International Journal of Pharmaceutics 483 (2015) 57–62

Preferential targeting of human erythrocytes infected with the malariaparasite Plasmodium falciparum via hexose transporter surface proteins

Kajal Devi Heikham, Ankit Gupta, Ambrish Kumar, Chandan Singh, Juhi Saxena,Kumkum Srivastava, Sunil K. Puri, Anil K. Dwivedi, Saman Habib, Amit Misra *CSIR - Central Drug Research Institute, Lucknow 226031, India

A R T I C L E I N F O

Article history:Received 27 August 2014Received in revised form 4 February 2015Accepted 5 February 2015Available online 7 February 2015

Chemical compounds studied in this article:Cornstarch (PubChem CID: 24836924)Quinine (PubChem CID 3034034)Fluorescein isothiocyantate (PubChem CID18730)40 ,6-diamidino-2-phenylindole (PubChemCID 2954)CDRI Compound 97/63

Kewords:StarchNanoparticlesGLUT1PfHT-1MalariaFlow cytometry quinineTrioxane

A B S T R A C T

Glucose uptake by Plasmodium-infected erythrocytes (RBC) is higher compared to uninfected RBC.Glucose is transported across the cell membrane by transporter proteins. Particles of median size146.3 � 18.7 nm, containing anti-malarial agents in corn starch were prepared for investigating: (a)whether the glucose moiety in starch targets RBC via hexose transporter(s), (b) whether there aredifferences in the extent of targeting to uninfected RBC versus infected RBC (iRBC) in view of higher cellsurface density of these proteins on iRBC and (c) whether targeting provides enhanced efficacy against P.falciparum in comparison to drugs in solution. Binding of these particles to RBC was target-specific, sinceit could be blocked by phloretin, an inhibitor of glucose transporters (GLUT), or competed out in a dose-dependent manner with D-glucose in a flow cytometry assay. Significant (P = 0.048, t-test) differences inextent of targeting to iRBC versus RBC were observed in flow cytometry. CDRI 97/63 incorporated inparticles was 63% more efficacious than its solution at 250 ng/ml, while quinine was 20% more efficaciousat 6.25 ng/ml in a SYBR Green incorporation assay. Preferential targeting of iRBC using an inexpensiveexcipient promises advantages in terms of dose reduction and toxicity alleviation.

ã 2015 Elsevier B.V. All rights reserved.

Contents lists available at ScienceDirect

International Journal of Pharmaceutics

journal homepage: www.elsev ier .com/locate / i jpharm

1. Introduction

Clinical management of severe and cerebral malaria, includingintravenous administration of potent anti-malarial agents, resultsin varying degrees of success (Olupot-Olupot and Maitland, 2013;Sheehy and Angus, 2011). There is therefore a need for strategiesthat can minimise drug dosage without inducing drug resistance(Barnes et al., 2008). Targeting anti-malarial agents preferentially,if not specifically, to RBC infected with the malaria parasite (iRBC)can potentially reduce drug exposure to non-target sites, anddeliver high local concentrations of the drug to the infected cell(Movellan et al., 2014).

* Corresponding author at: CSIR - Central Drug Research Institute, Sector 10A,Janakipuram Extension, Mohiuddinpur, Lucknow 226031, India. Tel.: +91 522 2772460; fax: +91 522 277 1940.

E-mail address: amit_misra@cdri.res.in (A. Misra).

http://dx.doi.org/10.1016/j.ijpharm.2015.02.0110378-5173/ã 2015 Elsevier B.V. All rights reserved.

The advantages of incorporating anti-malarial agents in nano-metre-sized drug delivery systems composed of pharmaceutically-acceptable materials have been amply demonstrated (Gupta et al.,2007; Haas et al., 2009; Jain et al., 2014; Movellan et al., 2014;Nayak et al., 2010).

It was hypothesized that particles bearing a hexose moietycapable of binding the human erythrocyte glucose transporterGLUT1, as well as the parasite hexose transporter PfHT1, would bepreferentially targeted to iRBC on account of a higher density ofeither or both these proteins on the iRBC surface (Slavic et al. 2011).GLUT1 density on RBC can vary as a result of recruitment of“masked” protein molecules (Zhang and Ismail-Beigi, 1998),suggesting a mechanism whereby this molecule may be upregu-lated on the iRBC surface. Synthetic bio-mimetic polymers with areducing glucose moiety have been visualized to interact with thisprotein (Park et al., 2003). Particles prepared using chitosan, whichhas N-acetyl glucosamine residues available for putative interac-tion with GLUT1/PfHT1 showed advantages in parasite killing invivo (Haas et al., 2009). Similarly, heparin (Valle-Delgado et al.,

58 K.D. Heikham et al. / International Journal of Pharmaceutics 483 (2015) 57–62

2013) and gellan (Recuenco et al., 2014) have been shown to targetnanoparticles to infected erythrocytes (iRBC) and interfere withRBC invasion and parasite growth. The present report describeselucidation of the possible mechanism whereby glucose-bearingparticles may target iRBC.

2. Material and methods

2.1. Preparation of particles

Particles were prepared using pharmaceutical grade corn starch(BDH, India), quinine (free base) and CDRI compound 97/63 whichis a water-soluble trioxane developed by CSIR-CDRI, India (Singhand Puri, 2001). As detailed in Fig. 1, particles containing drugswith or without fluorescein isothiocyanate (FITC, Sigma) as afluorescent label were prepared by adding solutions of starch andthe drug to a non-solvent under sonication (Van Soest et al., 2004).The nuclear stain DAPI and SYBR Green I were sourced fromMolecular Probes, and all other reagents from Sigma, unlessotherwise specified.

2.2. Pharmaceutical properties of particles

Particle size distribution was evaluated using a Zetasizer ZS(Malvern, UK). HPLC methods were developed and partiallyvalidated to estimate drug content and release profiles fromparticles. Briefly, a Shimadzu HPLC system was used to isocraticallyelute CDRI 97/63 from an analytical C18 Column (Lichrospher,Merck) at the flow rate of 1 ml/min using ammonium acetatebuffer (10 mM, pH-4) and acetonitrile in the ratio of 30:70 as themobile phase. Elution was monitored at 240 nm. Samples forcontent analysis were prepared by centrifugation of the colloidalsuspension at 12,000 rpm for 15 min and collecting the particlepellet and supernatant separately. The supernatant was lyophi-lized before extraction while the pellet was directly extracted withdimethyl sulphoxide (DMSO). DMSO extracts were dilutedappropriately in the mobile phase before injecting on to thecolumn. Quinine was estimated by the method of Croes et al.(1994) with minor modifications.

Drug release was studied by placing 5 ml of the formulation(�500 mg drug) in a dialysis bag of molecular weight cut-off100 KDa (Sigma) and securing the bag to the paddle of a USP

(1): Span85 (10 μl) in Aceto(25 ml), und er probe sonica tion at 20 W

200 μdropw

(2): CDRI 97 /63 (10mg/ml in ethy l acetate)+ 50μl of 0.5mg/ml FITC

200 μl, dropw ise

(4) Sonication , 5

(5) Rotary Evapo

Fig. 1. Preparation of starch particles by addition of drug and starch solutions to a nonfollowed by evaporation of solvents in a rotary evaporator.

dissolution rate apparatus containing 900 ml phosphate bufferedsaline (Electrolab, India). The paddle was run at 50 rpm and thebath temperature was maintained at 37 � 1 �C. One millilitresamples were withdrawn at indicated times, lyophilised andreconstituted for analysis as above.

2.3. Flow cytometry

Flow cytometry was carried out on a FACS Aria (BD Biosciences).With sanction and supervision of the Institutional EthicsCommittee of CSIR-CDRI, 900 ml aliquots of blood (donated byhealthy volunteers providing informed consent to Kohli BloodBank, Lucknow) were incubated with 100 ml of a solution of 97/63 + FITC in 10% DMSO-saline, or a suspension of FITC-taggedparticles in the same medium, ensuring equal concentrations ofthe compound and FITC in all incubation tubes. Two hours later, theinstrument was set up to discriminate between RBC, leukocytes,and platelets based on forward and side scatter.

Specificity of targeting was investigated in parasite cultures inhuman RBC, maintained as described below. To inhibit theinteraction of GLUT-1 with the particles, a solution of 5 mMphloretin (Park et al., 2003) prepared by taking 0.0137 g ofphloretin in 10 ml PBS and two logarithmic dilutions were addedto 107 RBC infected to 5.33% parasitemia with P. falciparum strain3D7. Similarly, four samples each of parasite culture were exposedto dilutions of 5% D-glucose, ensuring isotonic conditions byadjusting NaCl content in the buffer. Tubes containing untreatedculture and culture exposed to formulation in the absence of D-glucose were incubated in parallel for negative and positivecontrol.

2.4. Efficacy against P. falciparum in culture

The chloroquine-sensitive P. falciparum strain 3D7 was culturedin human RBC using Trager and Jensen’s protocol (Trager andJensen, 1978), as described previously (Ram, 2007). Briefly, RBCwere suspended in RPMI-1640 with 25 mM HEPES, 1% D-glucose,0.2% NaHCO3 and 0.5% Albumax II (Invitrogen). Culture wasinitiated at about 1% parasitemia in 1.5% haematocrit in 96-wellplates. About 1 h later, 50 ml of a DMSO-PBS solution containingindicated concentrations of quinine free base or CDRI 97/78 insolution or in starch particles was added to triplicate wells. After

ne (3):Starch, IP (100 mg), Distilled water, (10 0 ml)

• Autoc laved 15 min, 15 ps ig)• Coo led to room tempe rature• Tangen tial-flow filtration (0.2 μm)• Adde d Pluronu ic F68 to final concentration of 0.2%

l, ise

min

rator

(6) Res idue recon stituted in 5 ml phosp hate-bu ffered sa line: particle suspe nsion

-solvent (acetone) under sonication and in the presence of a surfactant (Span 85),

Fig. 2. Characterisation of particles. (a) Number-average particle size (bars, N = 6; �SD) of freshly-prepared particles in suspension in phosphate-buffered saline (PBS). (b)Polydispersity index of the same batches. (c) Representative result of size determination of particles suspended in PBS. (d) Scanning electron micrograph of air-dried particlesfollowing gold sputter coating. (e) Change in size during storage at 4–8 �C over a 3-week period. (f) Release (N = 2) of 97/63 from particles (open circles, open squares), versusrapid equilibration of the compound (open triangles, filled triangles) with release medium.

K.D. Heikham et al. / International Journal of Pharmaceutics 483 (2015) 57–62 59

48 h of incubation, cultures were lysed with 100 mL of a buffercontaining saponin and Triton-X100 in Tris-EDTA along with SYBRGreen. Relative fluorescence units (RFU) corresponding to dyeincorporation in parasite DNA were recorded using a spectrofluo-rimeter (FLUOstar, BMG Labtechnologies). Thin smears were alsoprepared at 48 h post treatment and stained with Geimsa’s stain. Atleast 300 cells were counted in 3 fields per slide. The number ofRBC positive for schizonts was recorded. Untreated wells devel-oped about 5% parasitemia during this period.

3. Results and discussion

3.1. Particle characterisitics

Particles prepared by the process shown in Fig. 1 had a mediansize of 150 nm (146.3 � 18.7 nm, N = 6), a polydispersity index of

Fig. 3. (a) The regions of interest (ROI) containing RBC and WBC as differentiated by forwand debris. (b) FITC-fluorescence histograms of RBC in human blood sample exposed to

tagged particles (curve 3) respectively. (c) FITC fluorescence associated with WBC in th

0.11 �0.05 and spherical morphology (Fig 2a–c). The preparationcomprised particles within a Gaussian size distribution rangingfrom 30 to 300 nm. No significant change in size occurred duringthree weeks of storage at 4–8 �C. Particle size distribution at thetime of preparation (Fig 2d), did not change significantly atsubsequent times (Fig 2e). Change in size during storage waswithin an upper control limit of 158.2 nm within a confidenceinterval of 95%, which was not statistically significantly differentfrom the original size in a paired t-test. The zeta potential of thepreparation was �69.2 � 2.2 mV.

The preparation contained about 100 (107.79 � 6.40) mg/ml ofthe candidate drug as assayed by HPLC, of which about 36% wasincorporated in the particles while the remainder was present insolution.

As shown in Fig 2f, drug release from the particles wascompleted within 24 h and followed first order kinetics

ard and side scatter. Scatter points outside the ROI indicate thrombocytes (platelets)fluorescence-free particles (curve 1), particles + FITC in solution (curve 2) and FITC-e same samples.

Table 1Fluorescence intensities resulting from exposure to FITC in solution versus particles (n = 2).

Stain MFI % Histogram area > RBC auto-fluorescence % Increase in MFI

Non-fluorescent particles 135 � 4 0.3 –

Non-fluorescent particles + FITC 156 � 8 1.2 15.56FITC-particles 189 � 6 8.9 40

60 K.D. Heikham et al. / International Journal of Pharmaceutics 483 (2015) 57–62

(R2 = 0.9982). In contrast, drug placed in the dialysis bag along withan equivalent amount of corn starch equilibrated with the releasemedium within 2 h, indicating that particle formation resulted incontrolled drug release.

3.2. Targeting to RBC and iRBC

Fig. 3 shows delineation of various cell populations in humanblood under the flow cytometry conditions employed, andfluorescence associated with RBC versus white blood cells (WBC)when exposed to FITC in a representative experiment.

Human RBC themselves showed little auto-fluorescence.Staining RBC with equal amounts of FITC, either in solution orin the form of FITC-tagged particles indicated greater association ofthe fluorescent label when FITC was bound to particles. The

Fig. 4. (a) Auto-fluorescence of cultured RBC (curve 1) was enhanced upon staining wiInhibition of fluorescence imparted to RBC by FITC-particles (curve 1) by IC50 of phloretimparted to RBC (curve 1) by different concentrations of D-glucose. (d) FITC-particles are

in solution, while un-infected RBC (DAPI-/FITC+, right bars) do not show as significant anRBC is competed out by 0.05% D-glucose. Means (n = 3) and S.D. (error bars) are shown

difference between mean fluorescence intensities (MFI, Table 1) ofthese two groups was significant (P < 0.01, Student’s t-test).

Since targeting of RBC by the particles is postulated to resultfrom affinity to the glucose transporter, two additional experi-ments were conducted to probe the role of inhibiting thisinteraction using a culture of human RBC infected with P.falciparum to parasitemia levels of about 5%.

Infected RBC themselves showed little auto fluorescence (Fig4a). Staining the RBC culture with FITC in solution imparted lessfluorescence to the RBC as compared to incubation with FITC-tagged particles. The mean fluorescence intensity (MFI) of RBCexposed to FITC in solution was 52.8 as against 132 when FITC-tagged particles were used. The distribution of fluorescenceintensities of RBC stained with FITC solution was narrow, sinceall cells rapidly equilibrated with FITC in solution. Particles,

th FITC in solution (curve 2), but even more by FITC-tagged particles (curve 3). (b)in (curve 2) and two logarithmic dilutions (3, 4 Competing-out of FITC fluorescencepreferentially targeted to infected RBC (DAPI+/FITC+, left bars) in comparison to FITC

increase in the proportion of RBC associated. The association of FITC-particles with.

K.D. Heikham et al. / International Journal of Pharmaceutics 483 (2015) 57–62 61

however, imparted varying degrees of fluorescence to differentRBC, presumably proportional to the number of particlesassociated with a cell and/or their localization on the cell surfaceor cytosol. Phloretin, an inhibitor of glucose transporters (Parket al., 2003) reduced the intensity of staining of RBC with FITC-particles in a dose-dependent manner (Fig 4b). Similarly, D-glucose, the natural ligand for GLUT-1 and PfHT-1, competed outfluorescence imparted to RBC by FITC-tagged particles in a dose-dependent manner in iso-osmotic conditions, as shown in Fig 4c.

When iRBC were stained with DAPI and exposed to FITC-particles, it was expected that only the cells bearing the parasitewould be stained with the nuclear dye. It was observed that DAPI+ RBC were preferentially targeted with the label when FITC-particles were used rather than FITC in solution (P < 0.05, Fig 4d).Un-infected RBC (DAPI�/FITC+) did not show as marked adifference in the proportion of RBC associated with FITC whenthe solution and particles were compared. Results depicted in Fig4d also show that the targeting of particles to both RBC and iRBCcould be inhibited by 0.05% D-glucose.

These observations suggest that particles were picked up byboth RBC and iRBC with greater avidity than the fluorescent label insolution. Further, iRBC picked up the particles with even greateravidity than uninfected RBC. Thus, the particles can not onlydeliver their payload to the target cell more efficiently than themolecular solution, they also discriminate between infected anduninfected RBC with higher efficiency.

3.3. Efficacy in vitro

The efficacy of CDRI 97/63 and quinine in solution wascompared with respective particle formulations in a culture ofthe P. falciparum strain 3D7 parasitizing human RBC.

5 10 100 50060008000

10000120001400016000

0 5 10 15 20 2510000120001400016000180002000022000

RFU

a

Conc (ng/ml)

b

Fig. 5. (a) Representative mean (N = 2, �S.D.) reduction of parasite nucleic acidsynthesis by antimalarial agent CDRI 97/63 in solution (triangles), or particlescontaining an equivalent amount (filled circles), especially at 250 ng/ml concen-tration, and blank particles (open circles). (b) Reduction of parasite nucleic acidsynthesis by quinine in solution (triangles), or particles containing an equivalentamount (filled squares), especially at 6.25 and 3.125 ng/ml and blank particles (opensquares).

Representative mean inhibition of parasite nucleic acid synthesis(SYBRGreen incorporation) by CDRI 97/63 or quinine free base, insolution, or incorporated in particles is depicted in Fig 5. Fitting adose (y) versus response (x) curve of the nature:

y ¼ A1 þ A2 � A1

1 þ 10 logx0�xð Þp

Using data shown in Fig. 5; where the A1 and A2 are the bottomand top asymptotes, log x0 is the centre of the dose-response curveand p is the variable Hill slope enabled calculation of the IC50 valuesof soluble and particle-incorporated 97/63 as 463.1 and 244.7 ng/ml respectively. Thus, incorporation of 97/63 in starch particlesreduced its IC50 to nearly half the intrinsic value.

Starch particles containing no drug did not kill parasites whenadded to cultures in amounts equivalent to drug-containingparticles (data not shown).

Drug-containing particles displayed optimal enhancement ofanti-malarial activity in a narrow concentration range. Atconcentrations that were too low to effectively kill the parasite,insignificant differences were observed in the efficacy of drugsolution and particles. Similarly, when drug concentrations werehigh enough to overcome rate-limiting diffusion and provideintracellular parasiticidal concentrations, the efficacy of particleswas no different from that of the drug in solution. However, in awindow of intermediate concentrations, the advantages of drugtargeting using particles were unambiguously apparent.

Thus, 97/63 was 63% more efficacious in the particle formula-tion at a concentration of 250 ng/ml as compared to a solution andthe efficacy of particles containing quinine was higher between theconcentration range of 3.1 and 6.25 ng/ml (20% and 22%). Aboveand below these concentrations, the advantages of drug targetingwere not apparent, such that quinine at a concentration of 25 ng/ml was more active in the soluble form than particles containing anequivalent amount. These results suggest that in vivo doses of theseagents can be reduced proportionately to achieve the sameefficacy. Further experiments are required to establish whetherparticle-incorporated antimalarials would also be targeted to RBC/iRBC in vivo, especially in the presence of cells of the mononuclearphagocyte system that can phagocytose such particles and removethem from circulation.

Parasite cultures were not synchronised for these experi-ments, but as parasitemia rose from �1% to �5% in untreatedwells, all erythrocytic stages of the parasite would have beenpresent during exposure to the drugs or particles. This mighthave resulted in different levels of PfHT1 expression on differentiRBC, leading to differences in particle uptake in proportion tothe density of the protein on the vacuolar membrane. Differen-tial particle uptake by RBC is also suggested by the spread of thehistograms in Fig. 4.

Clinical management of severe/cerebral malaria in India oftenrelies on specific antimalarial chemotherapy with quinine, but theintervention is successful only within a small time-window(Mishra et al., 2006; Sarkar et al., 2010). Current clinical practiceis therefore moving to water-soluble artesunate, especially inpaediatric malaria (Olupot-Olupot and Maitland, 2013). The optionof administering large doses of potent antimalarials like artemisinand artesunate is limited by systemic drug toxicity on the one handand rate-limiting efficacy on the other (Sheehy and Angus, 2011). Adrug delivery strategy that permits administration of an intrave-nous bolus of the drug rather than slow infusion and preferentiallytargets iRBC can therefore offer benefits in severe and cerebralmalaria. It is also interesting to speculate that hypoglycemiaobserved during severe malaria (Das et al., 1988) could privilegethe targeting of particles to glucose transporters. It is submitted

62 K.D. Heikham et al. / International Journal of Pharmaceutics 483 (2015) 57–62

that further in vivo evaluation of the proposed strategy of usingstarch particles or other materials capable of targeting erythrocyteGLUT and PfHT-1 may lead to optimization of an optimal drugdelivery system for use in cerebral malaria.

Acknowledgments

The authors thank Mr. A.L. Vishwakarma and the SophisticatedAnalytical Instruments Facility for assistance with flow cytometry.This is CSIR-CDRI Communication Number 8909, funded by theCouncil of Scientific and Industrial Research through grantsBSC0112 and ESC 0103 and Research Fellowships to HKD, AG,AK, and JS.

References

Barnes, K.I., Watkins, W.M., White, N.J., 2008. Antimalarial dosing regimens anddrug resistance. Trends Parasitol. 24, 127–134.

Croes, K., McCarthy, P.T., Flanagan, R.J., 1994. Simple and rapid HPLC of quinine,hydroxychloroquine, chloroquine, and desethylchloroquine in serum, wholeblood, and filter paper-adsorbed dry blood. J. Anal. Toxicol. 18, 255–260.

Das, B.S., Satpathy, S.K., Mohanty, D., Mohanty, S., Mishra, S.K., Satapathy, P.C.,Patnaik, J.K., Bose, T.K., 1988. Hypoglycaemia in severe falciparum malaria.Trans. R. Soc. Trop. Med. Hyg. 82, 197–201.

Gupta, Y., Jain, A., Jain, S.K., 2007. Transferrin-conjugated solid lipid nanoparticlesfor enhanced delivery of quinine dihydrochloride to the brain. J. Pharm.Pharmacol. 59, 935–940.

Haas, S.E., Bettoni, C.C., de Oliveira, L.K., Guterres, S.S., Dalla Costa, T., 2009.Nanoencapsulation increases quinine antimalarial efficacy against Plasmodiumberghei in vivo. Int. J. Antimicrob. Agents 34, 156–161.

Jain, S.A., Basu, H., Prabhu, P.S., Soni, U., Joshi, M.D., Mathur, D., Patravale, V.B.,Pathak, S., Sharma, S., 2014. Parasite impairment by targeting Plasmodium-infected RBCs using glyceryl-dilaurate nanostructured lipid carriers.Biomaterials 35, 6636–6645.

Mishra, S.K., Mohanty, S., Mohanty, A., Das, B.S., 2006. Management of severe andcomplicated malaria. J. Postgrad. Med. 52, 281–287.

Movellan, J., Urban, P., Moles, E., de la Fuente, J.M., Sierra, T., Serrano, J.L., Fernandez-Busquets, X., 2014. Amphiphilic dendritic derivatives as nanocarriers for thetargeted delivery of antimalarial drugs. Biomaterials 35, 7940–7950.

Nayak, A.P., Tiyaboonchai, W., Patankar, S., Madhusudhan, B., Souto, E.B., 2010.Curcuminoids-loaded lipid nanoparticles: novel approach towards malariatreatment. Colloids Surf. B Biointerfaces 81, 263–273.

Olupot-Olupot, P., Maitland, K., 2013. Management of severe malaria: results fromrecent trials. Adv. Exp. Med. Biol. 764, 241–250.

Park, K.H., Akaike, T., Lee, K.C., 2003. Visualization of specific interaction betweenbiomimetic glycopolymer containing reducing glucose moiety and HepG2 cellmediated by GLUT-1. J. Biomed. Mater. Res. A 65, 136–143.

Ram, E.V.S.R., 2007. Nuclear gyrB encodes a functional subunit of the Plasmodiumfalciparum gyrase that is involved in apicoplast DNA replication. Mol. Biochem.Parasitol. 154, 30–39.

Recuenco, F.C., Kobayashi, K., Ishiwa, A., Enomoto-Rogers, Y., Fundador, N.G., Sugi, T.,Takemae, H., Iwanaga, T., Murakoshi, F., Gong, H., Inomata, A., Horimoto, T.,Iwata, T., Kato, K., 2014. Gellan sulfate inhibits Plasmodium falciparum growthand invasion of red blood cells in vitro. Sci. Rep. 4, 4723.

Sarkar, P.K., Ahluwalia, G., Vijayan, V.K., Talwar, A., 2010. Critical care aspects ofmalaria. J. Intensive Care Med. 25, 93–103.

Sheehy, S.H., Angus, B.J., 2011. Malaria: severe, life-threatening. Clin. Evidence(Online) 2011, 0913.

Singh, C., Puri, S.K., 2001. Substituted 1,2,4-Trioxanes as Antimalarial Agents and aProcess of Producing the Substituted 1,2,4-Trioxanes. In: USPTO (Ed.), Council ofScientific and Industrial Research, New Delhi, IN), United States.

Slavic, K., Krishna, S., Derbyshire, E.T., Staines, H.M., 2011. Plasmodial sugartransporters as anti-malarial drug targets and comparisons with otherprotozoa. Malar. J. 10, 165.

Trager, W., Jensen, J.B., 1978. Cultivation of malarial parasites. Nature 273,621–622.

Valle-Delgado, J.J., Urban, P., Fernandez-Busquets, X., 2013. Demonstration ofspecific binding of heparin to Plasmodium falciparum-infected vs. non-infectedred blood cells by single-molecule force spectroscopy. Nanoscale 5,3673–3680.

Van Soest, J.J.G., Stappers, F.H.M., Van Schijndel, R.J.G., Gottlieb, K.F., Feil, H., 2004.Method for the Preparation of Starch Particles. In: USPTO (Ed.), EcosynthetixInc., Lansing, MI), USA.

Zhang, J.Z., Ismail-Beigi, F., 1998. Activation of Glut1 glucose transporter in humanerythrocytes. Arch. Biochem. Biophys. 356, 86–92.