard to the drug. In contrast, Toxoplasma gondii-infected cells do apparently not take up the compounds,
and the drugs are inactive against the liver stages of Plasmodium berghei, a mouse malaria parasite.
Conclusions/Significance: Our findings provide an explanation for the observed differences in activity of fosmidomycin and
FR900098 against different Apicomplexa. These results have important implications for future screens aimed at finding new
and safe molecular entities active against P. falciparum and related parasites. Our data provide further evidence that
parasite-induced new permeability pathways may be exploited as routes for drug delivery.
Citation: Baumeister S, Wiesner J, Reichenberg A, Hintz M, Bietz S, et al. (2011) Fosmidomycin Uptake into Plasmodium and Babesia-Infected Erythrocytes Is
Facilitated by Parasite-Induced New Permeability Pathways. PLoS ONE 6(5): e19334. doi:10.1371/journal.pone.0019334
Editor: Georges Snounou, Universite ´ Pierre et Marie Curie, France
Received January 24, 2011; Accepted March 27, 2011; Published May 4, 2011
Copyright: ? 2011 Baumeister et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This study was supported by the Deutsche Forschungsgemeinschaft (grants SFB 593 to S.B. and K.L., SE622/4-3 to F.S., and WI 2050/2-1 to J.W.) and by
a grant from the European Commission to H.J. (Integrated Project #018834) and via the EviMalaR network of excellence to K.L. and K.M. The funders had no role
in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: firstname.lastname@example.org
. These authors contributed equally to this work.
The antibiotic Fosmidomycin (Fos; 3-[formyl(hydroxy)amino]-
propylphosphonic acid; CID 572) and its derivative FR900098
162204) were described previously as inhibitors of DOXP
reductoisomerase (Dxr), the second enzyme in the biosynthesis
pathway of isoprenoids in P. falciparum, inhibiting its in vitro and in
vivo growth at high nanomolar concentrations . Several studies
have confirmed the presence of the other individual enzymatic
steps in this organism and its essential nature for parasite survival
[2,3]. In combination with the antibiotic drug clindamycin, Fos
has already been tested in phase II clinical trials against
uncomplicated malaria with good success [4,5,6,7,8]. Fos has an
exceptional safety profile in humans, even when given repeatedly
at a dose of 8 g/day . There is an ongoing need for new, safe
and affordable anti-malarials, in particular after reports of
decreased sensitivity against artemisinin-based monotherapy have
appeared in the literature .
Isoprenoids are a large and diverse group of natural compounds
fulfilling a large number of diverse cellular functions in all
biological systems, such as cell signaling processes, protein
modifications (prenylation), synthesis of the co-factor ubiquinone
and modifications of tRNAs, amongst others . The basic
building blocks for all these structures are isopentenyl diphosphate
(IPP) and its isomeric form, dimethylallyl diphosphate (DMAPP).
PLoS ONE | www.plosone.org1 May 2011 | Volume 6 | Issue 5 | e19334
Two alternative routes for their synthesis are known: most
eubacteria and plants follow the so-called 1-deoxy-D-xylulose-5-
phosphate (DOXP) pathway (also called methylerythritol phos-
phate (MEP) pathway) whereas eukaryotes and archaebacteria
mostly use the mevalonate (MEV) pathway  (see Figure S1).
The two pathways are fundamentally different, starting from
different compounds and employing distinct enzymes leading to
specific intermediate products. Unlike humans, almost all
apicomplexan parasites, including Plasmodium falciparum, the
causative agent of human malaria, and Toxoplasma gondii, causing
toxoplasmosis, are now known to synthesize isoprenoids exclu-
sively via the DOXP pathway in the apicoplast, an essential,
metabolically active, reduced plastid of endosymbiotic descent
found in almost all Apicomplexa .
Bioinformatic analyses of the published genome sequences from
several Apicomplexa of human and veterinary medical importance
(Plasmodium, T. gondii, Neospora caninum, Theileria and Babesia bovis)
have unequivocally shown the presence of all genes of the DOXP
pathway in all these organisms [13,14] (see also Table S1).
Notably, the Dxr protein sequences are highly similar among these
organisms, and residues known from 3D-structures of bacterial
Dxr to be important for Fos binding are well conserved (Figure
S2). They can be superimposed onto the respective amino acids in
a 3D-model of the T. gondii sequence . Given these facts it
could be assumed that Fos and FR are also active against those
parasites. Surprisingly, however, several reports have shown that
Fos does not kill T. gondii, Eimeria tenella and T. parva, even at very
high (.100) micromolar concentrations [1,14,15,16].
There are numerous reasons why drugs may be ineffective, but
the most obvious one is a failure of its uptake into the infected host
cell. We therefore started to investigate whether basic differences
exist in Fos uptake between T. gondii-infected fibroblasts and cells
infected with susceptible parasites, namely P. falciparum-infected
erythrocytes. The latter induce alterations in the permeability of
the red blood cell (RBC) plasma membrane for a variety of
different solutes and which are collectively called new permeability
pathways (NPP) .
Here we provide evidence that these pathways, which appear to
be absent in non-infected erythrocytes, greatly facilitate uptake of
the respective drugs into the infected cell. Likewise, Babesia
divergens, another apicomplexan parasite of erythrocytes that is also
known to possess NPP-like activities, was found to be susceptible to
Fos and its derivative FR. Again, Fos uptake was NPP-mediated.
Our results are consistent with the view that parasite-induced
changes of the host erythrocyte membrane are pre-requisites for
the uptake of these drugs into infected RBC. At the same time they
provide a likely explanation for the failure to kill other
apicomplexans where NPP appear to be absent or not required.
Identification of DOXP reductoisomerase activity in T.
gondii cell lysates and its inhibition by fosmidoymcin
In initial experiments we wished to formally prove that Fos-
inhibitable Dxr activity is present in T. gondii since functional data
on Dxr activity in T. gondii have not been reported so far. To
evaluate whether native Dxr from T. gondii (TgDxr) is an active
enzyme a highly specific and sensitive radiometric assay (see
Methods) was performed on lysates of tachyzoite-infected host
cells. In this assay, MEP formed from DOXP and NADPH by Dxr
is further converted into CDP-ME by recombinant E. coli YgbP,
the enzyme performing the next step in the DOXP pathway,
thereby incorporating the radioactively labeled phosphorus atom
from [a-32P]CTP (Figure 1A; Figure S1). The results show that (i)
significant Dxr acitivity is present only in lysates of cells infected
with tachyzoites (Figure 1B, compare lane 1 with lane 5) and (ii)
that the activity can be completely inhibited by Fos in a dose-
dependent manner (Figure 1B, lanes 2–4; 1C). We conclude that
TgDxr is an active enzyme that can be inhibited by low
micromolar concentrations of Fos in vitro.
DOXP reductoisomerase localizes to the apicoplast in T.
gondii tachyzoites and P. falciparum blood stages
We next wanted to confirm that Dxr resides in the apicoplast of
apicomplexan parasites. Notably, available proteomics data do not
provide direct evidence for the expression of Dxr neither in P.
falciparum nor in T. gondii (see Table S1), and neither has in situ
localization of Dxr in P. falciparum or in T. gondii been reported so
far. Previous targeting experiments had shown that the N-terminal
leader peptide of P. falciparum Dxr fused to GFP transported this
construct to the apicoplast of T. gondii tachyzoites . To prove
expression of Dxr in the apicoplast, polyclonal antibodies were
raised against recombinant PfDxr (Figure 2 and Figure S11) and
antibody staining was performed on intracellular T. gondii
tachyzoites and P. falciparum blood stages. Discrete anti-PfDxr
reactivity can be detected in T. gondii tachyzoites (Figure 2A) as
well as P. falciparum blood stages (schizonts, Figure 2B; for other
stages see Figures S3, S4, S5, S6, S7). Using co-localization with
the apicoplast-resident acyl carrier protein (PfACP) in the case of
P. falciparum and direct staining of the apicoplast DNA for T. gondii,
these structures were clearly identified as the apicoplast in both
organisms, indicating that Dxr is expressed in this organelle. For P.
falciparum this is in agreement with results showing that most
downstream intermediates of Dxr could be detected in all blood
T. gondii-infected fibroblasts do not take up FR
Having shown that T. gondii possesses Dxr activity that can be
inhibited by Fos we directly addressed the possibility that the
failure of the drug to kill tachyzoites could be due to its reduced
uptake into cells infected with T. gondii. To this end we synthesized
radiolabeled drug for transport studies. [14C]FR is much easier to
synthesize than [14C]Fos, and FR differs from Fos by a single
methyl group (Figure 1D) which neither alters its hydrophilicity
nor other physico-chemical properties (Table S2) nor overall shape
(Figure S8). FR is also twice as potent in inhibiting PfDxr than Fos
When we followed [14C]FR uptake into human foreskin
fibroblasts (HFF) for 15 min no significant cell-associated radio-
activity was seen for [14C]FR (Figure 3). However, [3H]-L-
glutamate ([3H]Glu) that served as control (see below) was readily
taken up, as described previously . Heavy infection of cells
(.50%) with tachyzoites did not result in higher [14C]FR counts,
in stark contrast to four-fold increased values for [3H]Glu
(Figure 3). This is consistent with the observed slight up-regulation
of transcripts of the epithelial high affinity glutamate transporter
EAAT3 upon T. gondii infection . Intracellular accumulation of
[3H]Glu could be significantly inhibited by pre-incubation with
unlabeled L-Glu. Extending the labeling period with [14C]FR to
2 hours in similar preliminary experiments did not result in
significantly different dpm between these two time points for both,
infected and non-infected HFF (unpublished observations). To
rule out that rapid efflux of [14C]FR via the P-glycoprotein efflux
pump was responsible for the observed results assays were
performed in the presence of the inhibitor verapamil; however,
this did not lead to significantly increased [14C]FR counts in those
cells (data not shown). Taken together, these experiments indicate
that T. gondii-infected fibroblasts do not take up FR.
Fosmidomycin Uptake into Infected Erythrocytes
PLoS ONE | www.plosone.org2May 2011 | Volume 6 | Issue 5 | e19334
Uptake of FR into P. falciparum-infected erythrocytes
depends on functional new permeability pathways (NPP)
Since P. falciparum induces permeability changes of the host cell
for a variety of different substrates, we next studied [14C]FR
uptake into P. falciparum-infected human red blood cells (Pf-iRBC).
We observed a time-dependent increase of radioactivity in Pf-
iRBC (Figure 4A). Strikingly, non-infected cells (RBC) showed
only minor amounts of cell-associated counts. This largely
increased uptake behavior into iRBC is reminiscent of compound
entry via the so-called new permeability pathways (NPP). This is
illustrated by uptake of the known NPP substrate [14C]L-Glu
(Figure 4B). To test whether [14C]FR entry is indeed via the NPP,
we inhibited these pathways pharmacologically with 50 mM 5-
nitro-2-(3-phenylpropylamino)-benzoic acid (NPPB), a well-known
NPP inhibitor (Figure 4 A,B). Consistent with its entry via the
NPP, FR uptake was considerably inhibited (.50%) by NPPB
(Figure 4A, B; iRBC+NPPB). Similar results were obtained with
another NPP inhibitor, furosemide (data not shown). Both
compounds decreased [14C]FR uptake dose-dependently (Figure
It has been reported previously that limited protease treatment
of Pf-iRBC with chymotrypsin almost completely abrogates NPP
activity . Accordingly, chymotrypsinization of Pf-iRBC
resulted in a drastic decrease of FR uptake to almost background
levels (Figure 4A, iRBC+chymptrypsin), as did the known NPP-
substrate L-glutamate (Figure 4B). Together, these results clearly
implicate a role of NPP in FR uptake into P. falciparum-infected
Figure 1. Fosmidomycin-sensitive Dxr activity in T. gondii. A Principle of the coupled radiometric assay for Dxr activity measurement. MEP
formed by Dxr from DOXP and NADPH is converted into [32P]CDP-ME in a second reaction using recombinant E. coli YgbP (IspD) enzyme and
[a-32P]CTP. B Analysis of the Dxr assay with T. gondii lysate in the presence of different concentrations of Fos (lanes 2–4) and without drug (lane 1) by
autoradiography of the TLC plate after separation of the reaction products. The appearance of [32P]CDP-ME is indicative of Dxr activity (arrowhead).
The smear at the bottom is [a-32P]CTP (*), the one above (**) is [a-32P]CDP whereas the prominent spot below [32P]CDP-ME is caused by [a-32P]CMP
(***). The production of CDP and CMP from CTP is presumably due to the presence of phosphatases in the whole cell lysates, which explains the
weaker corresponding signals with the recombinant E. coli Dxr protein (lane 6). As controls, either host cell lysate (negative control, lane 5) or 10 pg of
purified recombinant E. coli Dxr protein (41 U/mg; positive control, lane 6) was used. C Densitometric evaluation of the [32P]CDP-ME signal from B.
AU, arbitrary units. D Comparison of the structures of the Dxr substrate DOXP with of Fos and FR.
Fosmidomycin Uptake into Infected Erythrocytes
PLoS ONE | www.plosone.org3 May 2011 | Volume 6 | Issue 5 | e19334
FR entry via NPP is also observed in Babesia-infected
Some apicomplexan parasites of the genus Babesia also infect
mammalian erythrocytes and are considered emerging human
pathogens . Recent data provided evidence for permeability
changes in Babesia divergens-infected human RBC (Bd-iRBC) .
To test if FR uptake via NPP-like mechanisms is a general feature
of Apicomplexa-infected RBC we used B. divergens previously
adapted to in vitro culture in human RBC to measure [14C]FR
uptake. The results show that Bd-iRBC, like Pf-iRBC, have
significantly higher cell-associated [14C]FR counts than non-
infected cells (Figure 5). Uptake of [14C]FR is not as prominent as
in P. falciparum-infected erythrocytes, most likely because of a lower
trophozoite stage parasitemia, which is due to the current lack of
methods for stage-specific enrichment of Babesia-infected RBC
(average parasitemia of 20% trophozoite stage Bd-iRBC, com-
pared to 80% trophozoite stage Pf-iRBC). In good agreement with
a direct role of NPP-like mechanisms in FR uptake, inhibition by
furosemide leads to a statistically significant decreased [14C]FR
uptake (P=0.011 by paired t-test).
Having shown that FR also enters Bd-iRBC it was of interest
whether the drug would also inhibit the intracellular growth of B.
Figure 2. Intracellular localization of Dxr in T. gondii tachyzoites and P. falciparum blood stages. Antibody staining was performed on
fixed parasites with a polyclonal antibody raised against recombinant P. falciparum Dxr. A Discrete anti-PfDxr reactivity (green) can be detected in T.
gondii tachyzoites in the apicoplast (I), co-localizing with organellar DNA (blue; arrowheads in II; overlaid in III). Co-localization is visualized in IV using
the ‘Colocalization’ plugin from the ImageJ software suite. White pixels indicate the summed-up overlapping co-localizing signals from III,
superimposed onto the phase contrast image of the parasites within the vacuole. B Schizont stage of P. falciparum: (I) phase contrast of iRBC; (II) anti-
PfACP (red); (III) anti-PfDxr (green); (IV) DAPI stain (blue); (V) merged images of II-IV (co-localization of anti-PfACP and anti-PfDxr (yellow); (VI) merged
images of I and of co-localized pixels (using the ‘Colocalization’ plugin applied to II and III). Co-localization of anti-PfACP and anti-PfDxr (white). Scale
Figure 3. Uptake studies of [14C]FR and [3H]Glu into human fibroblasts. Duplicates of cells (HFF, non-infected; TgHFF, .50% infected with
tachyzoites) were incubated in parallel in phosphate-free ‘‘extracellular buffer’’ (see Material and Methods) for 15 min at 37uC in the presence of
either [14C]FR or [3H]Glu, respectively, of the same specific activity (1.45 mCi/ml, equaling 25 mM drug concentration per assay). Cell-associated
radioactivity was determined by scintillation counting of an aliquot of lysed cells and dpm for each compound were determined according to the
manufacturer’s instructions (shown as dpm6SD). In some assays cells were preincubated for 30 min in 1 mM unlabelled FR or L-Glu to test for uptake
specificity (indicated by ‘+FR’ and ‘+Glu’). This is one representative of three similar experiments.
Fosmidomycin Uptake into Infected Erythrocytes
PLoS ONE | www.plosone.org4May 2011 | Volume 6 | Issue 5 | e19334
divergens at pharmacologically relevant concentrations. A sharp
reduction in parasite growth (10% of untreated control after 72 h)
could already be seen with 10 mM FR (Table 1). Our results are in
very good agreement with recently published data showing that
Fos inhibits the intraerythrocytic growth of two other Babesia
species (B. bovis and B. bigemina) in bovine RBC at such low
concentrations (reported IC50,5 mM; .
Failure of FR to inhibit growth of liver stage P. berghei
parasites in vitro and in vivo
To our knowledge, the growth inhibiting activity of Fos and FR
for Plasmodia has only been studied on blood stage parasites.
Based on published microarray data of the liver stages of the
rodent parasite P. yoelii many of the genes for the DOXP pathway
are assumed to be expressed at least as strongly in liver stage as in
blood stage parasites (; Figure S10). We therefore tested
whether the stages of P. berghei multiplying in nucleated liver cells
(exo-erythrocytic forms; EEFs) are sensitive to increasing FR
concentrations (Figure 6). Human hepatoma cells infected with
GFP-expressing P. berghei sporozoites were incubated with 1, 10
and 100 mM FR, respectively, for 48 h and parasite replication
was recorded. These drug concentrations are between 2 to 200-
fold above the IC50for P. falciparum blood stages . In no case
were we able to detect differences in EEF growth (Figure 6A, B).
We then infected groups of susceptible C57BL/6 mice with
10,000 P. berghei sporozoites each and subsequently treated them
intraperitoneally with four high doses of 250 mg/kg FR at 12 h
intervals. This high dose regimen is expected to result in sustained
high drug levels in the liver [25,26]. However, no significant
difference (P.0.05; unpaired student’s t-test) of parasite load, as
quantified by real-time PCR of P. berghei 18S RNA transcripts in
the liver of these animals, could be observed after 42 h when
compared to those of uninfected control animals (Figure 6C). In
contrast, a control group that had received 60 mg/kg primaquine
at 0 h und 24 h post-infection showed a vast decline in parasite
load, illustrating the susceptibility of this parasite strain to in vivo
drug treatment. To further prove that the batch of FR used in the
above experiments was able to kill susceptible parasites, growth
inhibition of blood stages was monitored in a group of mice that
had received doses of 75 mg/kg FR every 8–12 hours for 5 days
starting at day 3 after iRBC inoculation. Parasitemia in treated
animals was 0.22%60.38% 7 days after infection, whereas that of
untreated controls was 17.1%66.8% (Figure 6D). Taken together
these results indicate that susceptibility of Plasmodium to FR
treatment can only be observed when the parasites reside in
erythrocytes but not when multiplying in nucleated liver cells.
We provide evidence that only erythrocytes with an increased
plasma membrane permeability for a variety of solutes upon
Figure 4. Infection-induced, time-dependent increase of uptake of [14C]FR into P. falciparum parasitized erythrocytes. Non-infected
(RBC) and red blood cells infected with P. falciparum (Pf-iRBC), respectively, were incubated in RPMI-1640 medium for 60 min containing [14C]FR (A)
or [14C]Glu (B) of the same specific activity. At the indicated time points aliquots were taken and the radiolabeled tracer in the cellular fraction was
quantified (see Material and Methods). Uptake of labeled compounds was also evaluated in the presence of the NPP-inhibitor 5-nitro-2-(3-
phenylpropylamino)-benzoic acid (NPPB) at 10 mM, or by pre-treatment of Pf-iRBC with 1 mg/ml chymotrypsin for 1 h at 37uC. Similarly treated RBC
served as controls. Cpm6SD for three experiments are shown.
Figure 5. Increased uptake of [14C]FR into erythrocytes
infected with Babesia divergens (Bd-iRBC). Bd-iRBC were incubated
in the presence of [14C]FR at 37uC for 20 min before the cells were
harvested and the amount of intracellular [14C]FR determined. Non-
infected RBC and Bd-iRBC treated with 100 mM furosemide (Furo)
served as controls. Shown are the mean values of three independent
experiments (6SD). The data represent the amount of intracellular
[14C]FR in relation to the amount quantified for RBC (control).
* indicates a p-value of 0.011.
Fosmidomycin Uptake into Infected Erythrocytes
PLoS ONE | www.plosone.org5 May 2011 | Volume 6 | Issue 5 | e19334
infection with P. falciparum or B. divergens parasites (so-called new
permeability pathways, NPP ) are permissive for uptake of the
highly charged drugs Fos and FR. In striking contrast, those
parasites that do not reside in erythrocytes and do not show this
property, such as T. gondii or the liver stages of P. berghei, are
obviously secluded from the drug’s action since these compounds
seem unable to traverse the plasma membrane of the respective
The high hydrophilicity of the compounds tested presumably
prevents their passage through the host cell plasma membrane in
the absence of a suitable transporter or altered membrane
permeabilities. This situation resembles that in a number of
organisms like the cyanobacterium Synechocystis sp. PCC6803 and
in Mycobacteria where Fos is known to inhibit the respective
bacterial Dxr enzymes but shows no activity against the intact
organisms [27,28,29]. A recent study provided evidence that Fos is
unable to enter Mycobacteria, presumably because they lack the
gene for a glycerol-3-P transporter (GlpT) , which is also
absent in the genome of Synechocystis sp. PCC6803 (unpublished
observation). In E. coli this transporter has been shown to be
responsible for Fos uptake , and introduction of this gene into
Brucella abortus is both required and sufficient to make these
bacteria Fos-susceptible . Lack of uptake is presumably also
the reason why a distant relative to Apicomplexa, the dinoflagel-
late Perkinsus marinus, is insensitive to high concentrations of Fos
, despite the presence of the whole DOXP pathway in its
genome  (Figure S2).
From more than 40 compounds known to act on apicoplast
targets, Fos and FR are by far the most hydrophilic compounds at
physiological pH. Both drugs bear significant structural resem-
Table 1. In vitro effects of 72 h treatment with FR900098 on
growth of B. divergens.
FR900098 [m mM]
% parasitemia (cells
infected/100 cells±SD)% of untreated
1 36.264.4 83.6
100 1.860.2 4.2
Figure 6. Effect of FR900098 on P. berghei liver stage parasites (exoerythrocytic forms, EEF) in vitro and in vivo. Human hepatoma cells
infected with 10,000 GFP-expressing sporozoites and treated with 100 mM, 10 mM and 1 mM FR, respectively, were fixed 48 h p.i. Parasites were
visualized with a monoclonal antibody directed against HSP70 and Alexa488-labelled secondary antibody (*), and Hoechst 33342 to stain the nuclei.
A Immunofluorescence images of untreated and 100 mM-treated EEFs, respectively, showing no changes in morphological appearance upon FR
treatment. Bar represents 10 mm. B Quantitative assessment of drug effects on parasite growth at the three different FR concentrations. Given are
numbers of EEFs/well (mean value6SEM). C Quantitation of in vivo effect of FR treatment by real time PCR to assess parasite load in livers of infected
C57BL/6 mice either untreated, treated with FR or, as a positive control, treated with primaquine. Parasite load on the y-axis is given as the
percentage of the mean parasite load (6SEM) in PBS-treated control mice. D Effect of the batch of FR used in A-C on blood stage P. berghei in
infected C57BL/6 mice. Parasitemia was determined by means of Giemsa-stained thin blood smears at the indicated time points after infection.
Fosmidomycin Uptake into Infected Erythrocytes
PLoS ONE | www.plosone.org6 May 2011 | Volume 6 | Issue 5 | e19334
blance to DOXP (see Figure 1D) but the latter is not known to be
involved in any human metabolic pathway as a precursor or
product . Moreover, in the Human Metabolome Database
there is no indication for its presence in body fluids (http://www.
hmdb.ca/metabolites/HMDB12173). It is therefore likely that no
specific transporter for these compounds exist in the host cell
plasma membrane. Only erythrocytes that, upon infection with
either Plasmodium or Babesia, alter their membrane permeability in
turn become permissive for Fos or FR uptake. To our knowledge
NPP-like permeability changes have not been reported to date for
any nucleated cells infected by other Apicomplexa. Of note, P.
berghei sporozoites modulate the volume-regulated anion channel
activity of its target cells upon infection . However, the altered
activity is a mere response of the host cell to the increased cell
volume and not a prerequisite for parasite survival .
Although our data provide clear evidence that Fos or FR uptake
into infected erythrocytes requires functional NPP, at present we
do not know any molecular entity that is involved in this initial
entry process. The existence and importance of NPP in Plasmodium
(and by comparable mechanisms also in Babesia)-infected cells for
the acquisition of many essential nutrients is largely undisputed,
however, the exact way this is accomplished and the molecules
being involved is much less so (for recent discussions see [36,37]).
The recognition of NPP-specific anti-parasitic compounds in this
and other studies [38,39] may aid in the identification of the
molecular make-up of the NPP, for instance by exploiting
differential screens of Plasmodium mutants for Fos resistance.
Fos (in combination with clindamycin) is the first anti-malarial
drug that successfully completed phase II trials  for which
specific cell entry via the NPP pathway is shown. Our findings
reinforce the previously suggested concept of exploiting this
phenotype for targeting drugs into iRBC [40,41,42]. In some
parasite lines the development of resistance to antimalarials such
as blasticidin and leupeptin has been found to be associated with
alterations of a so-called plasmodial surface anion channel (PSAC)
[43,44], indicating the existence of a pathway that involves a
parasite-encoded transporter, which may undergo mutations
under selective pressure. However, since it is likely that more
than a single mechanism constitute what is collectively called NPP
, it is also possible that in some instances the actual transporter
is an activatable host cell protein. In such cases, the chances for
resistance development at the point of drug transport should be
minimal since there is no selective pressure on the respective host
genes. So far experimentally induced resistance against Fos has
only been reported to be caused by amplification of the Dxr gene
NPP are generally described as transporting a broad range of
substrates, with a preference for anions and electroneutral
compounds over cations, and with the rate of permeation
influenced by size and hydrophobicity of the solute in a number
of cases . However, simple correlations between size, charge or
hydrophobicity and permeation rates are not apparent, and a
combination of different properties could play a role in other cases
. Fos and FR are small but very hydrophilic at physiological
pH, whereas two other anti-plasmodial compounds (pentamidine;
T16) taken up via the same route [38,39] are larger and much
more lipophilic (Table S2). Therefore, predictions of inhibitors
that enter iRBC via NPP seem unreliable and instead rigorous
experimental testing is required in each case.
The concept that hydrophilic, charged molecules could be
potent drugs if their structures resemble natural metabolites for
which specific transport routes exist, has been recently emphasized
. Extending this idea to Plasmodium-infected cells we propose
that future screens aimed at finding such new anti-plasmodial
molecules should include more small hydrophilic metabolite-like
compounds specific to ‘parasite-only’ pathways (e.g. from the
apicoplast; ) in their chemical libraries (in contrast to the
current situation; see [47,48] for a general discussion). This
strategy would increase the chances of developing novel drug
candidates with a high, built-in safety profile, provided that they
can enter iRBC only via NPP. The fact that the target of Fos and
FR, Dxr, is absent from the human host adds an additional safety
level to its use. Interestingly, in a recent high-throughput screen of
nearly 2 million compounds against blood-stage P. falciparum,
among 13,500 active hits were also two purine analog phospho-
nates with physico-chemical properties comparable to Fos and FR
: CHEBI 390944 with an estimated IC50of 470 nM, and
CHEBI 641822 with an estimated IC50of 340 nM (; Figure
S8 and Table S2). It would be interesting to see whether the
presence of the purine moieties in these anti-plasmodial phospho-
nates make them substrates for specific transporters or if they can
enter iRBC only via NPP. The latter would further advocate an
appealing option , namely synthesizing chimeric drugs based
on active phosphonates like Fos/FR that are still taken up via the
NPP, and subsequently cleaved inside the cell to liberate two or
more inhibitory entities.
Our results provide a plausible explanation for the observed
differences in killing activity of Fos and FR against different
apicomplexan parasites. At present it is not known if any of the
other three membranes both compounds have to pass to reach
their target Dxr in the apicoplast are also impermeable for these
drugs in T. gondii, the liver stages of Plasmodium, or any of the other
apicomplexans that are not killed by Fos or FR. This aspect is of
general cell biological and biochemical interest and deserves
further studies. However, with regard to Fos’ and FR’s intended
application as drugs these potential barriers are of secondary
importance since the host cell plasma membrane is the first and
thus determining barrier that has to be overcome by any
pharmacologically active compound. Failure to do so eliminates
Fos and FR in their current formulation as potential drugs in those
Nevertheless, in pathogens that fail to take up Fos or FR the
DOXP pathway remains a prime drug target. Hope to improve
drug delivery to the pathogen comes from recent studies showing
that lipophilic Dxr inhibitors structurally unrelated to Fos and FR
can be developed . The hydrophilicity problem of phospho-
nates is well known and might be overcome by the development of
phosphonate esters , although cleavage of the ester linkage by
cellular esterases in the host cytosol might generate again a
charged Fos or FR molecule impermeable for the parasite plasma
membrane. Importantly, specific inhibitors for other enzymes of
the DOXP pathway in bacteria have been described recently
In conclusion, our results show that the parasite- and life cycle-
specific action of anti-infectives may be explained by differential
uptake into the infected host cell, which has important
implications for future screens aimed at finding safe, affordable
and potent molecular entities active against P. falciparum and other
Materials and Methods
All animal work was conducted in accordance with the current
German Protection of Animals Act (BGBl. I S. 1207), which
implements the directive 86/609/EEC from the European Union
and the European Convention for the protection of vertebrate
animals used for experimental and other scientific purposes. The
Fosmidomycin Uptake into Infected Erythrocytes
PLoS ONE | www.plosone.org7 May 2011 | Volume 6 | Issue 5 | e19334
protocol was approved by the ethics committee of the Max-
Planck-Institute for Infection Biology and the Berlin state
authorities (LAGeSo Reg# G0469/09).
All chemicals were from Sigma-Aldrich (except where noted).
Fosmidomycin and FR900098 were synthesized as previously
[1-14C]acetyl chloride as radioactive precursor. The final prepa-
ration was characterized by a specific activity of 0.26 mCi/mg
(58 mCi/mmol) and a radiochemical purity of 98.8%. [a-32P]CTP
and [1-14C]acetyl chloride was from GE Healthcare, [14C]- and
[3H]-labeled L-glutamic acid were obtained from Hartmann
Analytic GmbH, Braunschweig, Germany.
P. falciparum (isolate FCBR) was carried out by standard protocols
[57,58]. Human erythrocytes and plasma were purchased from
Blood Bank Universita ¨tsklinikum Gieben und Marburg GmbH,
Germany. Parasites were cultured in human erythrocytes (blood
group A+) at 37uC and a hematocrit of 2%. The culture medium
was RPMI1640 (PAA, Germany) supplemented with 10% human
plasma (A+; heat inactivated at 56uC for 30 min). Trophozoite-
infected erythrocytes were enriched to a parasitemia of .80–90%
by plasmagel floatation .
immortalized human foreskin fibroblasts (HFF; hTERT-BJ1;
Clontech) was performed as described previously .
Blood stage parasites of B. divergens (isolate
Rouen 1987;  were cultured in vitro in human A+erythrocytes
(Blood Bank Marburg, Germany) using RPMI1640 supplemented
with 10% human serum (A+; Blood Bank Marburg, Germany) and
0.2 mM hypoxanthine (c.c. pro, Germany) at 37uC and a
hematocrit of 5%. Flasks were flushed with a gas mixture of
90% N2, 5% O2and 5% CO2. The medium was changed daily
and cultures were diluted to 1% once they reached a parasitemia
The in vitro culturing and synchronization of
Production of anti-PfDxr antibodies
For the immunization of rabbits recombinant Dxr of P.
falciparum (PfDxr) was produced by an optimized version of a
previous protocol . A synthetic gene for PfDxr adapted to the
preferred codon usage of E. coli was inserted into the pQE31
expression vector (Quiagen, Hilden, Germany) providing an
amino-terminal His6 tag. E. coli XL1-blue pREP cells were
transformed with this construct and grown in Terrific Broth
medium at 37uC until an OD600of 1. After induction by the
addition of 1 mM IPTG the culture was continued for 15 h at
30uC reaching an OD600of 5 to 9. The cells were harvested by
centrifugation, resuspended in a 10-fold volume of IMAC buffer
(100 mM NaCl, 30 mM Tris-HCl, 2 mM 2-mercaptoethanol,
14% glycerol, pH 8.0) and disintegrated by ultrasonic treatment (3
times 5 min with 5 min pause, VS70T sonotrode, 30% pulse, 60%
amplitude). Insoluble material was removed by centrifugation
(75,000 g, 25 min, 4uC) and filtration through a 0.22 mm filter.
The soluble fraction was loaded on an immobilized cobalt column
(Talon Superflow, Clontech), which was eluted with a two-step
gradient of 50 and 150 mM imidazole in IMAC buffer. PfDxr was
obtained in the 150 mM imidazole fraction with a purity of ca.
90% as judged by SDS-PAGE. The specific activity determined in
a photometric Dxr enzyme assay was 2 U/mg . The protein
was concentrated by ultrafiltration and stored at 270uC. A typical
purification procedure starting with 8 flasks each containing
350 ml bacteria culture resulted in ca. 1 mg PfDxr. The relatively
low yield was due to the fact that most of the protein was produced
as inclusion bodies.
The purified enzyme was used for custom immunization
(Eurogentec, Seraing, Belgium) of 2 rabbits applying a protocol
including an initial injection followed by 3 booster shots. For each
injection 1.6 mg PfDxr were used. The sera were tested for
reactivity against PfDxr by Western blot. Only one of the two
rabbits developed a sufficiently high specific antibody titer (see
Immunofluorescence assays of T. gondii and P. falciparum
For indirect immunofluorescence assays, T. gondii-infected HFF
cells grown on 22 mm glass cover slips were processed exactly as
previously described ). Briefly, cells were fixed in 4%
paraformaldehyde, permeabilized in 0.25% Triton-X 100, and
blocked in 3% BSA fraction V (Fisher). P. falciparum-infected RBCs
were processed exactly as described by Tonkin et al. . All
primary and secondary antibody incubations were carried out at
room temperature for 1 h each in blocking solution, followed by
three 10 min washes in 0.1% Triton-X 100. Apicoplast and
nuclear DNA were stained with 2.8 mM 49,6-diamidino-2-
phenylindole (DAPI, Invitrogen) for 5 min (in PBS) right after
the secondary antibody step, and mounted on glass slides using
Fluoromount-G (Southern Biotechnology Associates, Inc). Sam-
ples were examined on a Leica DM IRBE, 100 W Hg-vapor lamp
and an Orca-ER digital camera (Hamamatsu). Images were
captured and analyzed using Openlab software (Improvision). For
co-localization analysis the ‘co-localization’ plugin of the ImageJ
program (Rasband, W.S., ImageJ, U. S. National Institutes of
Health, Bethesda, Maryland, USA, http://rsb.info.nih.gov/ij/,
1997–2009) was used.
Human erythrocytes/P. falciparum.
infected erythrocytes (26108cells/ml with a parasitemia of
,80%) at the trophozoite stage were incubated at 37uC in
RPMI1640 medium without human serum in the presence of
[14C]FR or [14C]Glu, respectively, of the same specific activity
At the indicated time points, aliquots of 100 ml corresponding to
107cells were removed and spun through a layer of 200 ml
dibutylphtalate. After freezing the sample in liquid nitrogen the
bottom of the tube containing the cell pellet was cut, cells were
lysed in scintillation cocktail and subsequently radioactivity of the
samples was determined in a Beckman Coulter MR4000
scintillation counter. Inhibitor studies using NPPB, furosemide
and chymotrypsin were done as described previously , with the
exception that [14C]FR or [14C]Glu were added and the cultures
were then processed as described above.
Human erythrocytes/B. divergens.
influx of [14C]FR into B. divergens infected erythrocytes 108
infected cells with a parasitemia of 15–20% were washed three
times in Dulbecco’s phosphate buffered saline (DPBS) and
subsequently incubated in DPBS for 30 minutes at 37uC, to
reach a point of zero membrane transport. Thereafter, cells were
incubated in 1 ml DPBS containing 1.7 mCi/ml at 37uC for
20 min. The incubation was carried out in the presence or absence
of 100 mM furosemide. To separate the cells from the radioactive
medium, aliquots of 100 ml were centrifuged through a cushion of
600 ml dibutylphtalate at 13,000 rpm for 2 min. The resulting
pellet was carefully harvested and lysed in 2 ml scintillation
cocktail (Roth) and analyzed in a scintillation counter (Beckman
To determine the
Fosmidomycin Uptake into Infected Erythrocytes
PLoS ONE | www.plosone.org8 May 2011 | Volume 6 | Issue 5 | e19334
Human fibroblasts/T. gondii.
were cultured in 6-well plates (Corning) to confluency (16106
cells) in DMEM incl. 10% fetal bovine serum and penicillin/
streptomycin. Where appropriate they were infected in duplicate
with T. gondii tachyzoites (RHb1) resulting in an infection rate
.50% of cells 24 h post infection. Depending on the assay the
cells were pre-treated with 1 mM unlabelled FR or Glu for 30 min
before they were incubated with either [14C]FR or [3H]Glu,
respectively, of the same specific activity (1.45 mCi/ml, equaling
25 mM drug concentration per assay) for 15 min at 37uC. To rule
out a possible competition between phosphate present in DMEM
and the phosphonate FR for uptake into cells, labeling was
performed in phosphate-free ‘‘extracellular buffer’’ which mimics
the extracellular milieu . Labelling was terminated by 3
washes of the monolayers with ice-cold phosphate buffer (50 mM
sodium phosphate pH 8.2) containing 1 mM unlabelled FR and
Glu each. At this stage microscopic examination of the plates
showed that all monolayers were still intact and evenly infected.
Cell lysis was performed by incubating the plates on a shaker with
500 ml lysis solution (0.1 M NaOH, 0.1% SDS) for 1 h at 4uC,
before an aliquot (200 ml) was counted in a PerkinElmer
MicroBeta Trilux scintillation counter. Dpm were determined
according to the manufacturer’s instructions.
HFF (hTERT-BJ1) cells
Coupled radiometric Dxr activity assay
The assay was basically performed as described previously .
For preparation of the cell lysate, 10 ml of packed purified T. gondii
tachyzoites (ca. 107cells) were suspended in 600 ml assay buffer
(100 mM Tris-HCl, 20 mM NaF, 10 mM MgCl2, 1 mM MnCl2,
pH 8.0) supplemented with a protease inhibitor cocktail (4 mM
Pefabloc SC, 2 mg/ml aprotinin, 2 mg/ml leupeptin, 2 mg/ml
pepstatin A, 2 mg/ml antipain), disintegrated by sonication for
2 min (Sonoplus HD 70, Bandelin, Berlin, Germany; 30% pulse,
70% amplitude) and centrifuged (22,000 g, 20 min, 1uC). For the
inhibition test, 45 ml aliquots of the lysate were combined with 5 ml
of a 20-fold concentrated Fos solution in water and pre-incubated
for 5 min on ice. Then, the activity test was started by combining
10 ml of these aliquots with 10 ml of reaction mixture consisting of
8 mMDOXP,8 mM NADPH,
(400 Ci/mmol) and 0.1 mg/ml E. coli YgbP (IspD) in assay buffer.
After incubation at 37uC for 5 min, 0.7 ml samples were spotted
onto 10620 cm silica gel 60 HPTLC plates (Merck), which were
developed longitudinally for 320 min with a mixture of n-
propanol/ethyl acetate/H2O (6:1:3, v/v). For autoradiography,
the plates were exposed for 3 h to a Kodak X-Omat AR film.
P. berghei blood stage growth inhibition assay in mice
10 C57BL/6 mice were intravenously infected with 30,000 P.
berghei ANKA infected red blood cells from a NMRI donor mouse.
At day three post-infection (p.i.), parasitemia was determined by
means of Giemsa-stained thin blood smears at different time
intervals. The baseline parasitemia before addition of the first dose
of 75 mg/kg FR900098 was 4.9%60.66% (mean6standard
deviation). 5 mice were treated with 75 mg/kg FR900098
dissolved in PBS, which was given intraperitoneally every 8–
12 hours for 5 days. Control mice received PBS only.
In vitro effects of FR900098 on exoerythrocytic forms of P.
Labtek slides with 30,000 HuH7 human hepatoma cells 
wereinfected with 10,000sporozoitesperwell(cl.507,constitutively
expressing GFP; ) and treated in triplicate with 100 mM, 10 mM
and 1 mM FR900098, respectively. 48 hours after infection cells
were fixed with cold methanol. Parasites were visualized with a
mouse monoclonal antibody to HSP70, followed by a goat
Alexa488-labelled antibody directed against mouse IgG, and
Hoechst 33342 to stain the nuclei. EEFs were counted per well.
In vitro effects of FR900098 on B. divergens
Erythrocytes infected with B. divergens were cultured in 24-well
plates in triplicates for 72 h in the presence or absence of FR at
different concentrations as described above. The parasitemia at
time point zero was 1% in each well. The culture medium was
changed daily, and parasitemia was monitored by counting
Giemsa-stained blood films. From each blood film three different
areas were counted. Parasitemia is expressed as the average of
infected cells (from triplicates) per 100 cells.
Assessment of the in vivo activity of FR900098 on P.
berghei liver stage development
Five C57BL/6 mice were infected with 10,000 sporozoites and
treated via i.p. injection with 250 mg/kg FR at 0, 12, 24 and
36 hours p.i. 5 control mice received PBS instead of FR but were
otherwise treated identical. As a control for successful treatment of
infection four infected mice received 60 mg/kg primaquine given
0 and 24 h p.i.
For determination of parasite load total [69,70] liver RNA was
isolated 42 h p.i. using the RNeasy kit (Qiagen), and cDNA
synthesized with the RETROScript kit (Ambion), according to the
manufacturer’s instructions. Real time PCR was performed using
the StepOne Plus real-time PCR system and Power SYBR Green
PCR Master Mix (Applied Biosystems), according to the manufac-
turer’s instructions, using gene-specific primers for the P. berghei
18SrRNA (for: 59 AAGCATTAAATAAAGCGAATACATCCT-
TAC 39; rev: 59 GGAGATTGGTTTTGACGTTTATGTG 39)
and the mouse GAPDH gene (for: 59 CGTCCCGTAGA-
CAAAATGGT 39; 59 TTGATGGCAACAATCTCCAC 39). Real
time PCR was performed in triplicates, with 1 cycle of 95uC for
15 min, followed by 40 cycles of 95uC for 15 s, 55uC for 15 s, and
60uC for 45 s. Relative copy numbers were determined with the
DDCt method .
pathway for the biosynthesis of the isoprenoid precur-
sors IPP/DMAPP. The pathways shown are based on MetaCyc
 and were drawn using the Pathway Tools software .
Numbers drawn in blue are enzyme EC numbers.
Comparison of the mevalonate and DOXP
select bacteria and plastid or apicoplast-containing
organisms. Sequences were taken from NCBI and aligned
using MUSCLE at http://www.phylogeny.fr/. Residues colored
in red, green and black in the E. coli sequence are also highly
conserved in all other Dxr proteins and have been implicated in
binding/interaction with the substrate DOXP and/or NADPH
(see ). Amino acids colored brown in the T. gondii sequence (aa
23 and 67) correspond to the aa following the predicted cleavage
site of either the signal sequence (determined with SignalP 3.0) or
the apicoplast targeting sequence (taken the first aa of the E. coli
sequence as reference point), respectively.
Sequence alignment of Dxr proteins from
falciparum. For details see Fig. 1B.
Localization of PfDxr in ring stages of P.
Fosmidomycin Uptake into Infected Erythrocytes
PLoS ONE | www.plosone.org9 May 2011 | Volume 6 | Issue 5 | e19334
stages of P. falciparum. For details see Fig. 1B.
Localization of PfDxr in early trophozoite
stages of P. falciparum. For details see Fig. 1B.
Localization of PfDxr in mid-late trophozoite
stages of P. falciparum. For details see Fig. 1B.
Localization of PfDxr in late trophozoite
P. falciparum right before lysis. For details see Fig. 1B.
Localization of PfDxr in merozoites stages of
phosphonates. Structures of Fos and FR (A) and their
superimposed 3D-structures (B). Aligned calculated three-dimen-
sional conformer coordinates for Fos and FR were retrieved from
PubChem (http://pubchem.ncbi.nlm.nih.gov) and visualized us-
ing Chimera . (C) Structures of the two anti-plasmodial
phosphonates described in . See http://www.ebi.ac.uk/
chemblntd/ and also the main text for details.
1D- and 3D-structures of anti-plasmodial
by two NPP inhibitors (NPPB and furosemide). Uptake of
[14C]FR into infected erythrocytes in the presence of different
concentrations (as indicated in the figure) of NPPB (5-Nitro-2-(3-
phenylpropylamino)benzoic acid) or furosemide, respectively, was
determined in triplicates (see Materials & Methods). The counts
were normalized to the uptake of [14C]FR into infected
erythrocytes in the absence of inhibitor (100%), and the respective
uptake determined for non-infected erythrocytes (0%).
Dose-dependend inhibition of [14C]FR uptake
genes in P. yoelii. The values are based on microarray data of
the rodent parasite P. yoelii  and were extracted from the P.
yoelii gene entries (accessible via the respective cross-references
from the P. falciparum gene entries; see Table S1) at PlasmoDB
(http://www.plasmodb.org/). M values denote the relative
expression level between pairs of conditions, expressed as base-2
logarithm (M=61 means a 2-fold difference in expression
between the compared samples). BS: mixed erythrocytic stages
when parasitemia was at 5–10%. LS24, 40, 50: Isolated liver
stage-infected hepatocytes 24, 40 or 50 hrs, respectively, after in
vivo infection. The data indicate that some of the genes for the
DOXP pathway are even stronger expressed in liver than in blood
stages (negative M values).
Expression levels of the DOXP pathway
antisera. The western blot of a T. gondii cell lysate with pre-
immune sera from two rabbits (lanes 1 & 2) and the respective
Western blot analysis of rabbit anti-PfDxr
hyper-immune sera after PfDxr immunization (lanes 3 & 4) is
shown. It clearly shows in lane 3 a very prominent band,50 kDa.
This size correlates very well with a predicted molecular weight of
48.8 kDa of the mature protein (i.e. without a cleaved bipartite
apicoplast targeting sequence; see also Fig. S2). The other rabbit
serum did not contain specific antibodies upon PfDxr immuniza-
biosynthesis in five Apicomplexa. Given are the EC
numbers, enzyme names and respective accession numbers in
either EuPathDB (http://eupathdb.org/eupathdb) for T. gondii, N.
caninum, P. falciparum and T. parva, or NCBI (http://www.ncbi.nlm.
nih.gov) for B. bovis. Designations printed in blue mean that
MassSpec data have been deposited in EuPathDB for this protein
(such data are available only for Plasmodium and T. gondii),
indicating that this protein is present in the respective parasite
Enzymes of the DOXP pathway of isoprenoid
pounds known to act on apicoplast targets in Plasmodi-
um and/or T. gondii. Compounds were compiled from the
literature [72,73]. L-glutamic acid and pantothenic acid as
physiological NPP substrates (indicated in yellow) and five other
anti-plasmodials (blue) are included for comparison (see main text
for details). Corresponding data were retrieved from PubChem
(http://pubchem.ncbi.nlm.nih.gov/). LogD were calculated using
the service ADME Boxes (http://www.pharma-algorithms.com/
webboxes/). Compounds up to hexachlorophene are sorted
according to their decreasing LogD. CID, compound ID at
PubChem; MW, molecular weight; XlogP3, calculated Log10of
the partition coefficient in octanol-water  (http://www.sioc-
ccbg.ac.cn/software/xlogp3); LogD, Log10 of the apparent
octanol-water partition coefficient D at various pH; TPSA, polar
surface area of substance . * known to enter iRBC via NPP
Physico-chemical properties of select com-
We are grateful to Dr. Dominique Soldati-Favre for parasite supply and for
helpful discussions, to Dr. Henri Vial for providing the B. divergens strain,
and to Dr. Geoffrey I. McFadden for donating the anti-ACP antibodies.
We thank Dr. Richard Lucius for providing lab space and for his
continuous support to FS, Dajana Henschker for technical assistance and
Dr. Anton Aebischer for critically reading the manuscript.
Conceived and designed the experiments: S. Baumeister JW HJ FS .
Performed the experiments: S.Baumeister JW AR MH S. Bietz OSH MK
JF FS . Analyzed the data: S. Baumeister S. Bietz JW DSR KM KL HJ FS .
Wrote the paper: S. Baumeister JW KM KL FS .
1. Jomaa H, Wiesner J, Sanderbrand S, Altincicek B, Weidemeyer C, et al. (1999)
Inhibitors of the nonmevalonate pathway of isoprenoid biosynthesis as
antimalarial drugs. Science 285: 1573–1576.
2. Cassera M, Gozzo F, D’alexandri F, Merino E, Del Portillo H, et al. (2004) The
methylerythritol phosphate pathway is functionally active in all intraerythrocytic
stages of Plasmodium falciparum. J Biol Chem 279: 51749–51759.
3. Odom AR, Van Voorhis WC (2010) Functional genetic analysis of the
Plasmodium falciparum deoxyxylulose 5-phosphate reductoisomerase gene. Mol
Biochem Parasitol 170: 108–111.
4. Borrmann S, Adegnika AA, Matsiegui PB, Issifou S, Schindler A, et al. (2004)
Fosmidomycin-clindamycin for Plasmodium falciparum infections in african
children. J Infect Dis 189: 901–908.
5. Borrmann S, Adegnika AA, Moussavou F, Oyakhirome S, Esser G, et al. (2005)
Short-course regimens of artesunate-fosmidomycin in treatment of uncompli-
cated Plasmodium falciparum malaria. Antimicrob Agents Chemother 49:
6. Borrmann S, Lundgren I, Oyakhirome S, Impouma B, Matsiegui PB, et al.
(2006) Fosmidomycin plus clindamycin for treatment of pediatric patients aged 1
to 14 years with Plasmodium falciparum malaria. Antimicrob Agents Chemother
7. Missinou MA, Borrmann S, Schindler A, Issifou S, Adegnika AA, et al. (2002)
Fosmidomycin for malaria. Lancet 360: 1941–1942.
8. Olliaro P, Wells T (2009) The global portfolio of new antimalarial medicines
under development. Clinl Pharmacol Ther 85: 584–595.
Fosmidomycin Uptake into Infected Erythrocytes
PLoS ONE | www.plosone.org 10 May 2011 | Volume 6 | Issue 5 | e19334
9. Kuemmerle HP, Murakawa T, Sakamoto H, Sato N, Konishi T, et al. (1985)
Fosmidomycin, a new phosphonic acid antibiotic. Part II: 1. Human
pharmacokinetics. 2. Preliminary early phase IIa clinical studies. Int J Clin
Pharmacol Ther Toxicol 23: 521–528.
10. Dondorp AM, Yeung S, White L, Nguon C, Day NP, et al. (2010) Artemisinin
resistance: current status and scenarios for containment. Nat Rev Microbiol 8:
11. Holstein SA, Hohl RJ (2004) Isoprenoids: remarkable diversity of form and
function. Lipids 39: 293–309.
12. Eisenreich W, Bacher A, Arigoni D, Rohdich F (2004) Biosynthesis of
isoprenoids via the non-mevalonate pathway. Cell Mol Life Sci 61: 1401–1426.
13. Seeber F, Soldati-Favre D (2010) Metabolic pathways in the apicoplast of
apicomplexa. Int Rev Cell Mol Biol 281: 161–228.
14. Clastre M, Goubard A, Prel A, Mincheva Z, Viaudmassuart M, et al. (2007) The
methylerythritol phosphate pathway for isoprenoid biosynthesis in coccidia:
Presence and sensitivity to fosmidomycin. Exp Parasitol 116: 375–384.
15. Lizundia R, Werling D, Langsley G, Ralph SA (2009) The Theileria apicoplast
as a target for chemotherapy. Antimicrob Agents Chemother 53: 1213–1217.
16. Ling Y, Sahota G, Odeh S, Chan JM, Araujo FG, et al. (2005) Bisphosphonate
inhibitors of Toxoplasma gondi growth: in vitro, QSAR, and in vivo investigations.
J Med Chem 48: 3130–3140.
17. Kirk K (2001) Membrane transport in the malaria-infected erythrocyte. Physiol
Rev 81: 495–537.
18. Cooper B, Chebib M, Shen J, King NJC, Darvey IG, et al. (1998) Structural
selectivity and molecular nature of L-glutamate transport in cultured human
fibroblasts. Arch Biochem Biophys 353: 356–364.
19. Fouts AE, Boothroyd JC (2007) Infection with Toxoplasma gondii bradyzoites has a
diminished impact on host transcript levels relative to tachyzoite infection. Infect
Immun 75: 634–642.
20. Baumeister S, Winterberg M, Duranton C, Huber S, Lang F, et al. (2006)
Evidence for the involvement of Plasmodium falciparum proteins in the formation of
new permeability pathways in the erythrocyte membrane. Mol Microbiol 60:
21. Zintl A, Mulcahy G, Skerrett H, Taylor S, Gray J (2003) Babesia divergens, a
bovine blood parasite of veterinary and zoonotic importance. Clin Microbiol
Rev 16: 622–636.
22. Alkhalil A, Hill D, Desai S (2007) Babesia and plasmodia increase host
erythrocyte permeability through distinct mechanisms. Cell Microbiol 9:
23. Sivakumar T, Aboulaila M, Khukhuu A, Iseki H, Alhassan A, et al. (2008) In
vitro inhibitory effect of fosmidomycin on the asexual growth of Babesia bovis and
Babesia bigemina. J Protozool Res 18: 71–78.
24. Tarun AS, Peng X, Dumpit RF, Ogata Y, Silva-Rivera H, et al. (2008) A
combined transcriptome and proteome survey of malaria parasite liver stages.
Proc Natl Acad Sci USA 105: 305–310.
25. Tsuchiya T, Ishibashi K, Terakawa M, Nishiyama M, Itoh N, et al. (1982)
Pharmacokinetics and metabolism of fosmidomycin, a new phosphonic acid, in
rats and dogs. Eur J Drug Metab Pharmacokinet 7: 59–64.
26. Murakawa T, Sakamoto H, Fukada S, Konishi T, Nishida M (1982)
Pharmacokinetics of fosmidomycin, a new phosphonic acid antibiotic.
Antimicrob Agents Chemother 21: 224–230.
27. Dhiman RK, Schaeffer ML, Bailey AM, Testa CA, Scherman H, et al. (2005) 1-
Deoxy-d-xylulose 5-phosphate reductoisomerase (IspC) from Mycobacterium
tuberculosis: towards understanding mycobacterial resistance to fosmidomycin.
J Bacteriol 187: 8395–8402.
28. Ershov Y, Gantt R, Cunningham FX, Gantt E (2002) Isoprenoid biosynthesis in
Synechocystis sp. strain PCC6803 is stimulated by compounds of the pentose
phosphate cycle but not by pyruvate or deoxyxylulose-5-phosphate. J Bacteriol
29. Woo Y-H, Fernandes RPM, Proteau PJ (2006) Evaluation of fosmidomycin
analogs as inhibitors of the Synechocystis sp. PCC6803 1-deoxy-D-xylulose 5-
phosphate reductoisomerase. Bioorg Med Chem 14: 2375–2385.
30. Brown A, Parish T (2008) Dxr is essential in Mycobacterium tuberculosis and
fosmidomycin resistance is due to a lack of uptake. BMC Microbiol 8: 78.
31. Sakamoto Y, Furukawa S, Ogihara H, Yamasaki M (2003) Fosmidomycin
resistance in adenylate cyclase deficient (cya) mutants of Escherichia coli. Biosci
Biotechnol Biochem 67: 2030–2033.
32. Sangari FJ, Perez-Gil J, Carretero-Paulet L, Garcia-Lobo JM, Rodriguez-
Concepcion M (2010) A new family of enzymes catalyzing the first committed
step of the methylerythritol 4-phosphate (MEP) pathway for isoprenoid
biosynthesis in bacteria. Proc Natl Acad Sci USA 107: 14081–14086.
33. Stelter K, El-Sayed NM, Seeber F (2007) The expression of a plant-type
ferredoxin redox system provides molecular evidence for a plastid in the early
dinoflagellate Perkinsus marinus. Protist 158: 119–130.
34. Grauvogel C, Reece KS, Brinkmann H, Petersen J (2007) Plastid isoprenoid
metabolism in the oyster parasite Perkinsus marinus connects dinoflagellates and
malaria pathogens-new impetus for studying alveolates. J Mol Evol 65: 725–729.
35. Prudencio M, Derbyshire ET, Marques CA, Krishna S, Mota MM, et al. (2009)
Plasmodium berghei-infection induces volume-regulated anion channel-like activity
in human hepatoma cells. Cell Microbiol 11: 1492–1501.
36. Baumeister S, Winterberg M, Przyborski J, Lingelbach K (2010) The malaria
parasite Plasmodium falciparum: cell biological peculiarities and nutritional
consequences. Protoplasma 240: 3–12.
37. Staines HM, Alkhalil A, Allen RJ, De Jonge HR, Derbyshire E, et al. (2007)
Electrophysiological studies of malaria parasite-infected erythrocytes: current
status. Int J Parasitol 37: 475–482.
38. Stead AMW, Bray PG, Edwards IG, Dekoning HP, Elford BC, et al. (2001)
Diamidine compounds: selective uptake and targeting in Plasmodium falciparum.
Mol Pharmacol 59: 1298–1306.
39. Biagini GA, Richier E, Bray PG, Calas M, Vial H, et al. (2003) Heme binding
contributes to antimalarial activity of bis-quaternary ammoniums. Antimicrob
Agents Chemother 47: 2584–2589.
40. Ginsburg H, Stein WD (1987) New permeability pathways induced by the
malarial parasite in the membrane of its host erythrocyte: potential routes for
targeting of drugs into infected cells. Biosci Rep 7: 455–463.
41. Staines HM, Ellory JC, Chibale K (2005) The new permeability pathways:
targets and selective routes for the development of new antimalarial agents.
Comb Chem High Through Screen 8: 81–88.
42. Gero AM, Dunn CG, Brown DM, Pulenthiran K, Gorovits EL, et al. (2003)
New malaria chemotherapy developed by utilization of a unique parasite
transport system. Curr Pharm Des 9: 867–877.
43. Hill D, Pillai AD, Nawaz F, Hayton K, Doan L, et al. (2007) A blasticidin S-
resistant Plasmodium falciparum mutant with a defective plasmodial surface anion
channel. Proc Natl Acad Sci USA 104: 1063–1068.
44. Lisk G, Pain M, Gluzman IY, Kambhampati S, Furuya T, et al. (2008) Changes
in the plasmodial surface anion channel reduce leupeptin uptake and can confer
drug resistance in Plasmodium falciparum-infected erythrocytes. Antimicrob Agents
Chemother 52: 2346–2354.
45. Dharia NV, Sidhu AB, Cassera MB, Westenberger SJ, Bopp SE, et al. (2009)
Use of high-density tiling microarrays to identify mutations globally and
elucidate mechanisms of drug resistance in Plasmodium falciparum. Genome Biol
46. Staines HM, Rae C, Kirk K (2000) Increased permeability of the malaria-
infected erythrocyte to organic cations. Biochim Biophys Acta 1463: 88–98.
47. Dobson PD, Patel Y, Kell DB (2009) ‘Metabolite-likeness’ as a criterion in the
design and selection of pharmaceutical drug libraries. Drug Discov Today 14:
48. Dobson PD, Kell DB (2008) Carrier-mediated cellular uptake of pharmaceutical
drugs: an exception or the rule? Nat Rev Drug Discov 7: 205–220.
49. Gamo FJ, Sanz LM, Vidal J, de Cozar C, Alvarez E, et al. (2010) Thousands of
chemical starting points for antimalarial lead identification. Nature 465:
50. Muregi FW, Ishih A (2010) Next-generation antimalarial drugs: hybrid
molecules as a new strategy in drug design. Drug Develop Res 71: 20–32.
51. Deng L, Sundriyal S, Rubio V, Shi Z-z, Song Y (2009) Coordination chemistry
based approach to lipophilic inhibitors of 1-deoxy-D-xylulose-5-phosphate
reductoisomerase. J Med Chem 52: 6539–6542.
52. Wiesner J, Ortmann R, Jomaa H, Schlitzer M (2007) Double ester prodrugs of
FR900098 display enhanced in-vitro antimalarial activity. Arch Pharm Chem
Life Sci 340: 667–669.
53. Wang W, Li J, Wang K, Huang C, Zhang Y, et al. (2010) Organometallic
mechanism of action and inhibition of the 4Fe-4S isoprenoid biosynthesis
protein GcpE (IspG). Proc Natl Acad Sci USA 107: 11189–11193.
54. Wang W, Wang K, Liu YL, No JH, Li J, et al. (2010) Bioorganometallic
mechanism of action, and inhibition, of IspH. Proc Natl Acad Sci USA 107:
55. Kamiya T, Hashimoto M, Hemmi K, Takeno H (1980) Hydroxyaminohy-
drocarbon-phosphonic acids. US patent 4,206 156.
56. O¨hler E, Kanzler S (1995) Regioselective palladium(0) catalyzed amination of
carbonates of allylic hydroxyphosphonates with hydroxylamine derivatives: A
convenient route to phosphonic acids related to the antibiotic fosmidomycin.
Synthesis 1995: 539–543.
57. Trager W, Jensen JB (1976) Human malaria parasites in continuous culture.
Science 193: 673–675.
58. Lambros C, Vanderberg JP (1979) Synchronization of Plasmodium falciparum
erythrocytic stages in culture. J Parasitol 65: 418–420.
59. Pasvol G, Wilson RJ, Smalley ME, Brown J (1978) Separation of viable schizont-
infected red cells of Plasmodium falciparum from human blood. Ann Trop Med
Parasitol 72: 87–88.
60. Seeber F, Boothroyd JC (1996) Escherichia coli beta-galactosidase as an in vitro
and in vivo reporter enzyme and stable transfection marker in the intracellular
protozoan parasite Toxoplasma gondii. Gene 169: 39–45.
61. Gorenflot A, Brasseur P, Precigout E, L’Hostis M, Marchand A, et al. (1991)
Cytological and immunological responses to Babesia divergens in different hosts:
ox, gerbil, man. Parasitol Res 77: 3–12.
62. Giessmann D, Heidler P, Haemers T, Van Calenbergh S, Reichenberg A, et al.
(2008) Towards new antimalarial drugs: synthesis of non-hydrolyzable
phosphate mimics as feed for a predictive QSAR study on 1-deoxy-D-
xylulose-5-phosphate reductoisomerase inhibitors. Chem Biodivers 5: 643–656.
63. Harb OS, Chatterjee B, Fraunholz MJ, Crawford MJ, Nishi M, et al. (2004)
Multiple functionally redundant signals mediate targeting to the apicoplast in the
apicomplexan parasite Toxoplasma gondii. Eukaryot Cell 3: 663–674.
64. Ralph SA, Van Dooren GG, Waller RF, Crawford MJ, Fraunholz MJ, et al.
(2004) Metabolic maps and functions of the Plasmodium falciparum apicoplast. Nat
Rev Micro 2: 203–216.
Fosmidomycin Uptake into Infected Erythrocytes
PLoS ONE | www.plosone.org11May 2011 | Volume 6 | Issue 5 | e19334
65. Moudy R, Manning TJ, Beckers CJ (2001) The loss of cytoplasmic potassium
upon host cell breakdown triggers egress of Toxoplasma gondii. J Biol Chem 276:
66. Wiesner J, Hintz M, Altincicek B, Sanderbrand S, Weidemeyer C, et al. (2000)
Plasmodium falciparum: Detection of the deoxyxylulose 5-phosphate reductoisome-
rase activity. Exp Parasitol 96: 182–186.
67. Silvie O, Goetz K, Matuschewski K (2008) A sporozoite asparagine-rich protein
controls initiation of Plasmodium liver stage development. PLoS Pathog 4:
68. Janse CJ, Franke-Fayard B, Mair GR, Ramesar J, Thiel C, et al. (2006) High
efficiency transfection of Plasmodium berghei facilitates novel selection procedures.
Mol Biochem Parasitol 145: 60–70.
69. Friesen J, Silvie O, Putrianti ED, Hafalla JC, Matuschewski K, et al. (2010)
Natural immunization against malaria: causal prophylaxis with antibiotics. Sci
Transl Med 2: 40ra49.
70. Bruna-Romero O, Hafalla JC, Gonzalez-Aseguinolaza G, Sano G, Tsuji M,
et al. (2001) Detection of malaria liver-stages in mice infected through the bite of
a single Anopheles mosquito using a highly sensitive real-time PCR. Int J Parasitol
71. Caspi R, Altman T, Dale JM, Dreher K, Fulcher CA, et al. (2010) The MetaCyc
database of metabolic pathways and enzymes and the BioCyc collection of
pathway/genome databases. Nucleic Acids Res 38: D473–479.
72. Karp PD, Paley S, Romero P (2002) The Pathway Tools software.
Bioinformatics 18 Suppl 1: S225–232.
73. Henriksson LM, Bjorkelid C, Mowbray SL, Unge T (2006) The 1.9 A resolution
structure of Mycobacterium tuberculosis 1-deoxy-d-xylulose 5-phosphate reductoi-
somerase, a potential drug target. Acta Crystallogr D Biol Crystallogr D62:
74. Pettersen EF, Goddard TD, Huang CC, Couch GS, Greenblatt DM, et al.
(2004) UCSF Chimera-a visualization system for exploratory research and
analysis. J Comput Chem 25: 1605–1612.
Fosmidomycin Uptake into Infected Erythrocytes
PLoS ONE | www.plosone.org12 May 2011 | Volume 6 | Issue 5 | e19334
EC number Enzyme name (Abbr.)
5-phosphate synthase (Dxs)
reductoisomerase (IspC, Dxr)
cytidylyltransferase (IspD, YgbP)
erythritol kinase (IspE, YchB)
diphosphate synthase (IspF, YgbB)
diphosphate synthase (IspG, GcpE)
diphosphate reductase (IspH, LytB)
T. gondii N. caninum P. falciparum T. parva B. bovis
126.96.36.199 TGME49_008820 NCLIV_003330 PF13_0207 TP01_0516 BBOV_III002600A
188.8.131.527 TGME49_014850 NCLIV_051900 PF14_0641 TP02_0073 BBOV_III010740A
184.108.40.206 TGME49_106260 NCLIV_044490 PFA0340w TP03_0057 BBOV_I003560A
220.127.116.11 TGME49_106550 NCLIV_044740 PFE0150c TP02_0681 BBOV_II007070A
18.104.22.168 TGME49_055690 NCLIV_029230 PFB0420w TP03_0365 BBOV_IV002810A
22.214.171.124 TGME49_062430 NCLIV_025400 PF10_0221 TP02_0667 BBOV_II006930A
126.96.36.199 TGME49_027420 NCLIV_045740 PFA0225w TP03_0674 BBOV_III001660A
Table S2 Download full-text
LogD at pH
CID Name MW XLogP3 TPSA
1.7 6.5 7.4 8
4421 Nalidixic acid
611 L-Glutamic acid
6613 Pantothenic acid