Tropical Medicine and International Health doi:10.1111/j.1365-3156.2007.01991.

volume 13 no 2 pp 199–207 february 2008

Plasmodium falciparum from Pará state (Brazil) shows

satisfactory in vitro response to artemisinin derivatives and
absence of the S769N mutation in the SERCA-type PfATPase6
Isabel D. Ferreira1, Axel Martinelli1, Louise A. Rodrigues1, Ediclei L. do Carmo2, Virgı́lio E. do Rosário1,
Marinete M. Póvoa2 and Pedro Cravo1

1 Centro de Malária e Outras Doenças Tropicais ⁄ IHMT ⁄ UEI Biologia Molecular ⁄ UEI Malária, Lisbon, Portugal
2 Secretaria de Vigilância em Saúde, Instituto Evandro Chagas, Belém, Pará, Brazil

Summary objective To evaluate the in vitro efficacy of artesunate (ATN) and artemether (ATH) against
Plasmodium falciparum isolates from the Brazilian Amazon state of Pará and to search for mutations
and ⁄ or altered copy numbers in the putative resistance-associated pfcrt, pfmdr1 and pfATPase6 genes.
methods In vitro efficacy of ATN and ATH was successfully measured in 56 freshly collected
P. falciparum isolates, using a conventional WHO microtest with minor modifications. Single nucleotide
polymorphisms (SNPs) in the same isolates were inspected using DNA sequencing and ⁄ or PCR-RFLP.
We used real-time quantitative PCR to assess gene copy numbers.
results ATN and ATH geometric mean IC50s were 0.85 nm, 95% CI (0.55–1.15) and 3.0 nm, 95% CI
(1.5–4.5), respectively. There was extremely limited diversity of pfcrt and pfmdr1 genotypes and three
SNPs were identified in the pfATPase6 gene: one T to A synonymous mutation at nucleotide 2694 and
two non-synonymous (both G to A) mutations at nucleotides 110 and 1916, causing predicted ami-
noacid shifts of arginine to lysine and of glycine to aspartate, respectively. The previously reported
S769N mutation was not detected in any of the isolates inspected. In addition, no gene amplifications
were detected in a subset of eight isolates.
conclusion Artemisinin derivatives display satisfactory in vitro activity locally and the pfATPase6
gene is distinct from that reported in French Guiana, suggesting that those haplotypes have not been
introduced regionally.

keywords malaria, Plasmodium falciparum, artemisinin derivatives, molecular markers, Brazil

(Ministério da Saúde do Brasil 2006). In Pará, where this

Introduction
Malaria causes 300–600 million cases worldwide annually
of which 0.5–2.5 million are fatal. In sub-Saharan Africa,
where Plasmodium falciparum is the predominant species,
malaria accounts for at least 18.0% of all infantile deaths
(Rowe et al. 2006). In South America the number of
malaria infections and related mortality is lower than in
Africa, in part because the milder species of malaria
Plasmodium vivax predominates over P. falciparum. In the
Americas and in the Caribbean, 38% of the population in
21 countries live in areas of malaria risk, where 1.3 m cases
per year are reported (Coura et al. 2006). Thirty-six per
cent of these occur in Brazil (Coura et al. 2006), where
99.5% of all malarias occur within the so-called ‘Legal
Amazonic Region’ that includes all states of the North
Region of the country plus Mato Grosso and Maranhão
study was carried out, 122 442 cases were notified in 2005,
making it the second most malaria-affected state in Brazil
after Amazonas. The most frequent infecting species is
P. vivax (about 80% of the cases) but the incidence of
P. falciparum (the most virulent form of the parasite) has
been increasing (Ministério da Saúde do Brasil 2006),
raising the concern of Brazilian’s public health entities.
As in many other areas of the world, local malaria
control heavily relies on preventive or case management
chemotherapy, but increasing failure of the predominantly
used drugs has presented a serious obstacle towards
effective disease control. Resistance of P. falciparum has
been described in the amazonic region to mefloquine
(Calvosa et al. 2001), to quinine (Zalis et al. 1998), and
even to the quinine plus doxycycline combination (Leal
et al. 2003), the first-line treatment option recommended

ª 2008 Blackwell Publishing Ltd 199

Tropical Medicine and International Health
I. D. Ferreira et al. In vitro susceptibility and molecular typing of Plasmodium falciparum from Brazil
by the Brazilian National Antimalarials Policy until very
recently. Resistance to other drugs commonly used against
P. falciparum, such as chloroquine (Vieira et al. 2004) and
sulphadoxine–pyrimethamine (Cortese et al. 2002) has
been widespread in the Brazilian Amazonic region for
many years.
Reported chemoresistance has been further substanti-
ated by molecular analysis of malaria drug resistance
markers. For instance, the chloroquine-resistance-associ-
ated mutation pfcrtK76T has been shown to have
reached fixation in natural parasite populations of
Brazilian P. falciparum (Vieira et al. 2001, 2004), while
sulphadoxine–pyrimethamine-conferring mutations in the
pfdhfr and pfdhps genes have also been confirmed to be
highly frequent (Vasconcelos et al. 2000).
To counteract resistance, therapeutic approaches for
P. falciparum are being shifted towards the use of
Artemisinin-based Combination Therapy (ACT) nation-
wide. As the first reports regarding ACT efficacy trials
begin to emerge the results are encouraging. To this
purpose, Alecrim et al. (2006) have recently reported high
efficacy of the combination artemether + lumefantrine,
with the added benefit of this being a safe and well
tolerated treatment.
As the new treatment approaches show great promise, it
is essential to warrant their continued long-term efficacy.
Regular surveillance of in vitro responses to artemisinin
derivatives coupled with molecular examination of puta-
tive resistance modulators in local P. falciparum isolates
offers an exceptional means for timely detection of
potential resistance foci.
The mechanisms involved in mediating resistance to
artemisinin derivatives are not clearly understood. Point
mutations in the Pfcrt gene, originally investigated in the
context of chloroquine resistance, marginally influence the
response to artemisinin in a genetic transfection experi-
ment (Sidhu et al. 2002). However, their importance in
natural parasites populations has not been confirmed
(Jambou et al. 2005). An additional gene that has been
examined within the context of this phenotype is the multi-
drug resistance gene 1 (pfmdr1). There is evidence
indicating that natural parasite populations of P. falcipa-
rum harbouring multiple copies of pfmdr1 are more likely
to exhibit decreased susceptibility to artemisinin deriva-
tives (as well as to mefloquine) (Price et al. 2004). Genetic
disruption of one of the two pfmdr1 copies in the drug-
resistant FCB line resulted in increased susceptibility to
artemisinin in vitro, substantiating the view that the gene’s
copy number is able to modulate this phenotype (Sidhu
et al. 2006).
Jambou et al. (2005) alarmingly reported a significant
decrease in in vitro sensitivity to artemether (ATH) in
200
volume 13 no 2 pp 199–207 february 2008
P. falciparum isolates from French Guiana, located along
Brazil’s northern border. This reduced efficacy was
significantly associated with a mutation encoding a
S769N shift in the PfATPase6 protein (Jambou et al.
2005), an SERCA-type Ca2 + ATPase, believed to be the
primary target for artemisinins (Eckstein-Ludwig et al.
2003). Although the functional and causal significance of
this mutation have not yet been elucidated, these obser-
vations could indicate the initial stages of resistance
development. Critically, French Guyana’s physical prox-
imity to Brazil makes it possible for the potentially
resistant parasites to be carried across the border and
spread through local treatment-based selection.
This study reports results from a survey carried out with
P. falciparum isolates collected in Tucuruı́, a town located
in the Brazilian state of Pará. The in vitro sensitivity of
those samples to artesunate (ATN) and ATH was assessed
and possible mutations or alterations in copy numbers
were sought in the pfcrt, pfmdr1 and pfATPase6 genes.
Materials and methods
Study site and sample collection
Tucuruı́ is located in the Southeast of the Pará State
(Figure 1). Tucuruı́ is home to the fourth largest hydro-
electric dam in the world. Construction of this dam in 1981
caused enormous environmental changes, deforestation
and human migration to the region and elevated malaria to
the epidemic level (Vasconcelos et al. 2000). The town has
a permanent population of 81 000 inhabitants. The climate
is typically tropical, with an average temperature of 24 C
and relative humidity >85%. The average annual rainfall
exceeds 2500 mm; there is one rainy season from Decem-
ber to May and one dry season from June to November.
The Annual Parasite Index in the Tucuruı́ Council is
between 10 and 49.9, classifying the risk of malaria
transmission in this region as medium (Ministério da Saúde
do Brasil 2006).
For this work blood samples from suspected malaria
patients were collected at Posto Liler Leão and Posto de
Saúde GETAT health facilities in Tucuruı́, during
September 2005. The study received ethical approval from
the Evandro Chagas Institute Ethical Committee.
After confirming P. falciparum monoinfection through
inspection of thin and thick smears, 5 ml of venous blood
was collected into EDTA-containing Monovettes; 500 ll of
each sample was spotted onto WhatmanTM no. 4 filter
paper and used for subsequent DNA extraction and
parasite genotyping. The remaining blood was used for
P. falciparum in vitro susceptibility testing. The tests were
performed at the Hospital Regional de Tucuruı́.
ª 2008 Blackwell Publishing Ltd
Tropical Medicine and International Health
I. D. Ferreira et al. In vitro susceptibility and molecular typing of Plasmodium falciparum from Brazil
RR AP
PA
AM
AC
TO
RO
MT
GO
DF
MS
SP
PR
Figure 1 Location of Tucuruı́, where SC
samples used in this study were originally RS
collected.
In vitro activity of artemether and artesunate
Venous blood of each patient was used to determine the
in vitro susceptibility of parasites to ATN and ATH as
described elsewhere in detail (Ferreira et al. 2007). Briefly,
Giemsa-stained blood smears were analysed by optical
microscopy and percentage parasitaemia was determined.
Subsequently, blood samples were washed twice in
RPMI1640 medium and supernatant and white blood cells
were discarded after each wash. The red blood cells were
resuspended in 1:1 volume RPMI1640 medium and
percentage P. falciparum parasitaemia was normalized to
0.3–0.8%, through the addition of uninfected erythrocytes.
Drug sensitivity analysis was performed by a modified
version of the standard WHO MarkIII micro-test (WHO
1997). Ten microlitres of a parasite suspension at 0.3–
0.8% parasitemia and 5% haematocrit were added to
90 ll of medium in the absence or presence of different
concentration of ATN (range 0.05–36.5 nm) or ATH
(range 0.2–146 nm) and incubated in 96-well tissue culture
plates in duplicates for each sample and drug, for 24–30 h,
in a 5% CO2 atmosphere. Thick smears were made of the
blood from each well, stained with Giemsa and examined
under microscope at 100 · magnification.
The number of schizonts with three or more nuclei was
counted out of 200 asexual parasites and the drug
concentration causing 50% inhibition of schizont forma-
tion compared with a drug-free well (IC50) was determined
by linear regression. To assess possible correlations
between the isolate’s response to the two drugs, IC50 values
of all tested samples for each drug were organized
ª 2008 Blackwell Publishing Ltd
volume 13 no 2 pp 199–207 february 2008
FRENCH
GUIANA SURINAME GUIANA
AMAPÁ
TUCURUÍ
MA CE RN
PB
PI PE
AL
SE
PARÁ
BA
MG
ES
TOCANTINS
RJ MATO GROSSO
according to increasing concentrations, paired between the
two drugs and then subjected to Spearman rank order
correlation testing (Wessa 2007). A correlation was con-
sidered to be significant if the value of P is <0.05.
Genotyping: detection of single nucleotide polymorphisms
Genomic DNA of the parasites was first extracted from
spotted blood by boiling in Chelex-100 (Plowe et al. 1995).
The same procedure was followed for P. falciparum Dd2,
Hb3 and 3D7 laboratory strains when those were used as
controls.
PfATPase6
For DNA sequencing, several PCR fragments encompass-
ing a partial sequence of the pfATPase6 gene spanning
nucleotides 28 to 3078 were amplified by PCR and
inspected by DNA sequencing in a subset of eight isolates
selected according to their response to ATH (i.e. the four
samples with the lowest ATH IC50 and the four with the
highest ATH IC50 were picked). The quality and concen-
tration of PCR products generated for each fragment were
ascertained by agarose gel electrophoresis and used as
template for DNA sequencing following purification with a
commercial kit (Qiagen QIAquick PCR Purification Kit).
DNA sequences were generated from both sense and anti-
sense primers and aligned to check for genetic polymor-
phisms, after resolving potential ambiguities. PCR primers
and reaction conditions have been published by Ferreira
et al. (2007).
201
Tropical Medicine and International Health volume 13 no 2 pp 199–207 february 2008

I. D. Ferreira et al. In vitro susceptibility and molecular typing of Plasmodium falciparum from Brazil

mutations N75E and K76T were screened through muta-

PCR-RFLP
Four single nucleotide polymorphisms (SNPs) in the
pfATPase6 gene were determined in all samples using
PCR-RFLP (Table 1). These were: one 2694 T to A
synonymous polymorphism previously reported in African
parasites (Ferreira et al. 2007), the non-synonymous
2306 G to A transition described in French Guiana
(encoding S769N), and two other non-synonymous (both
G to A transitions) mutations at nucleotides 110 and 1916,
uncovered in the course of the present work by DNA
sequencing as described in the previous section (encoding
R37K and G639D, respectively).
Of the four SNPs inspected by PCR-RFLP, only the
G110A and the T2694A created natural endonuclease
restriction sites. In the case of the remaining two SNPs,
primers were designed using PIRA-PCR software available
online (Ke et al. 2001), to contain an artificially introduced
mismatch that generated a restriction site for one of the
two alternative alleles (Table 1). Restriction endonucleases
were acquired from FermentasTM and restriction assays
were conducted according to the supplier’s specifications.
Classification of each sample was made after inspection of
its respective restriction pattern as depicted in Table 1, on
a 3% agarose gel stained with ethidium bromide.
Pfmdr1 and pfcrt
The pfmdr1 genetic polymorphisms N86Y, D1042N and
D1246Y were inspected in all isolates using PCR-RFLP
protocols (Ferreira et al. 2007). As for the pfcrt gene the
tion-specific PCR and PCR-RFLP, respectively (Alves et al.
2006; Ferreira et al. 2007).
Estimation of pfmdr1 and pfATPase6 copy number
DNA from the same subset of eight samples selected as
described in a previous section was analysed by real-time
quantitative PCR in order to assess the copy number of
Pfmdr1 and PfATPase6 genes. Genomic DNA was incu-
bated under conditions described elsewhere (Ferreira et al.
2006) and the copy number was estimated using the 2–
DDct method of relative quantification (Livak & Schmitt-
gen 2001). Briefly, the clone 3D7 was used as calibrator for
both genes (this strain is known to carry only a single copy
of each of those genes) and Pfßactin was used as the
housekeeping gene to normalize data. The P. falciparum
Dd2 laboratory strain, known to harbour four copies of the
pfmdr1 gene, was used as a control (Cowman et al. 1994).
The experiments were repeated at least three times for each
sample and gene.
Results
Drug response
Sixty-six samples were collected and freshly tested in
vitro on site. Successful in vitro drug test results for both
ATH and ATN were obtained for 56 samples (85%).
Failure to get test results for particular samples was
mainly attributed to poor in vitro growth, undesirable

Table 1 Information on PCR-RFLP assays to detect single nucleotide polymorphisms in the PfATPase6 gene

Target Change Coordinates Expected fragment

Primer pair SNP (aa) (amplicon size) Enz. sizes after restriction

PfATP6-110F CGTTGAACTTATTATATCTTTGTC G110A R37K )76 to 255 MbOII G 103 A 143

PfATP6-110R TTTCATATCTAATAAAGTTAACACG (331) 94 94
92 92
40*
PfATP6-1916F G1916A G639D 1889 to 2055 PagI G 166 A 144
TATAGGAGAAAATACATTTGAtCATG (166) 22*
PfATP6-1916R
ACATTCATTTCTCCAAGAAGAA
PfATP6-769S-F ACTTAGCTTTGCTTATAAAAAcTTAA G2306A S769N 2279 to 2551 BsPTI G 250 A 272
PfATP6-769S-R AATTATCCTTTTCATCATCTCC (272) 22*
PfATP6-2694F A2694T I898I 2650 to 2832 TasI A 181 T 142
GAATTGTTTTCTGTAGAACTGAAC (182) 1* 39*
PfATP6-2694R 1*
ATCTGATGCTTCTTTAGCTACC

Small-capped underlined nucleotides indicate mismatches introduced in the primers to artificially create restriction sites. Enz., enzyme.
*PCR band sizes which are likely to go undetected on agarose gels due to small size.

202 ª 2008 Blackwell Publishing Ltd

Tropical Medicine and International Health volume 13 no 2 pp 199–207 february 2008

I. D. Ferreira et al. In vitro susceptibility and molecular typing of Plasmodium falciparum from Brazil

25 110 (37) 1916 (639)

G G G A G G M M G G G G m G
IC50 artemether (nM)

20

15

10

5
2306 (769) 2694 (898)
0 M M
A* G G G G G A* M T A A m
0 1 2 3 4 5 6
IC50 artesunate (nM)

Figure 2 Correlation between the susceptibility patterns of

artemether and artesunate among all tested samples (R = 0.90;
P < 0.001).

contaminations or to complications in downstream
processing of samples after collection. Drug test results
indicated that ATN and ATH geometric mean IC50s
were 0.85 nm, 95% CI (0.55–1.15) and 3.0 nm, 95% CI
(1.5–4.5), respectively. There was a perceptible mono-
tonic relation in increased sensitivity between the two
drugs (Figure 2), which was confirmed by statistical
analysis (P < 0.001).
Figure 3 Agarose gels (2.5%) depicting typical PCR-RFLP results
for each SNP inspected in the PfATPase6 gene [G, A, T: nucleo-
tides located at the polymorphic site with their corresponding gel
banding patterns; m: mixture between two genotypes; M:
O’RangeRuler 100 bp DNA ladder (FermentasTM); *undigested
PCR product is used to illustrate the behaviour of a mutant
genotype, as control samples of mutant genotype were unavailable;
note: numbers represent nucleotide coordinates and numbers in
brackets, the corresponding residues].

previously published mutation encoding an S to N residue

Genotyping
Previously reported mutations were sought among the
pfmdr1 and pfcrt genes in all isolates included in the
study. Genotyping was successfully achieved for all 56
isolates for which in vitro data had been generated. Results
highlighted extreme predominance of given genotypes as
illustrated in Table 2. Indeed, with the exception of residue
1042 of the pfmdr1 gene, alleles pfcrt 75N, pfcrt 76T,
pfmdr1 86N and pfmdr1 1246Y appear to be fixed among
the local parasite population (Table 2).
In the case of the pfATPase6 gene, limited information
has been published on allelic variation. Thus only two
previously reported mutations were targeted a priori: one
synonymous T to A SNP uncovered in African parasites,
reported recently by our group (Ferreira et al. 2007) and a
change at codon 769 (Jambou et al. 2005). For each case, a
PCR-RFLP was designed as described in the Materials and
methods section, which produced the expected results in
the optimization phase and was subsequently successfully
applied to all 56 samples under scrutiny. Figure 3 presents
typical results and more detailed information on these
genotyping assays.
To our knowledge there was no published information
on DNA sequence of the pfATPase6 gene in Brazilian
isolates of Plasmodium falciparum prior to the present
study and for that reason, eight samples were chosen for
which the majority of the gene’s sequence was obtained.
Selection of these samples was done based on their
respective in vitro response to ATH as described previ-
ously (see Materials and methods section). Resulting

Table 2 Prevalence and number of isolates (N) harbouring mutations within the pfcrt and pfmdr1 genes (m indicates samples containing
mixed genotypes)

Gene Pfcrt Pfmdr1

Allele 75N 75E m 76K 76T m 86N 86Y m 1042N 1042D m 1246D 1246Y m

Prevalence (%) 100 0 0 0 100 0 100 0 0 89 9 2 0 100 0

N 56 0 0 0 56 0 56 0 0 50 5 1 0 56 0

ª 2008 Blackwell Publishing Ltd 203

Tropical Medicine and International Health
I. D. Ferreira et al. In vitro susceptibility and molecular typing of Plasmodium falciparum from Brazil
overlapping sequencing fragments were used to compile a
pfATPase6 consensus sequence for each sample (data not
shown) and all sequences thus generated were compared
between them and with that of the published reference
strain P. falciparum 3D7 to search for potential SNPs.
This allowed to uncover two further novel non-synony-
mous mutations, G110A and G1916A, encoding R37K
and G639D, respectively, as depicted in Table 1. Because
we wished to investigate the prevalence of these muta-
tions in all isolates available and check whether any of
them could correlate with differential drug responses,
PCR-RFLP assays were also developed in these cases
(Table 1) which were then applied to the remaining
samples. These PCR-RFLP assays were successfully
implemented and typical results for each of them can be
visualized in Figure 3.
Successful screening of all 56 samples for the above
SNPs data revealed major predominance of parasites
harbouring wild-type alleles in all cases (Figure 4). The
G2306A mutation (encoding S769N), previously reported
in French Guiana (Jambou et al. 2005), was absent in our
samples. The results are detailed in terms of allelic
prevalence in Figure 4. Interestingly there were two
unique nucleotide combinations at polymorphic positions
110, 1919 and 2694, reflecting two alternative haplo-
types: GGA (wild-type) or AAT (mutant), whose preva-
lence was 88% and 12%, respectively. There was no
correlation between the above haplotypes and the degree
of in vitro response to either drug included in this study
(data not shown).
100
89
85
100
80
Prevalence (%)
60
40
11
10
20 0 5 0 0
0
G A m G A m G A m
110 1916 2306
(R37K) (G639D) (S769N)
Figure 4 Prevalence of the SNPs inspected in the PfATPase6 gene
among all samples analysed (G, A, T: polymorphic nucleotides
which may be present at positions 110, 1916, 2306 or 2694; in
brackets is the predicted translation of the mutations encoded by
each allele).
204
volume 13 no 2 pp 199–207 february 2008
Assessment of gene copy numbers
We estimated gene copy numbers of pfmdr1 and pfAT-
Pase6 in the subsample of eight isolates, of which four
presented the lowest calculated ATH IC50s whilst the
remaining four presented the highest. Genomic DNA of the
well characterized clones P. falciparum 3D7 and of Dd2,
respectively, known to harbour one and four copies of the
pfmdr1 gene, were used as controls in all experiments.
Each sample was analysed on three independent occasions
and final results were expressed as mean copy number of
the three experiments. This inspection revealed that all
parasites analysed harboured a single copy of the two genes
studied, independently of their respective response to the
drug (Figure 5).
Discussion
Drug surveillance surveys assume special relevance in the
Amazon, an area which has traditionally been prone to the
emergence of drug-resistant forms of P. falciparum despite
this region probably sharing less than 5% of the world’s
parasite biomass. Thus, for example, two of the four
independent chloroquine resistance founder events have
been clearly positioned to South America (Wootton et al.
2002), while the first published evidence of decreased
in vitro efficacy of an artemisinin derivative has suggested
the existence of P. falciparum with relatively high ATH
IC50s in French Guiana already (Jambou et al. 2005).
Throughout most of Brazil’s malaria endemic regions,
increasing failure rates of quinine + doxycycline, the
former first option for the treatment of P. falciparum, are
forcing its replacement with artemisinin combination
treatment (artemether + lumefantrine). Collectively, the
83
4 Pfmdr1
PfATPase6
Mean copy number
12 2
5
1
A T m
2694
(I898I) 0
Dd2 0.2 0.3 0.5 0.5 22.1 22.5 23.0 23.2
Artemether IC50 (nM)
Figure 5 Mean copy number of the pfmdr1 and PfATPase6
genes in P. falciparum Dd2 (control) and eight samples chosen
according to their artemether IC50 value.
ª 2008 Blackwell Publishing Ltd
Tropical Medicine and International Health
I. D. Ferreira et al. In vitro susceptibility and molecular typing of Plasmodium falciparum from Brazil
above information underscores the need for rigorous drug
supervision in this area of the world.
Our main aims were to assess the baseline in vitro
activity of the artemisinin derivatives ATH and ATN
against fresh P. falciparum isolates from Pará, and to
evaluate their genotype regarding putative drug response
modulators. To our knowledge this was the first study to
address both issues in Brazil.
Collectively, the overall activity of both drugs appeared
to be satisfactory although four samples presented ATH
IC50s above 20 nm. There is, however, no easy approach
towards correct interpretation of the real significance of
these results, as the values which may provide an unam-
biguous insight of resistance thresholds have not been
ascertained in the case of artemisinin derivatives. In general
terms, the data obtained for the Brazilian isolates appear to
be within the trends of other endemic areas and are
consistent with the notion that, although natural parasite
populations of P. falciparum vary significantly with respect
to their susceptibility to this class of drugs, both intra- and
inter-regionally, there have not been any studies reporting
in vitro responses indicative of true resistance. Interest-
ingly, the data from Pará suggest that the geometric mean
IC50 values observed in the present study appear to be
more similar to those observed in Africa than to those
reported for south-East Asia, where chemotolerance
appears to be higher. As an example, a recent study from
Laos has reported a geometric mean IC50 to ATN of
5.0 nm among local P. falciparum populations (Mayxay
et al. 2007), a value which is more than fivefold higher
than the one observed in our report (0.85 nm). In contrast,
African parasites appear to be more susceptible to arte-
misinin derivatives with previously reported geometric
mean IC50s to ATN of 0.58, 0.73 and 2.2 nm from São
Tomé, Gabon and Senegal, respectively (Henry et al. 2006;
Ramharter et al. 2006; Ferreira et al. 2007).
One exception to a worldwide tendency for adequate
in vitro efficacy of artemisinin derivatives is the previous
report by Jambou et al. (2005), who described several
P. falciparum fresh isolates from French Guiana with IC50
levels to ATH above 40 nm, which indicated a trend
towards significantly low tolerance. The molecular analysis
of this parasite population revealed also that with a single
exception, six of those samples harboured a particular
mutation in the pfATPase6 gene enconding S769N. Thus
in summary that report appeared to suggest that artemis-
inin-tolerant mutants may already be present in the
Brazilian Amazon region. It is also a well-documented fact
that there exists a significant uncontrolled rate of human
migration between borders separating the different coun-
tries constituting the Amazonic region and several states
within the legal Brazilian Amazon share their borders with
ª 2008 Blackwell Publishing Ltd
volume 13 no 2 pp 199–207 february 2008
the other malaria-endemic nations. A significant share of
this migration is constituted by non-immune miners and
loggers who travel between the different countries and
states in search of new work opportunities. These are
extremely efficient pathogen carriers contributing to the
potential migration of the parasite itself and to the
consequent expansion of prospective drug-resistant
mutants across the region.
In light of the above perspective, our work indicates that
at least at the time of this survey, the parasites encountered
in Pará are likely to be different from those described in
French Guiana for two main reasons: first, the in vitro
phenotypic behaviour appears to differ markedly between
the two populations such that the overall ATH IC50s are
significantly lower in the case of the Brazilian samples;
these figures may be due to differences in artemisinin use
and therefore, its pressure in the two countries. Second, the
molecular profile between the two samples appears to be
distinct, as illustrated, for instance, by the observation that
no pfATPase6 S769N mutant could be detected among 56
isolates successfully analysed. However, we acknowledge
that a comprehensive comparison of the population struc-
ture between the Brazilian and French Guiana samples
would require additional work which was not contem-
plated in this study, such as extensive genotyping, using for
instance microsatellite data.
Three SNPs in the PfATPase6 gene were disclosed
among the different samples, two of which are being
described here for the first time to our knowledge, and
cause the predicted aminoacid shifts R37K and G639D.
Collectively the three SNPs form two distinctive haplotypes
each encoding its unique predicted protein. The R37K
mutation is predicted to lie in the conserved N-terminal
region of the protein, which may be important for
regulating its function, whilst the G639D is expected to be
positioned in the cytosolic region located between trans-
membrane domains 5 and 6. However, the biological
significance of the above mutations is difficult to forecast in
the absence of functional assays.
There are some limitations in attempting to infer how
the above pfATPase6 genotypes compare with others
originated from different regions as little published
information is available as yet. Our group has previously
carried a similar survey in fresh samples from West Africa
and the African parasite displayed only one single
synonymous SNP (Ferreira et al. 2007). This appears to
be consistent with the recent findings by Mugittu et al.
(2006) who have also assessed PfATPase6 SNPs in
Tanzania (East Africa) and reported that no mutation
was disclosed, although in this case particular
polymorphisms were targeted a priori and hence no
de novo sequencing took place.
205
Tropical Medicine and International Health
I. D. Ferreira et al. In vitro susceptibility and molecular typing of Plasmodium falciparum from Brazil
Other genetic markers were contemplated in this study
due to previous evidence suggesting they may be able to
modulate responses to multiple drugs, including arte-
misinin derivatives. In these cases however, the data
obtained is more difficult to frame and interpret in
relation to artemisinin responses due to natural geo-
graphical clustering of particular genotypes and ⁄ or pre-
vious selective pressure exerted by other drugs such as
chloroquine, quinine and mefloquine. To this purpose
our observations are in line with previous work dem-
onstrating strong selection of pfcrt mutants, most likely
by chloroquine (Vieira et al. 2001), and large predom-
inance of specific pfmdr1 genotypes in the Brazilian
Amazon (Zalis et al. 1998). Previous reports have also
described an association between increased copy number
of the pfmdr1 gene and mefloquine resistance, as well as
a tendency for increased tolerance to artemisinin deriv-
atives both in vitro and in vivo (Price et al. 2004; Alker
et al. 2007). A recent allelic knock-out study has
provided direct evidence towards the ability of pfmdr1
copy numbers in modulating the efficacy of both drugs
against the human malaria parasite (Sidhu et al. 2006).
Interestingly, we did not detect any occurrence of
pfmdr1 gene amplification despite extensive use of
mefloquine and quinine in this region in the past and
subsequent reports of resistance to these drugs (Zalis
et al. 1998; Calvosa et al. 2001). The lack of parasites
harbouring multiple copies of the gene may thus reflect a
distinct mechanism of resistance in this population.
Alternatively, the gene may be overexpressed at the
mRNA and protein levels in some parasites via molec-
ular mechanisms other than gene amplification; this
should perhaps be contemplated in future studies. We
acknowledge that these observations may reflect a
stochastic effect due to reduced sample size (eight
isolates tested).
In summary, we describe the examination of a natural
parasite population of P. falciparum from Brazil whose
response to artemisinin derivatives and respective geno-
typic profile were evaluated at a time when these drugs
were starting to be used as first-line treatment in this
country. Hence, the information reported here represents
a valuable set of data which can be used as a
comparative baseline record for future analogous studies.
We suggest that regular phenotypic as well as molecu-
lar surveillance of artemisinin derivatives constitute
desirable initiatives especially at a time when these
compounds are being introduced at a massive scale
worldwide in the form of combination therapy. Such
trials will facilitate continuous evaluation and updating
(if required) of current ACT treatment policies in this
area of the world.
206
volume 13 no 2 pp 199–207 february 2008
Acknowledgements
This study is part of the research conducted by RES-
MALCHIP, a European Research Consortium. Virgı́lio E.
do Rosário and Pedro Cravo were funded by the Instituto
de Higiene e Medicina Tropical ⁄ UNL, Portugal; Axel
Martinelli by FCT, MCES ⁄ FEDER; Isabel D. Ferreira by
RESMALCHIP (contract QLK2-CT-2002-01503). We
thank the staff of Posto Liler Leão and Posto GETAT,
Tucuruı́, José Maria de Souza Nascimento e José Mário
Veloso Peres, and all the patients who agreed to participate
in this study.
References
Alecrim MG, de Lacerda MV, Mourao MP et al. (2006) First
Brazilian experience with the use of artemether-lumefantrine
(Coartem) a fixed-dosed ACT combination. American Journal
of Tropical Medicine and Hygiene 75, 187.
Alker AP, Lim P, Sem R et al. (2007) Pfmdr1 and in vivo resistance
to artesunate-mefloquine in falciparum malaria on the Cambo-
dian-Thai border. American Journal of Tropical Medicine and
Hygiene 76, 641–647.
Alves J, Roque AL, Cravo P et al. (2006) Epidemiological char-
acterization of Plasmodium falciparum in the Republic of Cabo
Verde: implications for potential large-scale re-emergence of
malaria. Malaria Journal 5, 32.
Calvosa VS, Adagu IS & Povoa MM (2001) Plasmodium falci-
parum: emerging mefloquine resistance in vitro in Para State,
north Brazil. Transactions of the Royal Society of Tropical
Medicine and Hygiene 95, 330–331.
Cortese JF, Caraballo A, Contreras CE & Plowe CV (2002) Origin
and dissemination of Plasmodium falciparum drug-resistance
mutations in South America. Journal of Infectious Diseases 186,
999–1006.
Coura JR, Suarez-Mutis M & Ladeia-Andrade S (2006) A new
challenge for malaria control in Brazil: asymptomatic Plasmo-
dium infection – a review. Memórias do Instituto Oswaldo Cruz
101, 229–237.
Cowman AF, Galatis D & Thompsom JK (1994) Selection for
mefloquine resistance in Plasmodium falciparum is linked to
amplification of the pfmdr1 gene and cross-resistance to halo-
fantrine and quinine. Proceedings of the National Academy of
Sciences USA 91, 1143–1147.
Eckstein-Ludwig U, Webb RJ, Van Goethem ID et al. (2003)
Artemisinins target the SERCA of Plasmodium falciparum.
Nature 424, 957–961.
Ferreira ID, do Rosário VE & Cravo P (2006) Real-time quanti-
tative PCR with SYBR Green I detection for estimating copy
numbers of nine drug resistance candidate genes in Plasmodium
falciparum. Malaria Journal 5, 1.
Ferreira ID, Lopes D, Martinelli A, Ferreira C, do Rosário VE &
Cravo P (2007) In vitro assessment of artesunate, artemether
and amodiaquine susceptibility and molecular analysis of
putative resistance-associated mutations of Plasmodium
ª 2008 Blackwell Publishing Ltd
Tropical Medicine and International Health
I. D. Ferreira et al. In vitro susceptibility and molecular typing of Plasmodium falciparum from Brazil
falciparum from São Tomé & Prı́ncipe. Tropical Medicine &
International Health 12, 353–362.
Henry M, Diallo I, Bordes J et al. (2006) Urban malaria in Dakar,
Senegal: chemosusceptibility and genetic diversity of Plasmo-
dium falciparum isolates. American Journal of Tropical Medi-
cine and Hygiene 75, 146–151.
Jambou R, Legrand E, Niang M et al. (2005) Resistance of Plas-
modium falciparum field isolates to in-vitro artemether and
point mutations of the SERCA-type PfATPase6. Lancet 366,
1960–1963.
Ke X, Collins A & Ye S (2001) PIRA PCR designer for restriction
analysis of single nucleotide polymorphisms. Bioinformatics 17,
838–839.
Leal O, Leal EA, Borges FR Jr et al. (2003) Clinical-parasitological
response to treatment with quinine associated to doxycycline in
uncomplicated falciparum malaria. Revista da Sociedade
Brasileira de Medicina Tropical 36, 751–754.
Livak KJ & Schmittgen TD (2001) Analysis of relative gene
expression data using real-time quantitative PCR and the 2–DDCt
method. Methods 25, 402–408.
Mayxay M, Pongvongsa T, Phompida S, Phetsouvanh R, White NJ
& Newton PN (2007) Diagnosis and management of malaria by
rural community health providers in the Lao People’s Demo-
cratic Republic (Laos). Tropical Medical International Health
12, 540–546.
Ministério da Saúde do Brasil (2006) A situação epidemiológica da
malária no Brasil 2006. Available at http://www.saúde.gov.br/
svs.
Mugittu K, Genton B, Mshinda H & Beck HP (2006) Molecular
monitoring of Plasmodium falciparum resistance to artemisinin
in Tanzania. Malaria Journal 5, 126.
Plowe CV, Djimde A, Bouare M, Doumbo O & Wellems TE
(1995) Pyrimethamine and proguanil resistance-conferring
mutations in Plasmodium falciparum dihydrofolate reductase:
polymerase chain reaction methods for surveillance in Africa.
American Journal of Tropical Medicine and Hygiene 52, 565–
568.
Price RN, Uhlemann AC, Brockman A et al. (2004) Mefloquine
resistance in Plasmodium falciparum and increased pfmdr1 gene
copy number. Lancet 364, 438–447.
Ramharter M, Burkhardt D, Nemeth J, Adegnika AA & Kremsner
PG (2006) In vitro activity of artemisone compared with
Corresponding Author Pedro Cravo, Centro de Malária e Outras Doenças Tropicais ⁄ IHMT ⁄ UEI Biologia Molecular ⁄ UEI Malária,
Lisbon 1349-008, Portugal. Tel. ⁄ Fax: +351 213622458; E-mail: pcravo@ihmt.unl.pt
ª 2008 Blackwell Publishing Ltd
volume 13 no 2 pp 199–207 february 2008
artesunate against Plasmodium falciparum. American Journal of
Tropical Medicine and Hygiene 75, 637–639.
Rowe AK, Rowe SY, Snow RW et al. (2006) The burden of
malaria mortality among African children in the year 2000.
International Journal of Epidemiology 35, 691–704.
Sidhu AB, Verdier-Pinard D & Fidock DA (2002) Chloroquine
resistance in Plasmodium falciparum malaria parasites con-
ferred by pfcrt mutations. Science 298, 210–213.
Sidhu AB, Uhlemann AC, Valderramos SG, Valderramos JC,
Krishna S & Fidock DA (2006) Decreasing pfmdr1 copy number
in Plasmodium falciparum malaria heightens susceptibility to
mefloquine, lumefantrine, halofantrine, quinine, and artemisi-
nin. Journal of Infectious Diseases 194, 528–535.
Vasconcelos KF, Plowe CV, Fontes CJ et al. (2000) Mutations in
Plasmodium falciparum dihydrofolate reductase and dihydro-
pteroate synthase of isolates from the Amazon region of Brazil.
Memórias do Instuto Oswaldo Cruz 95, 721–728.
Vieira PP, das Gracas Alecrim M, da Silva LH, Gonzalez-
Jimenez I & Zalis MG (2001) Analysis of the PfCRT K76T
mutation in Plasmodium falciparum isolates from the Amazon
region of Brazil. Journal of Infectious Diseases 183, 1832–
1833.
Vieira PP, Ferreira MU, Alecrim MG et al. (2004) pfcrt Poly-
morphism and the spread of chloroquine resistance in Plasmo-
dium falciparum populations across the Amazon Basin. Journal
of Infectious Diseases 190, 417–424.
Wessa P (2007) Free Statistics Software, Office for Research
Development and Education, version 1.1.21. Available at http://
www.wessa.net/.
WHO (1997) Instructions for the Use of the In Vitro Micro-test
Kit for the Assessment of the Response of Plasmodium falci-
parum to Chloroquine, Mefloquine, Quinine, Amodiaquine,
Sulfadoxine ⁄ Pyrimetamine and Artemisinin. Document
CTD ⁄ MAL ⁄ 97.20. WHO, Geneva.
Wootton JC, Feng X, Ferdig MT et al. (2002) Genetic diversity
and chloroquine selective sweeps in Plasmodium falciparum.
Nature 418, 320–323.
Zalis MG, Pang L, Silveira MS, Milhous WK & Wirth DF (1998)
Characterization of Plasmodium falciparum isolated from the
Amazon region of Brazil: evidence for quinine resistance.
American Journal of Tropical Medicine and Hygiene 58, 630–
637.
207