A blasticidin S-resistant Plasmodium falciparummutant with a defective plasmodial surfaceanion channelDavid A. Hill, Ajay D. Pillai, Fatima Nawaz, Karen Hayton, Lanxuan Doan, Godfrey Lisk, and Sanjay A. Desai*

Laboratory of Malaria and Vector Research, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD 20852

Communicated by Louis H. Miller, National Institutes of Health, Rockville, MD, November 22, 2006 (received for review November 9, 2006)

Erythrocytes infected with malaria parasites exhibit marked in-creases in permeability to organic and inorganic solutes. Theplasmodial surface anion channel (PSAC), an unusual voltage-dependent ion channel induced on the host membrane afterinfection, may play a central role in these permeability changes.Here, we identified a functional PSAC mutant through in vitroselection with blasticidin S. Resistance to blasticidin S was gener-ated during culture and correlated with significant reductions inpermeability to multiple solutes, consistent with uptake via acommon pathway. Single channel recordings revealed markedchanges in PSAC gating with the addition of a subconductancestate not present in wild-type channels. The channel’s selectivityprofile and pharmacology also were significantly altered. Eventualloss of the mutant phenotype upon removal of selective pressureand slower growth of mutant parasites suggest that PSAC servesan important role in intracellular parasite survival. These findingsprovide solid evidence for the uptake of diverse solutes via PSACand implicate one or more parasite genes in expression of thischannel.

drug resistance � malaria � nutrient acquisition � fitness cost

The malaria parasite, Plasmodium falciparum, is a highlysuccessful intracellular pathogen. In the human host, repli-cation in erythrocytes leads to an exponentially increasingnumber of parasites and accounts for much of the clinicalsyndrome of malaria. To facilitate its rapid intracellular growth,the parasite remodels the erythrocyte cytoplasm (1), adds vir-ulence factors to the surface of infected cells (2), and increaseserythrocyte permeability to many small solutes. Increased up-take of solutes may be mediated by the plasmodial surface anionchannel (PSAC) (3). This unusual ion channel has a single-channel conductance of 20 pS in molar Cl� solutions andfast-f lickering voltage-dependent gating. PSAC also has anatypical selectivity profile for anions (SCN� � I� � Br� � Cl�)and effectively excludes Na� despite permeability to bulkycations (4). Various antagonists have quantitatively identicaleffects on PSAC’s open probability (5–7) and uptake of organicsolutes including sugars (8), amino acids (9), purines (10),organic cations (11), and some vitamins (12), suggesting thatPSAC functions as a shared route for the increased uptake ofthese diverse solutes after infection.

Nevertheless, PSAC’s role in solute transport is debatedbecause other ion channels have been observed by some workers(13–16) and because a single ion channel with broad permeabil-ity to diverse solutes but stringent exclusion of Na� is unprec-edented. It is also debated whether PSAC and other putativechannels are parasite-encoded proteins or modified humanproteins (5). Addressing these issues is an important hurdle fortransport studies in malaria and for drug development programsthat seek to target the parasite-induced permeability changes.

Blasticidin S is a fungal toxin (molecular weight 422.2) thatkills most prokaryotic and eukaryotic cells by inhibiting mRNA-directed translation on ribosomes (17). This agent has beenextensively used in molecular transfections of various cell types

because it can select for the expression of detoxifying enzymes.Here, we identify and characterize a P. falciparum mutant thatachieves blasticidin S resistance without the expression of het-erologous resistance genes. This mutant was refractory to syn-chronization with sorbitol, suggesting defects in the parasite-induced permeability changes. Our studies reveal markedlyaltered functional properties of PSAC resulting from changes inthe parasite genome. The altered properties of this mutantchannel are essential for blasticidin S resistance.

This PSAC mutant directly addresses several important ques-tions regarding how malaria parasites transform the erythro-cyte’s permeability. Our findings also have broad implicationsfor molecular biology studies that use blasticidin S in selectivetransfections.

ResultsIdentification and Characterization of a Blasticidin S-Resistant Mu-tant. Blasticidin S has been successfully used in P. falciparum genetictransfection experiments (18–20), including the production of amutant with effective silencing of all members of the large var genefamily (21). In these transfections, resistant parasites were selectedthrough expression of the Aspergillus blasticidin S deaminase(BSD), which converts blasticidin S to a nontoxic deaminohydroxyderivative. We therefore were puzzled when an attempted geneticdisruption in the FCB parasite isolate yielded blasticidin S-resistantparasites that lacked the bsd sequence from the transfection plas-mid, as evaluated by PCR and Southern blotting (data not shown).To definitively establish resistance via parasite-specific mecha-nisms, we challenged several commonly available parasite isolates(FCB, HB3, W2, and 7G8) with blasticidin S under normal in vitroculture conditions. Microscopic examination showed no detectableparasites within 4 days of drug pressure for all of these lines;however, resistant parasites subsequently were recovered fromFCB in each of four independent attempts, but never from the otherthree isolates. We then used limiting dilution to obtain a clonalpopulation of blasticidin S-resistant parasites, hereafter referred toas FCB-br1.

FCB-br1 Exhibits Reduced Permeabilities to Diverse Organic Solutes.During routine in vitro culture, we found that the mutantFCB-br1 parasites were refractory to synchronization in isotonicsorbitol solutions (22). Sorbitol synchronization is based on theincreased permeability of infected erythrocytes to this sugaralcohol. In an isotonic sorbitol solution, mature trophozoite-infected erythrocytes import sorbitol via PSAC and/or otherparasite-induced channels; continued uptake produces an os-

Author contributions: D.A.H. and A.D.P. contributed equally to this work; D.A.H., A.D.P.,K.H., G.L., and S.A.D. designed research; D.A.H., A.D.P., F.N., K.H., and G.L. performedresearch; D.A.H., A.D.P., G.L., and S.A.D. analyzed data; and D.A.H., L.D., and S.A.D. wrotethe paper.

The authors declare no conflict of interest.

Abbreviations: PhTMA, phenyltrimethylammonium; PSAC, plasmodial surface anion channel.

*To whom correspondence should be addressed. E-mail: sdesai@niaid.nih.gov.

www.pnas.org�cgi�doi�10.1073�pnas.0610353104 PNAS � January 16, 2007 � vol. 104 � no. 3 � 1063–1068

PHYS

IOLO

GY

Dow

nloa

ded

by g

uest

on

June

1, 2

021

motic imbalance and leads to cell lysis. Immature ring stageforms survive this treatment because they do not exhibit in-creased sorbitol permeability. Therefore, failure to synchronizeFCB-br1 cultures suggests defects in the parasite-induced sor-bitol permeability.

We evaluated and compared the induced permeabilities inFCB-br1 and its wild-type parent with a quantitative light-scattering assay for cell lysis (23). These experiments revealedstatistically significant decreases in the mutant’s permeabilitynot only to sorbitol, but also to phenyltrimethylammoniumchloride (PhTMA-Cl) and L-proline (Fig. 1A), structurally un-related solutes that also exhibit increased permeability afterinfection. Importantly, the ratios of permeability coefficients forthese solutes indicated that the decreases are not uniform in

magnitude, with sorbitol exhibiting a 16-fold reduction, whereasthe decrease in proline permeability was only 5.0-fold (Student’st test, P � 0.002; Fig. 1B). These findings implicate transport ofthese structurally dissimilar solutes via a shared ion channel; theyalso suggest that functional changes in this channel alter itsselectivity for permeating solutes in the FCB-br1 parasite. Ourobservations are not compatible with simple down-regulation ofan ion channel, which would produce quantitatively similarreductions for each solute.

We also found that sulfosuccinimidyl-6-(biotinamido)hexano-ate, a bulky water-soluble reagent with known PSAC perme-ability (4), had markedly reduced entry into FCB-br1 whencompared with FCB. We quantified the uptake of this reagentby measuring the extent of hemoglobin labeling (Fig. 1C).

Fig. 2. Electrophysiological differences between FCB and FCB-br1. (A) Whole-cell voltage-clamp recordings from erythrocytes infected with FCB or FCB-br1trophozoites with solution A in both bath and pipette compartments. Each group of traces represents superimposed current responses to 50-ms pulses from aholding potential of 0 mV to values between �100 and �100 mV in 10-mV increments. The FCB-br1 isolate induces markedly smaller Cl� currents. (B) Singlechannel recordings in solution A at imposed membrane potential (Vm) of �100 mV. Whereas PSAC activity on FCB is indistinguishable from that on other wild-typeisolates, activity on FCB-br1 is visibly altered. Black dashes on the sides of each trace represent the closed channel level; the subconductance level seen on FCB-br1is demarcated with a red line behind each of those traces. (Scale bars: A, 25 ms and 1.0 nA; B, 6.7 ms and 0.83 pA.) (C) All points histograms calculated from singlechannel recordings as in B for FCB and FCB-br1 isolates (black and red traces, respectively). Each histogram appears as a smooth curve because of narrow bin widths(0.02 pA) and calculation from �2 � 106 samples (20 s of single channel recording). The mean closed (c), open (o), and subconductance (s) levels are marked withone-letter abbreviations.

Fig. 1. The blasticidin S-resistant mutant, FCB-br1, exhibits reduced permeability to diverse solutes. (A) Osmotic lysis kinetics at 37°C in isotonic solutions ofsorbitol, PhTMA-Cl, or proline, as indicated. In each graph, upper and lower traces represent lysis kinetics for FCB and FCB-br1, respectively. Lysis of FCB-br1 isslower in each solute, indicating reduced permeability across the host erythrocyte membrane. (B) Ratios of permeability coefficients (PFCB/PFCB-br1) for the solutesin A, shown as the mean � SEM of four to five measurements. Permeability coefficients were calculated as the reciprocal of the time to 50% lysis, as formallyjustified for passive diffusion via PSAC (23). (C) Immunoblot showing reduced labeling of hemoglobin by sulfosuccinimidyl-6-(biotinamido)hexanoate in FCB-br1parasites (lane 2) when compared with FCB parasites (lane 1), as detected with a monoclonal antibody against biotin. The blot was stripped and reprobed withan anti-hemoglobin antibody to confirm equal protein loading. Because the extent of hemoglobin labeling correlates with reagent permeation via PSAC (4),the blot reflects reduced reagent permeability in FCB-br1.

1064 � www.pnas.org�cgi�doi�10.1073�pnas.0610353104 Hill et al.

Dow

nloa

ded

by g

uest

on

June

1, 2

021

Reduced uptake in the FCB-br1 isolate suggests that this reagententers infected erythrocytes via the same ion channel as sorbitol,PhTMA-Cl, and proline.

Direct inhibition of channels by blasticidin S is unlikelybecause its addition to osmotic lysis solutions at concentrationsup to 100 �g/ml does not reduce solute permeability for eitherFCB-br1 or FCB (data not shown). Washout of blasticidin S byculturing the FCB-br1 parasite for 2 weeks without this toxinrevealed sustained reductions in sorbitol permeability, defini-tively excluding direct effects of blasticidin S or its metabolites.Our studies implicate changes in the parasite genome.

Associated Defects in PSAC. We used the whole-cell voltage-clampmethod to examine changes in conductive Cl� currents inducedby the intracellular parasite. These experiments revealed adramatic reduction in the magnitude of whole-cell currents in theFCB-br1 isolate, but no significant change in their characteristicinward-rectifying voltage dependence (Fig. 2A). To explore themolecular basis of these defects, we obtained low-noise record-ings in the cell-attached configuration (Fig. 2B). Under ourexperimental conditions, only one channel type was seen onerythrocytes infected with FCB; single channel conductance,gating, and voltage dependence matched those previously de-scribed for PSAC on other parasite isolates (3, 5). Identicalexperiments with FCB-br1 revealed PSAC activity with obviouschanges in gating and conductance. Most conspicuously, singlechannel recordings exhibited frequent transitions to a levelbetween the open and closed states, known as a subconductancestate (24). This level was apparent in all points histograms fromextended recordings (Fig. 2C), indicating that these mutatedPSAC channels spend a significant fraction of time in this state.The correlation between defects in PSAC gating and reductionsin both whole-cell currents and organic solute permeabilitystrongly support a central role for PSAC in the transport of Cl�and the various solutes examined in Fig. 1.

We further explored this link with furosemide, a well studiedPSAC antagonist that interacts with an extracellular site distinctfrom the channel pore (5, 25). We found that furosemide wassignificantly less effective at inhibiting proline uptake by eryth-rocytes infected with FCB-br1 (Fig. 3A). Inhibitory dose re-sponses revealed an �40-fold reduction in furosemide affinitywhen compared with the wild-type channel in FCB parasites.Single channel recordings confirmed that less effective inhibitionof the mutant PSAC accounts for this change (Fig. 3B). Reducedpermeability of organic solutes, altered selectivity and gating,and reduced affinity for furosemide suggest that the FCB-br1parasite expresses a PSAC with a globally altered structure.

Reversion of Blasticidin S Resistance in Culture Leads to Loss ofAltered PSAC Activity. To investigate whether blasticidin S resistanceresults from irreversible changes in the parasite genome, we prop-agated FCB-br1 in the absence of blasticidin S and found that thedefective sorbitol lysis phenotype was lost after �4 weeks in culture(FCB-br1-rev in Fig. 4). Restored rapid lysis in sorbitol was asso-ciated with wild-type sensitivity to growth inhibition by blasticidinS. It also correlated with restored PSAC-mediated whole-cellconductances and with loss of the subconductance state in singlechannel recordings. We then rechallenged FCB-br1-rev with blas-ticidin S and grew out a new mutant, which we named FCB-br2.Although FCB-br2 may represent a mixed population of parasitesbecause limiting dilution was not used to generate individual clones,functional studies indicate it also has defective sorbitol uptake,reduced whole-cell Cl� conductance, and defective PSAC gating(Fig. 4). Thus, blasticidin S resistance is tightly linked to theidentified defects in PSAC activity. The relatively rapid reversionwe observed may reflect changes in the expression of genes, one ormore mutation hotspots, or transcriptional switches among paralo-gous genes.

Slower Growth of Parasites Carrying Functional PSAC Mutants. Nu-trient acquisition by the intraerythrocytic parasite may criticallydepend on PSAC-mediated increases in organic solute perme-ability (3). If this hypothesis is correct, the altered PSAC activityexpressed on FCB-br1 mutants may affect adversely parasitegrowth and survival. We therefore quantified in vitro growthrates for FCB and FCB-br1 with and without blasticidin Sselective pressure. When compared in the absence of blasticidinS, we found that FCB-br1 parasites grow significantly slower thanwild-type parasites (Fig. 5), suggesting that defects in PSACadversely affect expansion of in vitro cultures. Because FCB-br1’sgrowth rate was not affected detectably by the addition ofblasticidin S, the selected genomic changes appear to confercomplete resistance to blasticidin S under our conditions.

DiscussionP. falciparum mutants that carry heritable changes in hosterythrocyte permeability have not been described previously.

Fig. 3. Markedly reduced sensitivity to furosemide of PSAC expressed onFCB-br1. (A) (Upper) Osmotic lysis kinetics in proline at 20°C in the presence of0, 10, 25, and 200 �M furosemide (top to bottom traces in each graph,respectively). Proline uptake by FCB-br1 is less well inhibited by furosemide.(Lower) The corresponding dose responses for inhibition. Symbols representthe mean � SEM lysis rates normalized to 1.0 without furosemide, calculatedas described in ref. 23 from up to five measurements at each concentration.Solid lines represent a least-squares best fit to y � K0.5/(K0.5 � x) with K0.5estimates of 0.8 and 32 �M for FCB and FCB-br1, respectively. (B) Single PSACrecordings in solution A with 10 �M furosemide in bath and pipette. Vm ��100 mV. (Scale bars: 100 ms and 2 pA.) Closed channel currents are indicatedby dashes to the sides of each trace. Furosemide significantly reduces channelopenings (downward deflections from the closed channel level) for PSAC onthe FCB isolate but has a negligible effect on FCB-br1 channels when com-pared with their activities without inhibitor (Fig. 2).

Hill et al. PNAS � January 16, 2007 � vol. 104 � no. 3 � 1065

PHYS

IOLO

GY

Dow

nloa

ded

by g

uest

on

June

1, 2

021

The mutant described here can be recognized easily because itscultures cannot be synchronized by incubation in isotonic sor-bitol, as commonly used in malaria research laboratories to studyparasite developmental stages. This method is based on sorbi-tol’s high permeability through wild-type PSAC; it spares im-mature parasite stages and achieves synchronization of culturesbecause PSAC activity is not induced until some 18 h aftererythrocyte invasion. FCB-br1 parasites resist sorbitol synchro-nization because they express PSAC activity with markedly lowersorbitol permeability. These mutant channels also exhibit alteredselectivity for various solutes, defective gating, and reducedaffinity for the inhibitor furosemide. The identification of aPSAC mutant has several important implications for intraeryth-rocytic malaria parasite biology.

Evidence for a Single Ion Channel Shared by Diverse Solutes. Otherreports from electrophysiological studies have suggested thatthere may be multiple distinct ion channels on erythrocyteswith or without infection by malaria parasites (13–16). Apersistent question therefore has been whether different ionchannels mediate the uptake of each solute whose permeabil-ity increases after infection. Although inhibition by multipleinhibitors produces quantitatively parallel effects on PSACsingle channel recordings and the uptake of each organicsolute (5, 6, 26), some workers remain skeptical of this findingbecause available inhibitors may be nonspecific and becauseelectrophysiological methods cannot directly study the trans-port of uncharged solutes. The identified mutant representsindependent evidence for a single ion channel shared by thevarious solutes we tested (Fig. 1). Because this mutant alsoexhibits corresponding reductions in PSAC single channel andwhole-cell currents, the permeability of organic solutes nowcan be confidently linked to PSAC.

Parasite Genes Encode PSAC. Our findings require the involvementof parasite genes in the expression of PSAC activity. Humanerythrocytes lack mutable genetic material; they therefore cannotaccrue changes to account for functional mutants. PSAC gatingpolymorphisms identified through surveys of geographically diver-

gent parasite isolates also insinuate parasite genetic elements (5).PSAC’s unusual selectivity profile (4) and various other functionalproperties (3, 25–27) also distinguish it from known human anionchannels and suggest one or more parasite genes.

PSAC Serves an Important Function for the Intracellular Parasite.Stringent conservation of PSAC’s biophysical properties onphylogenetically divergent malaria parasites suggests that thechannel serves an important biological role for the intraeryth-rocytic parasite (28). One proposal is that it provides the parasitewith access to serum nutrients by increasing their permeabilityacross the host erythrocyte membrane (3); removal of metabolicwaste products also may be facilitated by PSAC. Other possibleroles include volume regulation of infected erythrocytes andchanging the ionic composition of erythrocyte cytosol to supportother parasite activities. Although each of these proposed func-tions is plausible, experimental limitations have thwarted at-tempts to determine PSAC’s exact biological role and have

Fig. 5. In vitro growth rates for FCB (circles) and FCB-br1 (triangles) with andwithout blasticidin S (open and filled symbols, respectively) quantified byusing SYBR Green I-based fluorescent detection of parasite DNA in arbitraryunits. Experiments evaluating growth rates by microscopic determination ofparasitemias produced similar results (data not shown).

BA

C

Fig. 4. Loss of resistance to blasticidin S is linked to the reappearance of wild-type PSAC activity. (A) Osmotic lysis kinetics in sorbitol for FCB (black trace), FCB-br1(red), FCB-br1-rev (green), and FCB-br2 (blue). Wild-type sorbitol permeability is restored in the blasticidin S-sensitive revertant; subsequent selection ofblasticidin S resistance yields reduced sorbitol permeability, similar to that seen with FCB-br1. (B) Correlation between in vitro growth inhibition by blasticidinS (red bar, normalized to 1 for complete microscopic clearance of parasites after 4 days in culture with blasticidin S), whole-cell chord conductance in solutionA (blue bar, mean � SEM of n � 10–29 cells for each isolate), and apparent sorbitol permeability coefficient (green bars, mean � SEM of n � 3–6 trials for eachisolate, calculated as the reciprocal of the lysis halftime). (C) Single PSAC recordings with the same conditions as in Fig. 2B. (Scale bars: 50 ms and 2 pA.) WhereasPSAC on FCB-br1-rev exhibits wild-type gating, FCB-br2 channel gating resembles that of FCB-br1.

1066 � www.pnas.org�cgi�doi�10.1073�pnas.0610353104 Hill et al.

Dow

nloa

ded

by g

uest

on

June

1, 2

021

raised questions regarding whether it serves any function for theintracellular parasite. The identified mutant supports an impor-tant biological role because the observed defects in PSACcompromise in vitro parasite survival. Both the slower growthrate of FCB-br1 and its reversion to the wild-type phenotypeupon removal of blasticidin S are consistent with a fitness costassociated with blasticidin S resistance (29).

How Does Blasticidin S Select for a PSAC Mutant? Blasticidin S hasbeen used as a selective agent in P. falciparum transfections withplasmids carrying the bsd gene. Our demonstration that resis-tance to blasticidin S can be selected without plasmids reveals apotential limitation to its use in transfections. However, we notethat resistance could be generated only from the FCB parasite.This isolate presumably carries a permissive genetic backgroundfor the selection of required changes in PSAC. It is possible thatthis unique background results from polymorphisms in morethan one parasite gene; in this scenario, acquisition of blasticidinS resistance may require multiple changes at the level of theparasite genome.

We propose that the altered PSAC activity in FCB-br1 con-tributes to blasticidin S resistance by reducing host membranepermeability to this toxin. This proposal is consistent with thestructural similarity of blasticidin S to nucleosides with knownpermeability through PSAC (10). It is also consistent with theobservation that blasticidin S has limited membrane permeabil-ity in some cell lines; in these cells, toxicity resulting frominhibition of translation occurs only when blasticidin S perme-ability is increased (30, 31). Finally, blasticidin S-resistant yeast,identified after UV treatment to increase mutation rates, appearto use reduced permeability as the primary mechanism ofresistance (32).

We envision that the selected changes in PSAC serve to alterits selectivity profile and prohibit blasticidin S access to itsintracellular target. Importantly, changes in the channel’sselectivity profile permit maintenance of essential transportfunctions such as adequate levels of nutrient uptake. Moredrastic changes such as production of nonfunctional channelspresumably would be lethal. Although PSAC is an acknowl-edged antimalarial target (7), these findings suggest it also maybe an important route for delivery of other hydrophilic anti-malarials to their intracellular targets. Such approaches willneed to consider resistance via selection of parasites withreduced permeability.

Materials and MethodsIn Vitro Selection of Mutant Parasites. Resistant P. falciparumparasites were selected by culturing under standard conditions inthe presence of 2.5 �g/ml blasticidin S�HCl (Invitrogen, Carls-bad, CA). At this concentration, microscopic examination re-vealed essentially complete sterilization of cultures within 4 daysafter addition of blasticidin S. In experiments with the FCBisolate, rare parasites were detected in smears as soon as 4 weekslater. Transient removal of blasticidin S pressure appeared tospeed the selection of fully resistant parasites. FCB-br1 wascloned from cultures of blasticidin S-resistant parasites by thelimiting dilution technique (33).

Osmotic Lysis Kinetics. Osmotic lysis experiments were per-formed as described in ref. 23. Brief ly, trophozoite-infectedRBCs were enriched to 95–99% by percoll/sorbitol separation(34) with modifications that improve recovery of infectederythrocytes with reduced sorbitol permeability (35). En-riched cells were washed in PBS (150 mM NaCl/20 mMNaH2PO4, pH 7.5), and resuspended at 0.5% hematocrit inbuffered osmotic lysis solution containing 280 mM sorbitol,280 mM proline, or 145 mM PhTMA-Cl supplemented with 20mM Na-Hepes/0.1 mg/ml BSA, pH 7.4. Osmotic swelling and

lysis in these solutions then were continuously followed byrecording transmittance of 700-nm light through the cellsuspension (DU640 spectrophotometer with Peltier tempera-ture control, Beckman Coulter, Fullerton, CA). This methodproduces estimates of permeation rates through PSAC thatquantitatively match those obtained by using either radioiso-tope f lux or patch-clamp recordings (23).

Sulfo-NHS Ester Labeling and Immunoblot. Intact enrichedtrophozoite-stage RBCs were labeled with 0.4 mM sulfosuc-cinimidyl 2-(biotinamido)-ethyl-1,3-dithiopropionate (PierceEndogen, Rockford, IL) at a hematocrit of 0.5% in PBS for 15min at room temperature. After terminating the labeling with100 mM Tris (pH 7.4), the cells were washed, hypotonicallylysed in 20 mM Na2HPO4/1 mM EDTA/1 mM EGTA, pH 8.6,and ultracentrifuged (100,000 � g, 1 h) to separate solubleintracellular proteins from membrane-associated proteins.Soluble fractions corresponding to 7.5 � 105 cells were sepa-rated in a reducing SDS 4–12% polyacrylamide gel, trans-ferred to nitrocellulose membrane, blocked and probed withhorseradish peroxidase-conjugated anti-biotin IgG (JacksonImmunoResearch, West Grove, PA), and visualized with Su-perSignal ECL Substrate (Pierce Endogen). The blot then wasstripped and reprobed with horseradish perox idase-conjugated anti-hemoglobin IgG (Bethyl Laboratories, Mont-gomery, TX).

Electrophysiology. Cell-attached recordings on trophozoite stage-infected RBCs were obtained as described in ref. 5 in symmetricbath and pipette solutions of 1,000 mM choline-Cl/115 mMNaCl/10 mM MgCl2/5 mM CaCl2/20 mM Na-Hepes, pH 7.4(solution A). This hypertonic solution increases the signal-to-noise ratio for single PSAC detection by permitting higher ratesof Cl� f lux through open channels and by reducing pipette RCnoise. When used in single channel recordings, furosemide waspresent at identical concentrations in both pipette and bathcompartments.

The whole-cell patch-clamp configuration was obtained withbrief electrical pulses to disrupt the membrane patch. Whole-cellchord conductances (�) were calculated between �100 mV and0 mV, where PSAC-mediated currents are largest and haveapproximately linear current-voltage profiles in both the wild-type and mutant channels.

Recordings in both configurations used quartz pipettes pulledto tip diameters �0.5 �m and resistances of 1–3 M in solutionA. All recordings were low-pass filtered at 5 kHz (8-pole Bessel)and digitized at 100 kHz. All-points histograms and all otheranalyses were carried out with home-written code.

Parasite Growth Rates. Growth rates for blasticidin S-sensitiveand -resistant parasites were determined by using matchedcultures seeded at 0.5% hematocrit and a parasitemia of 0.2%with and without 2.5 �g/ml blasticidin S. On each subsequentday, identical aliquots were harvested, washed, and subjectedto freeze-thaw before dilution to 0.5% hematocrit with SYBRGreen I nucleic acid stain (used at twice the recommendedconcentration; Invitrogen). Parasite DNA then was quantifiedas a marker of growth by measuring f luorescence (485 nmexcitation, 525 nm emission). Standard curves using serialdilutions of parasite cultures revealed a linear relationshipbetween f luorescence and parasitemia over the range used inour study (data not shown).

We thank Michelle Doll for help with cell culture and Thomas E.Wellems for comments on this work. This research was supported by theIntramural Research Program of the National Institutes of Health,National Institute of Allergy and Infectious Diseases, and by theMedicines for Malaria Venture.

Hill et al. PNAS � January 16, 2007 � vol. 104 � no. 3 � 1067

PHYS

IOLO

GY

Dow

nloa

ded

by g

uest

on

June

1, 2

021

1. Trager W, Rudzinska MA, Bradbury PC (1966) Bull W H O 35:883–885.2. Leech JH, Barnwell JW, Miller LH, Howard RJ (1984) J Exp Med 159:1567–

1575.3. Desai SA, Bezrukov SM, Zimmerberg J (2000) Nature 406:1001–1005.4. Cohn JV, Alkhalil A, Wagner MA, Rajapandi T, Desai SA (2003) Mol Biochem

Parasitol 132:27–34.5. Alkhalil A, Cohn JV, Wagner MA, Cabrera JS, Rajapandi T, Desai SA (2004)

Blood 104:4279–4286.6. Desai SA, Alkhalil A, Kang M, Ashfaq U, Nguyen ML (2005) J Biol Chem

280:16861–16867.7. Kang M, Lisk G, Hollingworth S, Baylor SM, Desai SA (2005) Mol Pharmacol

68:34–40.8. Homewood CA, Neame KD (1974) Nature 252:718–719.9. Ginsburg H, Kutner S, Krugliak M, Cabantchik ZI (1985) Mol Biochem

Parasitol 14:313–322.10. Upston JM, Gero AM (1995) Biochim Biophys Acta 1236:249–258.11. Staines HM, Rae C, Kirk K (2000) Biochim Biophys Acta 1463:88–98.12. Saliba KJ, Horner HA, Kirk K (1998) J Biol Chem 273:10190–10195.13. Staines HM, Powell T, Ellory JC, Egee S, Lapaix F, Decherf G, Thomas SL,

Duranton C, Lang F, Huber SM (2003) J Physiol 552:177–183.14. Duranton C, Huber SM, Tanneur V, Brand VB, Akkaya C, Shumilina EV,

Sandu CD, Lang F (2004) J Gen Physiol 123:417–426.15. Verloo P, Kocken CH, van der WA, Tilly BC, Hogema BM, Sinaasappel M,

Thomas AW, De Jonge HR (2004) J Biol Chem 279:10316–10322.

16. Bouyer G, Egee S, Thomas SL (2006) Blood Cells Mol Dis 36:248–254.17. Yamaguchi H, Yamamoto C, Tanaka N (1965) J Biochem (Tokyo) 57:667–677.18. Mamoun CB, Gluzman IY, Goyard S, Beverley SM, Goldberg DE (1999) Proc

Natl Acad Sci USA 96:8716–8720.19. Wang P, Wang Q, Sims PF, Hyde JE (2002) Mol Biochem Parasitol 123:1–10.20. Sidhu AB, Verdier-Pinard D, Fidock DA (2002) Science 298:210–213.21. Dzikowski R, Frank M, Deitsch K (2006) PLoS Pathog 2:e22.22. Lambros C, Vanderberg JP (1979) J Parasitol 65:418–420.23. Wagner MA, Andemariam B, Desai SA (2003) Biophys J 84:116–123.24. Hamill OP, Sakmann B (1981) Nature 294:462–464.25. Desai SA (2005) Nanomed Nanotech Biol Med 1:58–66.26. Lisk G, Kang M, Cohn JV, Desai SA (2006) Eukaryot Cell 5:1882–1893.27. Baumeister S, Winterberg M, Duranton C, Huber SM, Lang F, Kirk K,

Lingelbach K (2006) Mol Microbiol 60:493–504.28. Lisk G, Desai SA (2005) Eukaryot Cell 4:2153–2159.29. Andersson DI, Levin BR (1999) Curr Opin Microbiol 2:489–493.30. Contreras A, Carrasco L (1979) J Virol 29:114–122.31. Iordanov MS, Pribnow D, Magun JL, Dinh TH, Pearson JA, Chen SL, Magun

BE (1997) Mol Cell Biol 17:3373–3381.32. Ishiguro J, Miyazaki M (1985) Curr Genet 9:179–181.33. Rosario V (1981) Science 212:1037–1038.34. Aley SB, Sherwood JA, Howard RJ (1984) J Exp Med 160:1585–1590.35. Alkhalil A, Hill DA, Desai SA (November 3, 2006) Cell Microbiol, doi:10.1111/j.

1462-5822.2006.00834.x.

1068 � www.pnas.org�cgi�doi�10.1073�pnas.0610353104 Hill et al.

Dow

nloa

ded

by g

uest

on

June

1, 2

021