MOLECULAR

ii&iEMIcAL PARASITOLOGY

Molecular and Biochemical Parasitology 74 (1995) 157-166

Analysis of pfmdrl and drug susceptibility in fresh isolates of Plasmodium falciparum from Subsaharan Africa

Leonardo K. Bascoa, Jacques Le Bras”, Zane Rhoadesb, Craig M. Wilson*b

aLaboratoire de Parasitologic. H6pital B&hat-Claude Bernard, Paris 75877, France bDivision of Geographic Medicine and Pediatrics, VAB BBRB 562, School of Medicine, University of Alabama at Birmingham,

Birmingham, AL 35294-2170, USA

Received 5 May 1995; revision received 17 July 1995; accepted I7 July 1995

Abstract

Resistance of Plasmodium falciparum to many therapeutic agents is an increasing problem in most endemic areas. The role of the mdr-like gene products of P. falciparum in resistance to quinoline-containing compounds is not clear. The purpose of this study was to further examine the role of pfmdrl in drug resistance in fresh clinical isolates originating from Africa. Drug susceptibility testing (chloroquine, mefloquine, halofantrine and quinine) and a molecu- lar analysis of pfmdrl was completed for 51 fresh clinical isolates. A statistical association between the chloroquine sensitivity phenotype and an intragenic allele of pfmdrl was noted at a position, amino acid 86, which was previously associated with chloroquine resistance. There was little variation in the other intragenic alleles previously associated with chloroquine resistance. No correlation between pfmdrl intragenic allelic variation and susceptibility to meflo- quine, halofantrine or quinine was found. There was no association between gene copy number of pfmdrl and any drug resistant phenotype in an analysis of selected isolates. This, along with other data, suggests that mefloquine resis- tance may have arisen by two different mechanisms in African and Southeast Asian isolates. Much more variability in the polyasparaginated region of the pfmdrl gene was noted in this study than previously reported. In addition, fingerprint analysis using multiplex PCR revealed considerable genetic variability among these isolates.

Keywords: Malaria; Plasmodium falciparum; Drug resistance; Multiple drug resistant; Chloroquine; Mefloquine

1. Introduction

Resistance of P. falciparum to therapy with quinoline-containing compounds is a rapidly

evolving problem worldwide. The molecular

Abbreviations: mdr, multi-drug resistant; PCR, polymerase chain reaction; CLQ, chloroquine; MFQ, mefloquine; HAL, haolfantrine; QUIN, quinine.

* Corresponding author, Tel.: +I 205 9757608; Fax: +I 205 9335671.

mechanisms of P. falciparum involved in resistance against these drugs are still unclear [1,2]. The chloroquine (CLQ)-resistant phenotype is associ-

ated with drug efflux, analogous to an enhanced eMux of cytotoxic drugs in multi-drug resistant (mdr) mammalian cancer cells [3]. Resistance to most quinoline-containing drugs can be reversed

in both malaria parasites and cancer cells by co- administration of drugs referred to as resistance modulators [4]. In cancer cells, the mdr phenotype

Ol66-6853/95/$09.50 0 1995 Elsevier Science B.V. All rights reserved SSDI 0166-6851(95)02492-V

158 L.K. Basco et al. /Molecular and Biochemical Parasitology 74 (1995) 157-166

is associated with the amplification of the mdr gene. The overexpressed protein product, P- glycoprotein, functions as an efflux pump of cyto- toxic drugs and is the molecular target of the modulators [5,6]. A homologous protein has been identified and localized to the vacuolar membrane of P. falciparum [7].

On the basis of sequence homology with the mammalian mdr genes, putative drug resistance genes of P. falciparum, designated as pfmdrl and pfmdr2, have been cloned and sequenced [8- lo]. Two possible genetic alterations in pfmdrl have been associated with resistance to quinoline- containing compounds. Gene amplification of pfmdrl has been more closely linked to mefloquine (MFQ) and halofantrine (HAL) resistance in both in vitro-selected and clinical isolates of P. falciparum [9,11,12]. Conversely, a decreasing gene copy number of pfmdrl has been associated with increasing sensitivity to MFQ and decreasing sensitivity to CLQ in in vitro-selected parasites [ 13,141. Several point mutations in thepfmdrl gene were initially correlated with CLQ resistance [ 151. However, other more geographically limited stud- ies did not corroborate these findings [ 11,161. The pfmdr2 gene is probably unrelated to drug resis- tance since it is neither amplified nor mutated, and its expression level does not differ between sensi- tive and resistant parasites [ 10,171.

Previous studies on the pfmdrl gene have used culture-adapted strains of P. falciparum originating mainly from Southeast Asia, Papua New Guinea, and South America [8,9,1 l-171. Culture-adapted malaria parasites, however, are not fully representative of naturally occurring parasites since continuous in vitro culture induces the selection of subpopulations of parasites [ 18,191. In addition, a large majority of P.

falciparum isolates found in Southeast Asia, Papua New Guinea, and Brazil are CLQ-resistant, mak- ing a comparison between CLQ-susceptible and CLQ-resistant parasites difficult.

The purpose of this study was to examine the re- lationship between the pfmdrl gene and drug resis- tance in fresh clinical isolates to eliminate any potential bias that might be introduced by using in vitro-selected parasites. Fifty-one fresh clinical isolates of P. falciparum representing naturally-

occurring parasites in Africa were acquired for this analysis. These isolates were examined for drug sensitivity, a molecular analysis of pfmdrl was completed and the isolates were fingerprinted using a multiplex PCR technique. The results of these studies are described below.

2. Materials and methods

2.1. Parasites The CLQ-susceptible L-3/C6te d’Ivoire and L-

161Sierra Leone clones and the CLQ-resistant FCM29/Cameroon clone were maintained in a continuous culture using standard methods. Fifty- one clinical isolates were obtained at the Bichat- Claude Bernard Hospital, Paris, from malaria- infected travelers returning to France from various African countries between the months of June through November, 1993. Blood samples were drawn in titrated anticoagulant, washed in culture media with physical removal of the huffy coat layer of cells. The erythrocytes were cultured using standard techniques for 24-36 h to obtain schizonts before extracting DNA. The CLQ- susceptible 3D7 clone was used as a control for the pfmdrl gene copy number experiments [8,9].

In vitro susceptibility to CLQ, quinine (QUIN), MFQ and HAL of each isolate was determined using the semi-microtest [20]. The threshold 50% inhibitory concentration (ICsO) values for resis- tance to CLQ, QUIN, MFQ and HAL were 100, 600, 30 and 6 nM, respectively and are based on our previously published work [21,22]. The results of the sensitivity testing were kept in France until the sequence analysis of pfmdrl was completed, at which time the data sets and predictions were exchanged.

2.2. DNA sequencing DNA was extracted as previously reported.

Briefly, free parasites were isolated with saponin in NET buffer (150 mM NaCl, 10 mM EDTA, 50 mM Tris, pH 7.5), washed and resuspended in NET buffer and treated with 1% Sarkosyl, RNAse A (100 mg/ml, 1 h at 37”(Z), and proteinase K (200 mg/ml, 2 h at 50°C). DNA was extracted by sequential phenol, phenol-chloroform and chloroform treatment, and precipitated with

L. K. Basco et al. /Molecular and Biochemical Parasitology 74 (1995) 157-166 159

Table 1 Drug susceptibility profiles of four strains from which pfmdrf was fully sequenced

Parasite ICsa(nM)

Chloro- Meflo- Halo- Quinine quine quine fantrine

CLQ-susceptible L-3KBte d’lvoire L- 161Sierra Leone

CLQ-resistant FCM 29Xameroun BUR l/Burundi

34.9 33.2 6.88 236 35.9 92.6 11.6 252

1670 20.8 2.98 737 405 6.12 0.956 333

The results are given as 50% inhibitory concentration values from a standard 48-h hypoxanthine incorporation assay. Threshold values for resistance to chloroquine, mefloquine, halofantrine and quinine are 100,30,6 and 600 nM, respective- ly. L-3, L-16, and FCM 29 are clones; BUR I is a culture- adapted isolate.

ethanol. The DNA samples were shipped to the USA suspended in ethanol.

The entire sequence analysis ofpfmdrl was per- formed using a PCR-based single-stranded se- quencing protocol previously described [ 111.

The drug susceptibility studies were completed

Table 2 Sequence analysis of pfmdrl from the four strains described in Table 1

in France and were correlated with the molecular analysis data once the latter was completed in the USA. The prediction of CLQ susceptibility was ultimately based on the amino acid residue at posi- tion 86, with the presence of Tyr being indicative of CLQ resistance and Asn being CLQ-susceptible [15]. The statistical association between the in- tragenic alleles of the pfmdrl gene and CLQ sus- ceptibility was analyzed by a chi-square test.

2.3. DNA fingerprint Genomic variation of the African isolates was

characterized by a multiplex polymerase chain reaction using primers that are specific for genes encoding merozoite surface antigens 1 and 2 (MSA-1 and MSA-2) and circumsporozoite pro- tein (CSP), as described by Wooden et al. [23]. The amplified fragments of the polymerase chain reac- tion were separated by electrophoresis in 1.5% ag- arose gel and visualized with ethidium bromide and UV transillumination.

2.4. Gene copy number The gene copy number ofpfmdrl was quantified

as previously described using CSP as a single copy gene control [ 1 I]. The radioactive hybridization

Parasite Amino acid residue?

CLQ-S L-3/CBte d’Ivoire L- 161Sierra Leone

CLQ-R FCM 29Kameroon BUR 1 /Burundi

PredictionC CLQ-S CLQ-R (SE Asia)

CLQ-R (S America)

86

AAT AAT

TAT TAT

AAT TAT AAT

184

T-f-f T-l-f

TAT TAT

TAKf TAT TAT

1034

AGT ACT

AGT ACT

AGT AGT TGT

1042

AAT AAT

AAT AAT

AAT AAT GAT

1246 PolyNs b

GAT 8/l/8 CAT 71216

GAT 7/2/10 GAT 7/o/o

GAT - CAT - TAT

The full-length pfmdrf sequence of four African parasites was determined as described in the text. No new alleles were identified in this analysis. aAmino acid residue numbering is as determined in Ref. [8]. bPolyNs represents the polyasparaginated hinge region as described in the text. The data is given as the sequential number of predicted Asn/Asp/Asn residues. ‘Prediction of chloroquine sensitivity is based on the wild-type genotype and the resistant Southeast Asian and South American genotypes are as described in Ref. [15].

160 LX. Basco et al. /Molecular and Biochemical Parasitology 74 (199s) 157-166

signal was quantitated using a PhosphorImager 4OOE (Molecular Dynamics, Sunnyvale, CA) and Image Quant software for both the CSP and pfmdrl gene. Ratios of the quantitated signals were established for each of the isolates and com- pared to the ratio calculated for 3D7 which was set as the unity value.

3. Results

as described above. Twenty-four isolates were susceptible to CLQ; the other 27 isolates were CLQ-resistant. There were CLQ-resistant isolates from all geographic regions with a slight bias toward isolates obtained from countries east of Nigeria. Five isolates were resistant to MFQ and one of these isolates was resistant to HAL. All of the isolates were sensitive to QUIN. The drug sen- sitivity profiles and country of origin of these isolates is included as part of Tables 3 and 4.

3.1. Isolates and sensitivity testing The 51 isolates used in this study are from 19 3.2. pfmdrl sequence analysis

different countries of Subsaharan Africa and rep- The full-length sequence of pfmdrl from three resent a broad geographic sampling. In vitro sus- reference clones and one of the isolates was exam- ceptibility testing was completed on all 51 isolates ined looking for any allelic variants not previously

Table 3

pfmdrl sequence of 24 chloroquine-susceptible African isolates

Isolatesa Amino acid residuesb

86 184 1034 1042 1246 PolyNc

2 SEN 2

3SEN4

4GUB 1

5GUI 1

6 SIL 1

7 SIL 2 MFQ

8 SIL 4 MFQ 9 MA1 I MFQ 10 MAI 4

II MA1 5

12 IVC 1 MFQ, HAL

13 IVC 2

14 IVC 6

15 UPV 1

16 UPV 2

IITGGI

18CHA 1 19 CNG 1

20 CNG 3 21 ZAI 1

22 ZAI 2

23 ZAM 1

24 MAD 1

TAT*

AAT

AAT

AAT

AAT

AAT

AAT

AAT

AAT

TAT*

AAT

AAT

AAT

AAT

AAT

AAT

AAT

TAT*

AAT

AAT

AAT

AAT

AAT

-l-I-r

l-I-r

TAI-M

I-I-T

TAT

TAl-I-f

TAJTT

TAi-l-T

TTT

I-I-r

TTr

TAT

TAT

TAT

Tl-r TAT

TAT

TTT

TAAT

TAT

TAT

TAT

I-l-f

AGT

AGT

AGT

AGT

AGT

AGT

AGT

AGT

AGT

AGT

AGT

AGT

AGT

AGT AGT

AGT

ACT

ACT

AGT

AGT

ACT

AGT

A AT GAT

AAT GAT

AAT GAT

AAT GAT

AAT GAT

AAT GAT

AAT GAT

AAT GAT

AAT GAT

AAT GAT

AAT GAT

AAT GAT

AAT GAT

AAT GAT

AAT GAT

AAT GAT

AAT GAT

AAT GAT

AAT GAT

AAT CAT

AAT GAT

AAT GAT

AAT GAT

8/2J9

lnll

I/2/10

71218

I/2/9

7/2/7

IQl6

InllO

Il2l8

8/2/8

7l2lI

ll2l8

810/O

7l2llO

II219

l/2/8

I/2/8

71217

7/2l6-8

8/2l8-9

7/2/7-9

l/2/8

II219

‘Isolates are designated by the international code of their country of origin, followed by a number. The isolates marked MFQ and HAL are mefloquine and halofantrine resistant by sensitivity tesring. bAmino acid residue numbering is as determined in Ref. [8]. The prediction of CLQ was based on the criteria established in Ref.

[I$ In these isolates, the prediction was based totally on the amino acid residue at position 86 as described in the text. The isolates

which were incorrectly predicted are marked (*). CPolyNs represents the polyasparaginated hinge region as described in the text. The data is given as the sequential number of

predicted Asn/AsplAsn residues. Polyclonality is indicated by the variable number of Asn residues in the same isolate.

L.K. Basco et al. /Molecular and Biochemical Parasitology 74 (1995) 157-166 161

identified. The in vitro drug susceptibility pattern of each of these parasites is presented in Table 1. The CLQ-resistant BURI isolate presented an identical sequence as that of FCM 29, with the exception of the absence of parts of the polyasparaginated region (Table 2). Except for two intragenic alleles (amino acids 86 and 184) and the polyasparaginated region, the DNA sequences of the two CLQ-susceptible ~-3 and ~-16 clones were identical to that of FCM 29. In comparison with the previously published sequences [8,11,151,

Table 4 pfmdrl sequence of 27 chloroquine-resistant African isolates

Isolatesa Amino acid residuesb

no new intragenic alleles were found in these clones.

For the 51 clinical isolates, the genomic se- quences of four codons previously associated with CLQ resistance, one variable codon not associated with drug resistance and the polyasparaginated region were determined. This data is presented in Tables 3 and 4 where the isolates are divided by their CLQ sensitivity. There was variation at nu- cleic acid residue 754 (A-T) and therefore at predicted amino acid 86 (Asn to Tyr) which had

86 184 to34 1042 I 246 PolyNf

1 SEN 3

2GUI2

3 SIL 3

4 SIL s

5 SIL 6

6MAI2

7 MA1 3

8 IVC 3

9 IVC 4

10 IVC 5

11 UPV 3

12 BEN I

13 CAE 1

14 CAE 2

15 CAE 3

16 CAE 4

17CAE5

18CAE6

19RCA 1

20 CNG 2 21 RWA I

22 BUR I

23 BUR 2

24 BUR 3 MFQ 25 BUR 4

26 BUR 5

27 COM 1

TAT

AAT*

AAT*

AAT*

TAT

TAT

TAT

TAT

TAT

TAT

AAT*

TAT

TAT

TAT

TAT

TAT

TAT

AAT*

AAT*

TAT

TAT

TAT

TAT

AAT’

A/TAT

TAT

TAT

TTr AGT TTr AGT

TAT AGT

TTT AGT

T-IT AGT

I-I-I AGT

T-l-r AGT

TAT AGT

I-I-I AGT

T-I-I AGT

TAT AGT

TTT AGT

TAT AGT

I-I-I AGT

I-l-I- AGT

T-IT AGT

I-I-I AGT

TAT AGT

I-I-I AGT

TAT AGT

TAT AGT

TAT AGT

TAT AGT

TTT AGT

TAK’I AGT

TAT AGT

TAT AGT

AAT GAT 81219

AAT GAT 7/2l8-9

AAT GAT 71219

AAT GAT 71218

AAT GAT 71218

AAT GAT 7/2/9

AAT GAT 81218

AAT GAT 7/2/10

AAT GAT 71217

AAT GAT ml9

AAT GAT 7/2/10

AAT GAT 81216

AAT TIGAT 7/2/9- IO

AAT GAT 81219

AAT GAT al219

AAT GAT 71218

AAT GAT 81219

AAT GAT 71219

AAT GAT 71318 AAT TAT 712110

AAT GAT 8/O/O AAT GAT 7/o/o

AAT GAT 71219

AAT GAT 71217-8 AAT T/GAT 7l2l8

AAT GAT 71219

AAT GAT 71219

Ysolates are designated by the international code of their country of origin, followed by a number. The isolates marked MFQ are

mefloquine resistant by sensitivity testing.

bAmino acid residue numbering is as determined in Ref. [8]. The prediction of CLQ was based on the criteria established in Ref.

[IS]. In these isolates, the prediction was based totally on the amino acid residue at position 86 as described in the text. The isolates which were incorrectly predicted are marked (*).

‘PolyNs represents the polyasparaginated hinge region as described in the text. The data is given as the sequential number of

predicted Asn/Asp/Asn residues. Polyclonality is indicated by the variable number of Asn residues in the same isolate.

162 L.K. Basco et al. /Molecular and Biochemical Parasitology 74 (199s) 157-166

previously been associated with CLQ resistance. Twenty-two isolates (19 CLQ-resistant, three CLQ-susceptible) had TAT (Tyr) at nucleic acid position 754. Twenty-eight isolates (seven CLQ- resistant, 21 CLQ-susceptible) had AAT (Asn) at the same position. One CLQ-resistant isolate had both Tyr and Asn, indicating its polyclonality. Based on the predicted amino acid residue 86 (TAT = CLQ-resistant; AAT = CLQ-susceptible), CLQ susceptibility was correctly predicted in 80% (40150) of the isolates. The association of the amino acid 86 and CLQ susceptibility was statistically significant (P < 0.05).

At nucleic acid position 1049, the sequence was TAT (Tyr-184) in 20 isolates (11 CLQ-resistant, nine CLQ-susceptible) and TIT (Phe-184) in 26 isolates (15 CLQ-resistant, 11 CLQ-susceptible). Five isolates had both Tyr and Phe. Statistical analysis showed no significant difference (P > 0.05) of the sequences at position 184 between the CLQ-resistant and the CLQ-susceptible isolates. This variation in amino acid 184 had been noted previously, and similarly had not been linked with CLQ susceptibility. No isolates had the allelic changes associated with the ‘South American’ CLQ-resistance genotype identified by Foote et al. [ 151 (see Table 2). Only three isolates had TAT (Tyr-1246) or a mixture of GAT (Asp-1246) and TAT at nucleic acid position 4234; the remainder had Asp. The amino acids Ser-1034 and Asn- 1042 were invariable in all 5 1 isolates, as well as in three African clones.

Fig. 1. Multiplex PCR fingerprint analysis of African isolates. Multiplex PCR was performed using oligonucleotide primer pairs for CSP, MSA-I and MSA-2 as described in the text. This is a representative picture of the ethidium bromide-stained, resolved PCR products showing the nearly unique pattern for each isolate tested.

CSP, demonstrated the genomic variation of the isolates. The results of representative isolates are shown in Fig. 1. Analysis of the fingerprinting data shows that, with the possible exception of two pairs of isolates, each isolate appears unique.

3.4. Gene copy analysis of pfmdrl

The polyasparaginated linker region of the pfmdrl gene of these isolates was highly variable. Previous studies had found this region (nucleic acids 2426-2483) to have seven predicted Asn followed by two Asp residues and then a variable number of Asn residues, usually six to ten. This analysis showed a predicted initial seven or eight Asn residues, followed by two Asp residues (occa- sionally one or three) and another series of six to ten Asn residues. However, in three isolates, the Asp residues and the second series of Asn residues were totally absent.

The pfmdrl gene copy number in relation to CLQ and MFQ susceptibility is summarized in Table 5. Adequate quantities of DNA to complete these experiments in a reproducible manner were available for 13 isolates and two clones. Of the 15 DNA samples analyzed for gene amplification, only two isolates (IVC3, MADl) clearly had amplified the pfmdrl gene. The remainder had a single (one to two) copy of the gene. No correla- tion between pfmdrl gene copy number and any drug resistance could be determined.

4. Discussion

3.3. Isolate fingerprinting In the present study, the P. falciparum geno- A multiplex PCR fingerprinting technique, types and phenotypes were characterized by the in-

using the primers specific for MSA- 1, MSA-2 and tragenic allelic sequences of the pfmdrl gene,

L.K. Basco et al. /Molecular and Biochemical Parasitology 74 (1995) 157-166 163

Table 5 pfmdrl gene copy number and drug susceptibility

Isolate Gene copy number

CLQ- and MFQ-susceptible GUI SEN 1 UPV 1 ZAM 1 SIL 1 CNG 1 IVC 6

CLQ-resistant, MFQ-susceptible IVC 3 IVC 5 CAE 6

CLQ-susceptible, MFQ-resistant SIL 2 MAD 1 MA1 1 L-3/CBte d’lvoire (clone)

CLQ- and MFQ-resistant FCM 29Kameroon (clone)

1-2 1 1 l-2 l-2 l-2 l-2

2-4 1 l-2

1 l-3 1 1

1

The threshold 50% inhibitory concentration (ICN) value for chloroquine resistence was fixed at 100 nM. All mefloquine- susceptible isolates had I&, values ~20 nM. The ICw values of mefloquine-resistant isolates were 53.7 (SIL 2), 43.5 (MAD 1) and 30.6 nM (MA1 1). FCM 29 clone was considered to be mefloquine-resistant (range of IC, 18-32 nM). The gene copy number is expressed as the range of three to seven independent quantitation experiments. Numbers given as ranges represent intermediate values not resolved with multiple determinations and probably indicate polyclonality of the isolates.

polymorphism of the polyasparaginated hinge region, pfmdrl gene copy number, polymorphism of the MSA-1, MSA-2 and CSP genes, and in vitro drug susceptibility to CLQ, QUIN, MFQ and HAL. The main purpose of this study was to com- plete sensitivity testing and a molecular analysis of pfmdrl in fresh clinical isolates of P. falciparum. Previous studies looking at these same correlations had all been completed on in vitro-stabilized para- sites. The molecular studies described were com- pleted from DNA extracted within 24-36 h of sample collection to remove as much bias as possi- ble from in vitro selection of parasites.

The clinical isolates in this study represent a broad sampling of malaria isolates from Sub- saharan Africa with a bias toward the fran- cophone regions. The drug sensitivity testing

results were consistent with previous findings; CLQ resistance is widespread, and MFQ resis- tance and CLQ resistance occur most commonly as unrelated phenotypes in African isolates. This is in contrast to isolates from Southeast Asia where MFQ resistance developed as an additional phenotype in isolates which were already resistant to CLQ and many other compounds. Only one isolate from Burundi (BUR 3) was resistant to CLQ and MFQ, however this isolate was sensitive to QUIN and HAL.

A remarkable degree of genetic heterogeneity was noted from the molecular studies on these isolates. Nearly all of the isolates could be unique- ly distinguished using the multiplex PCR analysis. The genes which were used in this analysis repre- sent surface proteins with conserved regions flank- ing regions of variable lengths. The degree of variability observed here suggests a substantial di- versity among these potential immunogens. Other studies have found a high degree of heterogeneity among isolates but few have established that all of the isolates were essentially unique [24-27). In ad- dition, there was considerably more heterogeneity noted in the polyasparaginated region of the pfmdrl gene than had previously been noted [8,11,16]. This may reflect the true heterogeneity in non-in vitro-selected parasites. The function of the polyasparaginated region of pfmdrl is unknown and there is no association of the variation of this region with any phenotype.

The sequence analysis ofpfmdrl in these isolates did not reveal any new intragenic allelic variations. There was very little variation seen at predicted amino acid residues 1034, 1042 and 1246, which had previously been used to predict CLQ resis- tance in isolates from South America. There was considerable variation noted at predicted amino acid 184, which did not correlate with CLQ resis- tance. This is in agreement with previous studies [ 11,151. No correlation of any intragenic allelic variation and drug susceptibiltiy to quinine, halofantrine or mefloquine was found.

However, the results of the present study in- dicate that a single point mutation, which is pre- dictive of an amino acid change at position 86 (Asn to Tyr), is statistically associated with CLQ resistance. Using this single criterion, the CLQ

164 L. K. Basco et al. /Molecular and Biochemical Parasitology 74 (1995) 157-166

susceptibility of 40-50 isolates (one isolate had a mixture of amino acids) and three of three clones was correctly predicted. This is similar to the pro- portion of correct predictions (25127) by Foote et al. in African, Asian and Papua New Guinean culture-adapted isolates and clones where amino acid 86 was used to predict CLQ sensitivity [15]. Two studies which used culture-adapted parasites from limited geographic locations, the Sudan [16] and Southeastern Thailand [ 111, did not find any correlation of this particular point mutation with CLQ resistance. The association of a single point mutation with CLQ resistance is also teleologically unsatisfying. CLQ resistance did not arise clinical- ly as if it were associated with a single point muta- tion; that is, it took nearly 20-25 years to appear from the time the drug went into widespread use. The statistical association of a single point muta- tion in pfmdrl with CLQ resistance does suggest a linkage between pfmdrl and CLQ resistance, as has been suggested with other studies [13,14], but ultimately other genetic and molecular factors are involved [2,28,29].

There was no correlation found in this study be- tween gene copy number of pfmdrl and drug resis- tance phenotype. In the present study, only one of three CLQ-resistant isolates (IVC 3) and one of four MFQ-resistant isolates (MAD 1) had an amplified gene. Previous studies have also observed multiple copies of pfmdrl in some CLQ resistant isolates [8]. Two recent studies have observed the gene copy number of pfmdrl to decrease with in vitro back-selection for CLQ re- sistance in clones which had previously been se- lected for MFQ resistance and had increased pfmdrl copy number 113,141. Our previous studies have shown a correlation of increasing gene copy number with MFQ and HAL resistance in isolates from Southeast Asia [9,11]. The isolates from Southeast Asia were always also resistant to CLQ, QUIN and other unrelated compounds. This is in contrast to the isolates reported here and in other studies from the African continent where MFQ and CLQ resistance usually occur discretely. This suggests that MFQ resistance may have two dif- ferent phenotypes depending on the background resistance to other compounds. This is supported

by the observation that MFQ resistance in isolates from Southeast Asia can be modulated with verapamil while verapamil has no effect on MFQ resistance on isolates from Africa (Dennis E. Kyle, personal communication, and Ref. [30]).

In conclusion, this study does support a linkage between pfmdrl and CLQ resistance in fresh isolates of P. falciparum from Subsaharan Africa. This suggests the pfmdrl gene product may be one of the factors in a multifactorial process leading to CLQ resistance. There was no correlation in this study between MFQ resistance and gene copy number ofpfmdrl. This suggests a difference in the mechanism by which MFQ resistance has arisen in the African and Southeast Asian regions. Finally, the remarkable degree of genetic heterogeneity in the three genes which were used for fingerprinting these African isolates suggests a level of antigenic variability that is daunting for the prospects of vaccine development.

Acknowledgments

We are grateful to Dr. Philippe Deloron (IN- SERM U 13, Paris, France) for providing the para- sites from Burundi, to Dr. Tom McCutchan of NIH for the CSP clone and to Drs. Tom Unnasch, Naomi Lang-Unnasch and Anne Vezza for their careful review of this manuscript. Part of this work was supported by NIAID grant Kll AI00892-05 (CMW).

References

111

121

[31

141

Karcz, S. and Cowman, A.F. (1991) Similarities and dif- ferences between the multidrug resistance phenotype of mammalian tumor cells and chloroquine resistance in Plasmodium falciparum. Exp. Parasitol. 13, 233-240. Bray, P.G. and Ward, S.A. (1993). Malaria chemo- therapy: resistance to quinoline containing drugs in Plasmodium falciparum. FEMS Microbial. Lett. 113, l-7. Krogstad, D.J., Gluzman, I.Y., Kyle, D.E., Oduola, A.M., Martin, S.K., Milhous, W.K. and Schlesinger, P.H. (1987) Efllux of chloroquine from Plasmodium falciparum: Mechanism of chloroquine resistance. Sci- ence 238, 1283-1285. Martin, S.K., Oduola, A.M.J. and Milhous, W.K. (1987)

LX. Basco et al. /Molecular and Biochemical Parasitology 74 (1995) 157-166 165

Reversal of chloroquine resistance in Plasmodium falciparum by verapamil. Science 235, 899-901.

[5] Riordan, J.R., Deuchars, K., Kartner, N., Alon, N., Trent, J. and Ling, V. (1985) Amplification of P- glycoprotein genes in multidrug-resistant mammalian cell lines. Nature 316, 817-819.

[6] Comwell, M.M., Safa, A.R., Feldsted, R.L., Gottesman, M.M. and Pastan, (1986) Membrane vesicles from multidrug-resistant human cancer cells contain a specific 150 to 170 kDa protein detected by photoafftnity label- ing. Proc. Natl. Acad. Sci. USA 83, 3847-3850.

[7] Cowman, A.F.. Karcz, S., Galatis, D. and Culvenor, J.G. (1991) A P-glycoprotein homologue of Plasmodium jidciparum is localized on the digestive vacuole. J. Cell Biol. 113, 1033-1042.

[8] Foote, S.J., Thompson, J.K., Cowman, A.F. and Kemp, D.J. (1989) Amplification of the multidrug resistance gene in some chloroquine-resistant isolates of P. falciparum. Cell 57. 921-930.

[9] Wilson. C.M., Serrano, A.E., Wasley, A., Bogenschutz, M.P., Shankar, A.H. and Wirth. D.F. (1989) Amplifica- tion of a gene related to mammalian mdr genes in drug- resistant Plasmodium falciparum. Science 244, 1184-1186.

[IO] Zalis, M.G., Wilson, C.M., Zhang, Y. and Wirth, D.F. (1993) Characterization of the Pfmdr2 gene for Plasmodium falciparum. Mol. Biochem. Parasitol. 62, 83-92.

[I I] Wilson, C.M., Volkman, S.K., Thaithong, S., Martin, R.K., Kyle, D.E., Milhous, W.K. and Wirth, D.F. (1993) Amplification of pfmdrf associated with mefloquine and halofantrine resistance in Plasmodium falciparum from Thailand. Mol. B&hem. Parasitol 57, 151-160.

[12) Cowman, A.F., Galatis, D. and Thompson, J.K. (1994). Selection for mefloquine resistance in Plasmodium

falciparum is linked to amplification of the pfmdrl gene and cross-resistance to halofantrine and quinine. Proc. Natl. Acad. Sci. USA 91. 1143-1147.

[ 131 Barnes, D.A., Foote, S.J., Galatis, D., Kemp, D.J. and Cowman, A.F. (1992) Selection for high-level chloro- quine resistance results in deamplification of the pfmdrl gene and increased sensitivity to mefloquine in Plasmodium falciparam. EMBO J. I 1, 3067-3075.

[14] Peel, S.A., Bright, P., Yount, B., Handy, J. and Baric, R.S. (1994). A strong association between mefloquine and halofantrine resistance and amplification, overex- pression. and mutation in the p-glycoprotein gene homolog (pfmdr) of Plasmodium fakiparum in vitro. Am. J. Trop. Hyg. 51, 648-658.

[IS] Foote, S.J., Kyle, D.E., Martin, R.K., Oduola, A.M.J., Forsyth, K., Kemp, D.J. and Cowman, A.F. (1990) Several alleles of the multidrug-resistance gene are close- ly linked to chloroquine resistance in Pfasmodium falciparum. Nature 345, 255-258.

1161 Awad-El-Kariem, F.M.. Miles, M.A. and Warhurst, D.C. (1992) Chloroquine resistant Plasmodium

1171

[ISI

1191

[201

v-11

I221

1231

1241

I251

1261

1271

falciparum isolates from the Sudan lack two mutations in the pfmdrl gene thought to be associated with chloro- quine resistance. Trans. R. Sot. Trop. Med. Hyg. 86, 587-589. Rubio, J.P. and Cowman, A.F. (1994). Plasmodium falciparum: The pfmdrl protein is not overexpressed in chloroquine-resistant isolates of the malaria parasite. Exp. Parasitol. 79, 137-147. Le Bras, J., Deloron, P., Ricour, A., Andrieu, B., Save], J. and Coulaud, J.P. (1983) Plasmodium falciparum: Drug sensitivity in vitro of isolates before and after adaptation to continuous culture. Exp. Parasitol. 56, 9-14. Pologe, L.G. and Ravetch, J.V. (1986). A chromosome rearrangement in Plasmodium falciparum histidine-rich protein gene is associated with the knobless phenotype. Nature 322, 474-477. Le Bras, J. and Deloron, P. (1983) In vitro study of drug sensitivity of Plasmodium falciparum: An evaluation of a new semi-microtest. Am. J. Trop. Med. Hyg. 32, 447-451. Le Bras, J., Rabenjarso, E., Ringwald, P. and Doury, J.C. (1991). Evolution de Ia chimioresistance du paludisme importe en France en 1990. Med. Hyg. 49, 2314-2320. Bouchad, O., Basco, L.K., Gillotin, C., Gimenez, F., Ramiliarsoa, O., Genissel, B., Bouvet, E., Farinotti, R., Le Bras, J. and Coulad, J.P. (1994). Clinical efficacy an pharmacokinetics of micronized halofantrine for the treatment of acute uncomplicated falciparum malaria in nonimmune patients. Am. J. Trop. Med. Hyg. 51, 204-213. Wooden, J., Gould, E.E., Paull, A.T. and Sibley, C.H. (1992) Plasmodium falciparum: A simple polymerase chain reaction method for differentiating strains. Exp. Parasitol. 75, 207-212. Creasey, A., Fenton, B., Walker, A., Thaithong, S., Oliveira. S., Mutambu, S. and Walliker, D. (1990) Gene- tic diversity of Plasmodium falciparum shows geographical variation. Am. J. Trop. Med. Hyg. 42, 403-413. Conway, D.J., Greenwood, B.M. and McBride, J.S. (1991) The epidemiology of multiple-clone Plasmodium fateiparum infections in Gambian patients. Parasitology 103, l-6. Babiker, H.A., Creasey, A.M., Bayoumi, R.A.L., Walliker, D. and Arnot, D.E. (1991) Genetic diversity of Plasmodium falciparum in a village in eastern Sudan. 2. Drug resistance, molecular karyotypes and the mdrl genotype of recent isolates. Trans. R. Sot. Trop. Med. Hyg. 85, 578-583. Bayoumi, R.A.L., Creasey, A.M., Babiker, H.A., Cariton, J.M.R., Sultan, A.A., Satti, G., Sohal, A.K., Walliker, D., Jensen, J.B. and Amot, D.E. (1993) Drug response and genetic characterization of Plasmodium falciparum clones recently isolated from a Sudanese village. Trans. R. Sot. Trop. Med. Hyg. 87, 454-458.

166 L.K. Basco et al. /Molecular and Biochemical Parasitology 74 (1995) 157- I66

[28] Wellems, T.E., Panton, L.J., Gluzman, I.Y., Rosario, V. E., Gwada, R.W., Walker-Jonah, A. and Krogstad, D.J. (1990) Chloroquine resistance not linked to mdr-like genes in Plasmodium fakiparum cross. Nature 345, 253-255.

[29] Wellems, T.E., Walker-Jonah, A. and Panton, L.J. (1991) Genetic mapping of the chloroquine-resistance

locus on Plasmodium falciparum chromosome 7. Proc. Natl. Acad. Sci. USA 88, 3382-3386,

[30] Oduola, A.M.J., Omitowoju, G.O., Gerena, L., Kyle, D.E., Milhous, W.K., Sowunmi, A. and Salako, L.A. (1993) Reversal of mefloquine resistance with penfluridol in isolates of Plasmodium falciponun from south-west Nigeria. Trans. R. Sot. Trop. Med. Hyg. 87, 81-83.