Novel S-adenosyl-L-methionine decarboxylase inhibitors as potent antiproliferative agents against intraerythrocytic Plasmodium falciparum parasites.

S-adenosyl-l-methionine decarboxylase (AdoMetDC) in the polyamine biosynthesis pathway has been identified as a suitable drug target in Plasmodium falciparum parasites, which causes the most lethal form of malaria. Derivatives of an irreversible inhibitor of this enzyme, 5'-{[(Z)-4-amino-2-butenyl]methylamino}-5'-deoxyadenosine (MDL73811), have been developed with improved pharmacokinetic profiles and activity against related parasites, Trypanosoma brucei. Here, these derivatives were assayed for inhibition of AdoMetDC from P. falciparum parasites and the methylated derivative, 8-methyl-5'-{[(Z)-4-aminobut-2-enyl]methylamino}-5'-deoxyadenosine (Genz-644131) was shown to be the most active. The in vitro efficacy of Genz-644131 was markedly increased by nanoencapsulation in immunoliposomes, which specifically targeted intraerythrocytic P. falciparum parasites.

Introduction

Polyamines (putrescine, spermidine and spermine) are critical components of cell growth and division, particularly in rapidly proliferating cells that include cancerous cells and numerous parasites (Casero and Marton, 2007; Heby et al., 2007). A number of enzymes in the biosynthesis of polyamines have been validated as suitable drug targets including ornithine decarboxylase (ODC) (Pegg, 2006) and S-adenosyl-l-methionine decarboxylase (AdoMetDC) (Pegg, 2009) as the two major enzymatic activities. Protozoan infections resulting in human parasitic diseases such as African sleeping sickness (caused by subspecies of Trypanosoma brucei), Chagas disease (Trypanosoma cruzi), leishmaniasis (Leishmania spp) and malaria (Plasmodium spp) are highly reliant on substantial amounts of polyamines for development and proliferation (Heby et al., 2003; Birkholtz et al., 2011). Of these diseases, malaria has a high disease incidence in most tropical regions of the world, with Plasmodium falciparum parasites being the most lethal.

AdoMetDC catalyses a chokepoint reaction to produce decarboxylated S-adenosyl-l-methionine (dcAdoMet) that is exclusively used for polyamine biosynthesis. The irreversible AdoMetDC inhibitor, 5′-{[(Z)-4-amino-2-butenyl]methylamino}-5′-deoxyadenosine (MDL73811), is 100-fold more effective than the ODC inhibitor, dl-α-difluoromethylornithine (DFMO), in curing murine T. b. brucei and T. b. rhodesiense infections. Treatment of T. brucei parasites with MDL73811 causes an intracellular hypermethylated state due to accumulation of S-adenosyl-l-methionine (Byers et al., 1991) in addition to the detrimental depletion of trypanothione (the principal polyamine-dependent thiol in trypanosomes) (Yarlett et al., 1991; Willert and Phillips, 2008). AdoMetDC is considered an attractive drug target in P. falciparum parasites due to its unique association with ODC in a heterotetrameric bifunctional protein complex, PfAdoMetDC/ODC (van Brummelen et al., 2008; Birkholtz et al., 2011; Williams et al., 2011). MDL73811 inhibits in vitro intraerythrocytic P. falciparum parasite proliferation (Wright et al., 1991; Das Gupta et al., 2005), however, this does not result in a hypermethylated state but only depletes intracellular polyamine levels, leading to cellular cytostasis (Wallace et al., 2003; van Brummelen et al., 2008; Birkholtz et al., 2011).

MDL73811, however, is not clinically useful against parasitic infectious due to poor blood brain barrier penetration, a short plasma half-life, poor oral bioavailability and limited metabolic activity (Wright et al., 1991; Das Gupta et al., 2005; Barker et al., 2009). Structure-guided design of MDL73811 derivatives, modified on the ribose and purine moieties through addition of halogens and methyl groups, resulted in a series of compounds with improved ADME toxicity profiles. These included improved aqueous solubility, decreased rates of hepatocyte and microsome clearance, minimal CYP inhibition and half the plasma protein binding capacity compared to MDL73811 (Bacchi et al., 2009; Barker et al., 2009; Hirth et al., 2009). Methylation of position 8 of the adenine group resulting in 8-methyl-5′-{[(Z)-4-aminobut-2-enyl]methylamino}-5′-deoxyadenosine (Genz-644131) displayed an increased inhibitory potency against heterologous TbAdoMetDC. This compound showed improved cellular toxicity against different T. brucei parasite strains, with a longer plasma half-life and improved blood brain barrier penetration in in vivo mice (Bacchi et al., 2009; Barker et al., 2009). Here, the MDL73811 derivatives were assessed for inhibitory activity against heterologous PfAdoMetDC as well as for inhibition of intraerythrocytic P. falciparum parasite proliferation in vitro.

Methods and materials

MDL73811 and derivatives

MDL73811 (5′-{[(Z)-4-amino-2-butenyl]methylamino}-5′-deoxyadenosine), Genz-644131 (8-methyl-5′-{[(Z)-4-amino-2-butenyl]methylamino}-5′-deoxyadenosine), Genz-644043 (2′-fluoro-5′-{[(Z)-4-amino-2-butenyl]methylamino}-5′-deoxyadenosine), and Genz-644053 (2-chloro-5′-{[(Z)-4-amino-2-butenyl]methylamino}-5′-deoxyadenosine) (Table 1) were synthesised by the Genzyme Corporation (www.genzyme.com, 2013; Hirth et al., 2009).

Table 1

Conformational search analysis of MDL73811 and its derivatives.

		Lowest overall energy conformation#	Lowest syn energy conformation#	Number of conformations generated#	Bioactive syn conformational number⁎, #
MDL73811 5′-{[(Z)-4-amino-2-butenyl]methylamino}-5′-deoxyadenosine		−131.6	−124.0	200	22
Genz-6441318-methyl-5′-{[(Z)-4-amino-2-butenyl]methylamino}-5′-deoxyadenosine		−132.9	−130.7	207	7
Genz-6440432′-fluoro-5′-{[(Z)-4-amino-2-butenyl]methylamino}-5′-deoxyadenosine		−131.0	−117.8	193	106
Genz-6440532-chloro-5′-{[(Z)-4-amino-2-butenyl]methylamino}-5′-deoxyadenosine		−136.0	−127.8	209	64

All energies are given in kcal/mol.

Generated conformations were ranked from the lowest to highest energy i.e. the bioactive syn conformation of Genz-644131 was ranked the 7th lowest energy conformation from 207 conformations generated.

Recombinant PfAdoMetDC enzyme inhibition assays

The monofunctional form of PfAdoMetDC (harmonised gene construct) as well as bifunctional PfAdoMetDC/ODC were heterologously expressed in Escherichia coli BL21 Star™ (DE3) cells and purified via Strep-tag affinity chromatography as previously described (Birkholtz et al., 2004; Williams et al., 2011). To determine the enzyme inhibition activity of the MDL73811 derivatives, an isotope-based bioassay that measures the release of radiolabeled CO2 (Birkholtz et al., 2004; Williams et al., 2011) was performed with 5 μg of either monofunctional PfAdoMetDC or bifunctional PfAdoMetDC/ODC in the presence of 1 μM of each derivative. Results were normalised to the specific activity (nmol/min/mg) of an uninhibited control to determine % inhibition.

Determination of the inhibition constant of Genz-644131 against PfAdoMetDC

The apparent inhibition constant (Kapp) of Genz-644131 was determined with Kitz and Wilson time dependent enzyme kinetics for irreversible inhibitors as described (Kitz and Wilson, 1962). Using the isotope-based bioassay described above, residual enzyme activity was measured following pre-incubation (37 °C) at fixed time intervals (2, 4 and 6 min) with varying inhibitor concentrations (0.02, 0.05 and 0.1 μM) and 1 μg of either monofunctional PfAdoMetDC or bifunctional PfAdoMetDC/ODC, each in duplicate. The reciprocal of the slope in the primary graph (kapp) was plotted against the reciprocal of the inhibitor concentrations to yield the secondary plot, from which the kinact and the Kapp values were derived.

Homology modelling and conformational analysis

P. falciparum and T. brucei homology models were generated using the human AdoMetDC crystal structure (PDBid 3DZ2) as template similar to a previously described model (Wells et al., 2006). Molecular shape-based alignment between MDL73811 derivatives and the homology model was performed with vROCS (v3.1.0 OpenEyeScientific Software, Inc., Santa Fe, NM, USA, www.eyesopen.com, 2010) (Supplemental data; Fig. S1 and Table S1). Conformational analysis was performed using Conformation Search and the Minimisation module of Discovery Studio 3.0 suite (Accelrys, Inc.). Detailed methods for homology modelling, molecular shape based alignment and conformational analysis are provided in Supplemental data, S1.

In vitro cultivation of intraerythrocytic P. falciparum parasites and IC50 determination of MDL73811 derivatives

Intraerythrocytic P. falciparum parasites (strain 3D7; chloroquine sensitive) were maintained in P. falciparum culture media as described (Verlinden et al., 2011). Intraerythrocytic parasites were synchronised to a 95% ring stage population with a 5% sorbitol solution. The effect of MDL73811 derivatives on the proliferation of intraerythrocytic P. falciparum parasites at 37 °C for 96 h was determined using a SYBR Green I-based fluorescence assay as described (Verlinden et al., 2011). MDL73811 and Genz-644131 were dissolved in 1xPBS and Genz-644043 and Genz-644053 in DMSO and incubated with ring stage intraerythrocytic P. falciparum parasites (1% parasitaemia, 1% haematocrit) at specific concentrations, serially diluted 2-fold in culture medium (final 0.1% (v/v)) non-lethal DMSO concentration (Grobusch et al., 1998). Sigmoidal dose–response curves were fitted to the data using SigmaPlot 11.0 with non-linear regression yielding the IC50 values (concentration at which 50% inhibition of parasite proliferation was observed).

Determining parasite recovery following Genz-644131 inhibition

The ability of the products of polyamine metabolism to rescue parasites from the inhibitory effect of Genz-644131 was determined. Ring stage intraerythrocytic P. falciparum parasites (1% parasitaemia, 1% haematocrit) were treated with Genz-644131 (2xIC50) in the presence of exogenous spermidine trichloride (non-toxic concentration, 250 μM, results not shown) and 500 μM of the polyamine oxidase inhibitor, aminoguanidine (Niemand et al., 2013), for 96 h at 37 °C and parasite proliferation determined with a SYBR Green I-based assay (Stjernborg and Persson, 1993; Lee and Sayre, 1998).

Subsequently, to determine the ability of ring stage intraerythrocytic P. falciparum parasites (1% parasitaemia, 1% haematocrit) to recover after Genz-644131 treatment, the latter was withdrawn following 24 h incubation at 37 °C at 2 × IC50, the parasites washed and subsequently resuspended in normal culture media. Samples were removed at 12 h intervals over 96 h and DNA content was determined for the treated, untreated and the Genz-644131 withdrawn parasite populations using a SYBR Green I-based assay.

Spermidine uptake in intraerythrocytic P. falciparum parasites

Intraerythrocytic P. falciparum parasites were purified to ⩾ 95% parasitaemia with magnetic separation (Trang et al., 2004; Teng et al., 2009). [3H]spermidine uptake was initiated by combining equal volumes of cell suspension and radiolabelled solution (1 μCi/ml [3H]spermidine at 5 nm extracellular concentration, PerkinElmer, in 125 mM NaCl, 5 mM KCl, 20 mM glucose, 25 mM HEPES and 1 mM MgCl2, pH 7.1). At predetermined time intervals following incubation at 37 °C, the reactions were terminated by dibutyl phthalate sedimentation. A 10 μl sample of the aqueous phase was transferred to a scintillation vial to determine the extracellular concentration of [3H]spermidine. The remaining cell pellet was lysed with 0.1% (v/v) Triton X-100, the proteins precipitated with 5% w/v trichloroacetic acid and cell debris (including membrane fractions) removed with centrifugation before measuring the radioactivity present in the aqueous supernatant. The rapid initial association of radiolabel with the cells, due to polyamines trapped in the extracellular space as well as adhering to the cell surface was determined and subtracted from the total measured radioactivity (cmp’s) to determine the intracellular concentration of polyamines. The data are given as a distribution ratio of intracellular to extracellular spermidine (Saliba et al., 1998).

Comparative IC50 determination of Genz-644131 incorporated with Pheriod® technology

A micellular formulation of a nanoparticle structure, Pheroid®, consisting mainly of 43.8% plant and essential ethylated and PEGylated polyunsaturated fatty acids, was prepared in incomplete RPMI 1640 medium as described (Grobler et al., 2008; Steyn et al., 2011). The suspension was subsequently filtered (0.22 μm), diluted 50 × with sterile water and homogenised with Genz-644131 powder to yield an encapsulated suspension with a final compound concentration of 23.75 mM. The encapsulation efficiency of Pheroid® was analysed with confocal laser scanning microscopy (Grobler et al., 2007). Subsequently, ring stage intraerythrocytic P. falciparum parasites (1% parasitaemia, 1% haematocrit) were treated with Genz-644131 encapsulated Pheroid® at a 20-fold dilution (24 μM initial starting concentration) and incubated for 96 h at 37 °C. Parasite proliferation of cell suspensions treated with Genz-644131 encapsulated into Pheroid® was determined, and normalised against non Genz-644131 encapsulating Pheroid® in order to derive the IC50 value using a SYBR Green I-based assay.

Comparative IC50 determination of Genz-644131 with immunoliposomes

Immunoliposomes were prepared by the lipid film hydration method (MacDonald et al., 1991) and covalently functionalised with IgG antibody fragments prepared specifically for intraerythrocytic P. falciparum parasites (Urbán et al., 2011a). For encapsulation of Genz-644131, 1.74 mg of the compound was incorporated into a 1.5 ml immunoliposome suspension to give a final concentration of encapsulated Genz-6644131 at 30 μM (10% encapsulation efficiency). Ring stage intraerythrocytic P. falciparum parasites (1% parasitaemia, 1% haematocrit) were treated with Genz-644131 encapsulated immunoliposomes at a 50-fold dilution (0.6 μM starting concentration) and incubated for 96 h at 37 °C. Parasite proliferation of cell suspensions treated with Genz-644131 encapsulated into immunoliposomes was determined as described above.

Statistical analyses

All data are representative of at least three independent biological experiments (n ⩾ 3), each performed in triplicate. Statistical analysis was performed using either paired or unpaired Student’s t-tests. Data were analysed using GraphPad Prism 5.0 or SigmaPlot 11.0.

Results

Inhibitory effect of MDL73811 derivatives on heterologous monofunctional PfAdoMetDC and bifunctional PfAdoMetDC/ODC

MDL73811 and Genz-644131 showed the highest percentage inhibition of monofunctional PfAdoMetDC at 98% and 100%, respectively (n = 3, P < 0.01, paired Student’s t-test) (Fig. 1). The inhibition of the PfAdoMetDC domain of heterologous bifunctional PfAdoMetDC/ODC was similar at 96% and 98%, respectively (n = 3, P < 0.01, paired Student’s t-test) (Fig. 1). By contrast, Genz-644043 only inhibited monofunctional PfAdoMetDC by 57% (n = 3, P = 0.14, paired Student’s t-test) whereas Genz-644053 had no inhibitory effect (Fig. 1).

Fig. 1

The inhibitory activities of MDL73811 derivatives against monofunctional PfAdoMetDC (grey) and the PfAdoMetDC domain of bifunctional PfAdoMetDC/ODC (black). MDL73811 and the three derivatives (1 μM) were incubated with either 5 μg monofunctional PfAdoMetDC or bifunctional PfAdoMetDC/ODC for 30 min at 37 °C. Specific activity (nmol/min/mg) of the monofunctional and bifunctional PfAdoMetDC domains were normalised against the uninhibited enzymes. Data are representative of three independent experiments performed in triplicate, ±SEM. ∗∗P < 0.01, paired Student’s t-test. Where not shown, the error bars fall within the symbols.

The activity of PfAdoMetDC decreased in a concentration dependent manner following pre-incubation with Genz-644131 for both monofunctional and bifunctional forms of the protein (Fig. 2A and C). A non-significant increase in the Kapp value of 0.36 ± 0.10 μM (n = 3, P > 0.05, unpaired Student’s t-test) was determined for Genz-644131 on monofunctional PfAdoMetDC (Fig. 2B) compared to 0.22 ± 0.09 μM for MDL73811 (results not shown). However, there was a significant difference in the inhibition of bifunctional PfAdoMetDC/ODC by Genz-644131 (Kapp at 0.18 ± 0.02 μM (Fig. 2D) compared to MDL73811 (0.53 ± 0.09 μM, results not shown) (n = 3, P = 0.02, unpaired Student’s t-test).

Fig. 2

Enzyme kinetics of Genz-644131 against monofunctional and bifunctional PfAdoMetDC. Kitz-Wilson inhibition kinetics was used to determine the Kapp for Genz-644131 against monofunctional (A and B) and bifunctional PfAdoMetDC (C and D). Percentage activity was determined from residual enzyme activity, following pre-incubation with Genz-644131 at 0.02 (circles), 0.05 (squares) or 0.1 μM (triangles) concentrations ([I]) at specific time intervals (0–6 min) (Et). The ln(E/E0) of the activity at a specific inhibitor concentration was plotted against the pre-incubation time points using non-linear regression. The reciprocal of the slopes (1/kapp) of the primary plots (A and C) was plotted against the reciprocal of the specific inhibitor concentrations using non-linear regression (B and D), from which the kinact (inverse of the y-intercept) and the Kapp (slope multiplied by kinact) were derived (Kitz and Wilson, 1962). Data are representative of three independent experiments performed in triplicate, ±SEM and all values fell into the 95% confidence interval of the mean. Where not shown, the error bars fall within the symbols.

The efficiency of inactivation (depicted by the kinact/Kapp ratio) of MDL73811 compared to Genz-644131 against monofunctional PfAdoMetDC was not significantly different at 1.91 and 1.17 μM−1 min−1, respectively. However, the inactivation efficiency for Genz-644131 against the PfAdoMetDC domain of bifunctional PfAdoMetDC/ODC is 1.6-fold higher (kinact/Kapp = 3.50 μM−1 min−1) compared to MDL73811 (kinact/Kapp = 2.19 μM−1 min−1). Moreover, there was a ∼3-fold increase in the inactivation efficiency of Genz-644131 against bifunctional PfAdoMetDC/ODC compared to monofunctional PfAdoMetDC (kinact/Kapp ratios of 3.50 vs. 1.17 μM−1 min−1), respectively. Genz-644131 seems to therefore be a more effective inhibitor of PfAdoMetDC compared to MDL73811, with marked preference for the protein when found in its native conformation in bifunctional PfAdoMetDC/ODC.

Due to the observed differences in the inhibition of MDL73811 and its derivatives against PfAdoMetDC, the binding capacity of these compounds to PfAdoMetDC was analysed in silico. Previously, it was shown that purine nucleoside AdoMetDC inhibitors adopt unusual syn conformations and that modification of these inhibitors does not only affect protein ligand interactions but also alter conformational preferences (Tolbert et al., 2001) (Supplemental data S1). MDL73811 has an energy difference of 7.6 kcal/mol between the lowest energy conformation and its bioactive syn conformation (Table 1). However, the 8-methyl substitution on the purine rings of MDL73811, yielding Genz-644131, increases the conformational preference for the bioactive syn conformation of the latter (energy difference equal to 2.2 kcal/mol, Table 1). By contrast, the energy differences for both the halogen substituted Genz-644043 and Genz-644053 derivatives were higher, with the latter also showing a distorted N-(Z)-4-aminobutenyl-N-methyl tail conformation due to interference of the chlorine substitution at position 2 of the purine ring (Supplemental data S1).

The selected binding pose of Genz-644131 (Fig. 3A) shows similar interactions to other AdoMetDC substrate analogues (Fig. 3B) (McCloskey et al., 2009). Moreover, conformational search analysis of the N-(Z)-4-aminobutenyl-N-methyl tail of Genz-644131 showed that the lowest absolute energy of the Genz-644131 syn conformation is reached when the tail assumes the predicted binding conformation (-126.9 kcal/mol) (Fig. 3C). Notably, the 8-methyl substitution on the purine ring does not show any steric hindrance within the protein active site and therefore would not negatively affect ligand binding.

Fig. 3

A predicted binding pose for Genz-644131 to PfAdoMetDC highlighting conserved residues with T. brucei and human protein equivalents. (A) A homology model of PfAdoMetDC bound with Genz-644131 in the active site. The ribbon representing the β-chain is coloured either bright blue or red indicating conserved and non-conserved residues. Likewise, the α-chain ribbon is coloured in a lighter shade. (B) The interacting residues between PfAdoMetDC and Genz-644131. Green lines represent hydrogen bonds formed between the protein and ligand. (C) Representation of the systematic conformational search of the [(Z)-4-amino-2-butenyl] methylamino tail with the lowest energy conformations in yellow. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Genz-644131 is active against in vitro intraerythrocytic P. falciparum parasites

The IC50 of the MDL73811 derivatives was determined on intraerythrocytic P. falciparum parasites in vitro (96 h incubation at 37 °C). Treatment of intraerythrocytic P. falciparum parasites with Genz-644131 resulted in a significant, 2-fold decrease in the IC50 compared to MDL73811 (IC50 = 0.97 ± 0.06 μM vs. 2.21 ± 0.07 μM , n = 5, P < 0.01, unpaired Student’s t-test). The IC50 values of Genz-644043 and Genz-644053 against intraerythrocytic P. falciparum parasites were significantly higher (25.6 ± 8.4 and 22.4 ± 7.5 μM; n = 4, P > 0.05, unpaired Student’s t-test) than that of the parent compound, MDL73811.

The ability of exogenous polyamines to rescue the inhibitory effect of Genz-644131 on intraerythrocytic P. falciparum parasites was established by determining the IC50 of Genz-644131 in the presence and absence of exogenous spermidine (250 μM). No significant change in the IC50 value of Genz-644131 in the presence (IC50 = 0.94 ± 0.03 μM) or absence (IC50 = 0.97 ± 0.06 μM) of spermidine (n = 3, P = 0.89, paired Student’s t-test) was observed (Fig. 4A), indicating that spermidine could not antagonise the inhibitory effect of Genz-644131. However, P. falciparum infected erythrocytes are capable of taking up exogenous spermidine, with [3H]spermidine reaching a distribution ratio of 1.4 ± 0.4 (n = 6) following 60 min incubation (Fig. 4B).

Fig. 4

Uptake of [3H]spermidine, with rescue and reversibility of Genz-644131 inhibited intraerythrocytic P. falciparum parasites in vitro. (A) Initial ring stage intraerythrocytic P. falciparum parasites were treated with Genz-644131 (serial dilution) alone (squares) or in the presence of 250 μM spermidine (circles, 0.5 μM aminoguanidine present) for 96 h at 37 °C. Parasite proliferation is expressed as a percentage of untreated parasite proliferation at 100%. Data are representative of n ⩾ 3 independent experiments performed in triplicate, ± SEM. (B) Time course for the uptake of [3H]spermidine (5 nM extracellular concentration) into intraerythrocytic P. falciparum parasites (circles) at 37 °C over 60 min averaged from five independent experiments and shown ± SEM. A distribution ratio of 1.4 ± 0.4 was obtained, where a ratio of 1 indicates that the radiolabelled polyamine has equilibrated to levels equal to the extracellular levels. (C) Initial ring stage intraerythrocytic P. falciparum parasites were either treated with Genz-644131 (2 × IC50, squares) for 96 h at 37 °C or treated with Genz-644131 (2 × IC50, triangles) for 24 h at 37 °C before replacing the culture media thereby removing the Genz-644131 before incubating the parasites for a further 96 h at 37 °C. Untreated initial ring stage intraerythrocytic P. falciparum parasites (circles) incubated at 37 °C for 96 h was included as a positive control for parasite proliferation. Samples were taken every 12 h and DNA content was measured as relative fluorescence units using SYBR Green I-based assay. Data are representative of n ⩾ 3 independent experiments performed in triplicate, ±SEM. Where not shown, the error bars fall within the symbol. (D) Morphological monitoring of the stage specificity of parasites treated with Genz-644131 (2 × IC50), analysing percentage distribution in each life-cycle stage. Treated parasites indicated that Genz-644131 arrested parasite development during the trophozoite stage compared to untreated parasites.

The recovery of intraerythrocytic P. falciparum parasites (1% parasitaemia, 1% haematocrit) after limited exposure to Genz-644131 (2× IC50) for 24 h was determined after washing out the compound, followed by additional incubation of parasite cultures for a further 96 h before measuring DNA content as an indicator of parasite proliferation. There was a significant increase in DNA levels observed for untreated parasites over the two life-cycles analysed (Fig. 4C). Genz-644131 treated ring-stage intraerythrocytic P. falciparum parasites (2× IC50) were able to recover after 24 h of drug pressure and continue to proliferate following drug removal (Fig. 4C). However, continuous exposure of intraerythrocytic P. falciparum parasites to Genz-644131 (2× IC50) for 96 h resulted in a stage-specific inhibition of parasite proliferation, with parasites arrested in the trophozoite stage (∼24 h post-invasion) within the first life-cycle (Fig. 4D).

Effect of Genz-644131 encapsulated in nanovectors on in vitro antiplasmodial activity

To improve membrane translocation of Genz-644131 and possibly the in vitro activity against intraerythrocytic P. falciparum parasites, Genz-644131 was encapsulated into two types of nanovectors, a submicron micellular emulsion formulation, Pheroid® (Grobler et al., 2008), and a parasite targeting immunoliposome system (Urbán et al., 2011a). The compound encapsulated into Pheroid® did not show a significant decrease in the in vitro IC50 (0.97 ± 0.06 μM compared to 0.67 ± 0.29 μM; n = 4, P > 0.05, unpaired Student’s t-test; Fig. 5A). By contrast, Genz-644131 encapsulated immunoliposomes showed a significant 32-fold decrease in the in vitro IC50 compared to non-encapsulated Genz-644131 (0.97 ± 0.06 μM vs. 0.031 ± 0.004 μM; n = 3, P < 0.01, unpaired Student’s t-test) (Fig. 5B).

Fig. 5

The effect of encapsulation of Genz-644131 in different nanovectors on its in vitro anti-plasmodial activity. Parasite proliferation of ring stage intraerythrocytic P. falciparum parasites was monitored with a SYBR Green I-based assay over 96 h at 37 °C and IC50 determined from dilution series. Dose–response curves for Genz-644131 alone (squares) compared to incorporated into (A) Pheroid® or (B) immunoliposomes (circles). Data are representative in each instance of three independent experiments performed in triplicate or quadruplicate, ±SEM. Where not shown, the error bars fall within the symbols.

Discussion

Polyamine biosynthesis enzymes have been the target of various parasitic disease intervention strategies (Birkholtz et al., 2011) as highlighted by the clinical treatment of T. brucei infections through DFMO inhibition of ODC activity (Van Nieuwenhove et al., 1985). Of the other enzymatic activities associated with polyamine biosynthesis, inhibition of AdoMetDC shows promise as a therapeutic target in P. falciparum, with MDL73811 a 1000-fold more potent against intraerythrocytic P. falciparum parasites compared to DFMO (Wright et al., 1991). Although MDL73811 is an irreversible inhibitor of AdoMetDC activity, it has poor drug-like characteristics for Plasmodium (Wright et al., 1991) and Trypanosoma parasites (Barker et al., 2009), which led to the synthesis of pharmacokinetically amenable derivatives. These derivatives of MDL73811 were used here to determine (1) their efficacy in inhibiting the PfAdoMetDC protein and (2) their antiproliferative activity against intraerythrocytic P. falciparum parasites in vitro.

Several comparisons can be drawn between the treatment of P. falciparum and T. brucei parasites with the lead derivative, Genz-644131. Firstly, the AdoMetDC protein from both these parasites responds similarly to Genz-644131 treatment. PfAdoMetDC has a near conserved active site compared to AdoMetDC homologues from human and T. brucei parasites, despite an overall low sequence identity (21% and 23%, respectively (Wells et al., 2006)). As a result, MDL73811 inhibits AdoMetDC from both P. falciparum and T. brucei parasites at comparable levels and in a similar manner as indicated by their respective micromolar Kapp values (Bitonti et al., 1990; Das Gupta et al., 2005; Williams et al., 2011). However, Genz-644131 potently inhibits monofunctional and bifunctional PfAdoMetDC similarly to TbAdoMetDC (Barker et al., 2009) with Kapp values in the nanomolar range (Barker et al., 2009).

The 1.6-fold decrease in Kapp between MDL73811 and Genz-644131 observed for the bifunctional PfAdoMetDC/ODC is explained by the 8-methyl substitution on the purine ring of Genz-644131, which promotes the preferred bioactive syn conformation (Pegg, 2009). However, Genz-644131 is ∼7-fold less effective in inhibiting monofunctional PfAdoMetDC compared to the T. brucei enzyme (kinact/Kapp ratios of 1.17 μM−1 min−1 for PfAdoMetDC compared to 7.78 μM−1 min−1 for TbAdoMetDC (Barker et al., 2009)). The association of PfAdoMetDC with ODC in the biologically relevant bifunctional protein PfAdoMetDC/ODC has been shown to result in the modulation of plasmodial AdoMetDC activity (Birkholtz et al., 2011; Williams et al., 2011). Rate-limiting and equimolar synthesis of putrescine and dcAdoMet by the ODC and AdoMetDC activities is enabled by a decrease in AdoMetDC activity when associated in the bifunctional complex with ODC in comparison to its monofunctional PfAdoMetDC form, respectively (Williams et al., 2011). Here, although comparative inactivation efficiencies are seen for Genz-644131 for the monofunctional and bifunctional proteins, this inhibitor shows a ∼3-fold increase in specificity and rate of inhibition of the AdoMetDC domain of the bifunctional protein. This can be attributed to the lower substrate Km of PfAdoMetDC in the bifunctional protein compared to the monofunctional protein, which probably reflects differences between active site conformations of these two proteins and consequently, their binding affinities for Genz-644131 (Williams et al., 2011). Interestingly, the simultaneous inhibition of both activities of the bifunctional PfAdoMetDC/ODC with Genz-644131 and DFMO is additive as was also shown for MDL73811 and DFMO on in vitro P. falciparum parasites (Wright et al., 1991; van Brummelen et al., 2008) (Supplemental data S2).

In contrast to the marked improvement (>10-fold) in the in vitro antiproliferative efficacy of T. brucei parasites treated with Genz-644131 compared to MDL73811 (Barker et al., 2009), Genz-644131 only shows marginal (2-fold) improvement in the in vitro IC50 against intraerythrocytic P. falciparum parasites. The antiproliferative effect observed with Genz-644131 was not plasmodicidal to the parasite, similar to treatment with MDL73811 and DFMO, with parasite proliferation recovering after limited Genz-644131 exposure (24 h at 2× IC50). Both MDL73811 and DFMO treatment result in a cytostatic effect since inhibition is negated by the uptake of exogenous polyamines (Assaraf et al., 1987; Das Gupta et al., 2005). Co-treatment of parasites with MDL73811 and exogenous spermidine did not abolish the inhibitory effect of MDL73811 on parasite proliferation, and it was previously suggested that intraerythrocytic P. falciparum parasites are incapable of spermidine uptake, since exogenously supplied putrescine, but not spermidine, was capable of overcoming biosynthesis inhibition caused by a variety of inhibitors (Assaraf et al., 1987; Das Gupta et al., 2005). Likewise, co-treatment of parasites with Genz-644131 and exogenous spermidine also did not abolish the inhibitory effect of Genz-644131 on parasite proliferation. However, recent work clearly indicates that exogenous spermidine is taken up by isolated P. falciparum trophozoite-stage parasites (Niemand et al., 2012). Once inside the infected erythrocyte unit, the parasite is able to efficiently take up spermidine across the plasma membrane in a concentration dependent manner, mediated by an electrogenic process energised by the parasite’s membrane potential (Niemand et al., 2012). In addition, here we report that exogenous spermidine is taken up by P. falciparum infected erythrocytes. Therefore, the inability of spermidine to abolish Genz-644131 inhibition does not appear to be due to the inability of the parasite to take up spermidine. Genz-644131 shows improved in vivo cellular toxicity against different T. brucei parasite strains (Bacchi et al., 2009; Barker et al., 2009). When this compound was tested in a murine malaria model for in vivo antimalarial activity, Genz-644131 significantly (P < 0.001) reduced P. berghei parasitaemia by 89% when dosed in the Peters model for 4 days at 100 mg/kg/day. Animals dosed with 20 mg/kg/day showed a 37% (P = 0.002) reduction. However, in no case was there sterile cure, as all animals had detectable parasitaemia levels on day 4 (Supplemental data S3). This may be due to the cytostatic effect described above.

The evidence provided does however not exclude the possibility of off target effects of Genz-644131 on P. falciparum parasites including its binding to purine deaminases and polyamine oxidases as observed for MDL73811 (Hirth et al., 2009), particularly to P. falciparum adenosine deaminases (Reyes et al., 1982) and erythrocytic polyamine oxidases (Byers et al., 1992). However, Genz-644131 (at 2× IC50) arrested parasite development in a stage-specific manner during the trophozoite stages (18–26 h post invasion), as previously described for MDL73811 (Wright et al., 1991). This corresponds to the requirement of polyamines due to the stage-specific expression of PfAdoMetDC/ODC (18–30 h post-invasion) during the trophozoite stage of the asexual cycle (van Brummelen et al., 2008). The parasite arrested temporal phenotype induced by Genz-644131 therefore corresponds to the expression profile of PfAdoMetDC in the parasite as the target for this compound.

Another explanation for the relatively poor activity of Genz-644131 against intraerythrocytic P. falciparum parasites may be due to the poor membrane permeability of the compound itself, as low membrane permeability was previously shown to be the only inferior in vitro ADME characteristic of the MDL73811 derivatives (Barker et al., 2009). Additionally, for any compound to access the intraerythrocytic P. falciparum parasites, they would need to cross the parasite plasma membrane (PPM), parasitophorous vacuolar membrane (PVM) and erythrocyte membrane (Lingelbach and Joiner, 1998). Extracellular T. brucei parasites are surrounded by only a single plasma membrane (Vennerstrom et al., 2004; Orhan et al., 2006) and cannot synthesise purines de novo, and therefore have to acquire host purines (Hassan and Coombs, 1988). Adenosine, of which MDL73811 is a structural analogue, is actively transported into T. brucei parasites by the T. brucei nucleotide transporter 1 (TbNT1) (de Koning et al., 2005) and the T. brucei aminopurine transporter (TbAT1 or P2). The latter transporter has been confirmed to actively transport MDL73811 (Goldberg et al., 1997; Maser et al., 1999), which could explain the low nanomolar in vitro IC50 values of MDL73811 in these parasites. As in T. brucei, Plasmodium parasites also do not synthesise purines de novo and has to recruit exogenous purines from the host (Downie et al., 2008). In contrast to T. brucei, multiple transport mechanisms enable the uptake of purines into intraerythrocytic P. falciparum parasites including host purine transporters (Quashie et al., 2008) as well as parasite derived transporters, PfNT1 and PfNT4 localised in the PPM (Carter et al., 2000; Parker et al., 2000). Although the latter are low-affinity transporters (Downie et al., 2008), their ability to transport MDL73811 and derivatives needs further investigation.

The development of novel lipid based nanovectors provides a solution for the challenges facing malaria chemotherapeutics, since it has the potential to mediate sustained, targeted drug release thereby increasing the drug plasma half-life, lowering dosage requirements and reducing drug toxicity (Maurer et al., 2011). Additionally, lipid-based nanovectors has been shown to eliminate off-target effects by delivery to specific targeted cells, thereby improving the therapeutic efficacy of the compound (Forrest and Kwon, 2008). This makes lipid based nanovector drug delivery systems ideal for treatment of intracellular pathogens (Alving, 1986; Armstead and Bingyun, 2011; Maurer et al., 2011). Previous encapsulations of chloroquine (Urbán et al., 2011a) and fosmidomycin (Urbán et al., 2011b) resulted in a 10- and 7.5-fold decrease in the in vitro IC50 values of these compounds.

To investigate possible enhancement of the uptake of Genz-644131 into intraerythrocytic P. falciparum parasites, the compound was incorporated into two nanovector drug delivery systems: Pheroid® and immunoliposomes (Steyn et al., 2011; Urbán et al., 2011a). The Pheroid® system is a nanovector carrier developed from a submicron micellular emulsion formulation, typically ranging in size from 80 to 300 nm (Steyn et al., 2011). These micellular structures can be manipulated in terms of structure, size and morphology to enhance the solubility properties of intended compounds, by entrapment and delivery of compounds across cellular membranes (Steyn et al., 2011). Liposomes are synthetic lipid bilayers of up to 200 nm that have the ability to increase drug bioavailability by encapsulating compounds into the hydrophilic core of the lipid bilayer system. Moreover, the liposomal preparations were orientated with half anti-glycophorin A antibodies, specific for intraerythrocytic P. falciparum parasites, enhancing the selectivity of Genz-644131 to the parasite (Urbán et al., 2011a). Here, the 32-fold decrease observed in the in vitro IC50 of immunoliposome encapsulated Genz-644131 against intraerythrocytic P. falciparum parasites suggests that the uptake of Genz-644131 by itself into intracellular P. falciparum parasites is restricted. However, the activity of compounds can be improved by either enhancing the chemical pharmacokinetic properties through medicinal chemistry, or encapsulating the compound into drug delivery systems. Although the Genz-644131 immunoliposome combination has not been tested in vivo, other immunoliposomal drug suspensions tested against murine mice infections improved the pharmacokinetic profiles of the drugs tested (Owais et al., 1995; Agrawal and Gupta, 2000). Encapsulation of Genz-644131 with immunoliposomes may also reduce non-specific off-target effects and in an sustainable release of the drug to prolong its plasma half-life.

The combination of Genz-644131 with a novel nanovector drug delivery system therefore provides the most promising result obtained thus far with this nanovector delivery system against intraerythrocytic P. falciparum parasites in vitro, and could be evaluated in novel antimalarial drug development.