HAL Id: pasteur-00921203
https://hal-riip.archives-ouvertes.fr/pasteur-00921203
Submitted on 20 Jun 2014
HAL is a multi-disciplinary open access
archive for the deposit and dissemination of sci-
entic research documents, whether they are pub-
lished or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est
destinée au dépôt et à la diusion de documents
scientiques de niveau recherche, publiés ou non,
émanant des établissements d’enseignement et de
recherche français ou étrangers, des laboratoires
publics ou privés.
A molecular marker of artemisinin-resistant Plasmodium
falciparum malaria.
Frédéric Ariey, Benoit Witkowski, Chanaki Amaratunga, Johann Beghain,
Anne-Claire Langlois, Nimol Khim, Saorin Kim, Valentine Duru, Christiane
Bouchier, Laurence Ma, et al.
To cite this version:
Frédéric Ariey, Benoit Witkowski, Chanaki Amaratunga, Johann Beghain, Anne-Claire Langlois, et
al.. A molecular marker of artemisinin-resistant Plasmodium falciparum malaria.. Nature, Nature
Publishing Group, 2014, 505 (7481), pp.50-5. �10.1038/nature12876�. �pasteur-00921203�
ARTICLE
doi:10.1038/nature12876
A molecular marker of artemisinin-
resistant Plasmodium falciparum malaria
Fre
´
de
´
ric Ariey
1,2
{, Benoit Witkowski
3
, Chanaki Amaratunga
4
, Johann Beghain
1,2
{, Anne-Claire Langlois
1,2
, Nimol Khim
3
,
Saorin Kim
3
, Valentine Duru
3
, Christiane Bouchier
5
, Laurence Ma
5
, Pharath Lim
3,4,6
, Rithea Leang
6
, Socheat Duong
6
,
Sokunthea Sreng
6
, Seila Suon
6
,CharMengChuor
6
, Denis Mey Bout
7
, Sandie Me
´
nard
8
{, William O. Rogers
9
, Blaise Genton
10
,
Thierry Fandeur
1,3
, Olivo Miotto
11,12,13
, Pascal Ringwald
14
, Jacques Le Bras
15
, Antoine Berry
8
{, Jean-Christophe Barale
1,2
{,
Rick M. Fairhurst
4
*, Françoise Benoit-Vical
16,17
*, Odile Mercereau-Puijalon
1,2
* & Didier Me
´
nard
3
*
Plasmodium falciparum resistance to artemisinin derivatives in southeast Asia threatens malaria control and elimination
activities worldwide. To monitor the spread of artemisinin resistance, a molecular marker is urgently needed. Here,
using whole-genome sequencing of an artemisinin-resistant parasite line from Africa and clinical parasite isolates from
Cambodia, we associate mutations in the PF3D7_1343700 kelch propeller domain (‘K13-propeller’) with artemisinin
resistance in vitro and in vivo. Mutant K13-propeller alleles cluster in Cambodian provinces where resistance is
prevalent, and the increasing frequency of a dominant mutant K13-propeller allele correlates with the recent spread
of resistance in western Cambodia. Strong correlations between the presence of a mutant allele, in vitro parasite survival
rates and in vivo parasite clearance rates indicate that K13-propeller mutations are important determinants of
artemisinin resistance. K13-propeller polymorphism constitutes a useful molecular marker for large-scale surveillance
efforts to contain artemisinin resistance in the Greater Mekong Subregion and prevent its global spread.
The emergence of Plasmodium falciparum resistance to artemisinin
derivatives (ART) in Cambodia threatens the world’s malaria control
and elimination efforts
1,2
. The risk of ART-resistant parasites spread-
ing from western Cambodia to the Greater Mekong Subregion and to
Africa, as happened previously with chloroquine- and sulphadoxine/
pyrimethamine-resistant parasites
3–5
, is extremely worrisome. Clinical
ART resistance is defined as a reduced parasite clearance rate
1,6–10
,
expressed asan increased parasiteclearance half-life
11,12
, or a persistence
of microscopically detectable parasites on the third day of artemisinin-
based combination therapy (ACT)
2
. The half-life parameter correlates
strongly with results from the in vitro ring-stagesurvival assay(RSA
0–3 h
)
and results from the ex vivo RSA
13
, which measure the survival rate of
young ring-stage parasites to a pharmacologically relevant exposure
(700 nM for 6 h) to dihydroartemisinin (DHA)—the major metabolite
of all ARTs. However, the present lack of a molecular marker hampers
focused containment of ART-resistant parasites in areas where they
have been documented and hinders rapid detection of these parasites
elsewhere, where ACTs remain the most affordable, effective antima-
larials. To detect and monitor the spread of ART resistance, a molecu-
lar marker for widespread use is needed.
Recent genome-wide analyses of P. falciparum isolates have pro-
vided evidence of recent positive selection in geographic areas of ART
resistance
9,14–16
. Whereas parasite heritability of the clinical phenotype
is above 50%, no reliable molecular marker has yet been identified. One
possible explanation is that the parasite clearance half-life is not only
determined by the intrinsic susceptibility of a parasite isolate to ART,
but also by its developmental stage at the time of ART treatment and
host-related parameters such as pharmacokinetics and immunity
17
.
This issue was recently highlighted in patients presenting discordant
data between parasite clearance half-life in vivo and RSA
0–3 h
survival
rate in vitro
13
. Moreover, genome-wide association studies (GWAS)
are confounded by uncertainties about parasite population structure.
Recent evidence for several highly differentiated subpopulations of ART-
resistant parasites in western Cambodia
15
suggests that distinct emer-
gence events might be occurring. An alternative strategy to discover a
molecular marker is to analyse mutations acquired specifically by
laboratory-adapted parasite clones selected to survive high doses of
ART in vitro, and use this information to guide analysis of polymorph-
ism in clinical parasite isolates from areas where ART resistance is well
documented at both temporal and geographical levels. Here we used
this strategy to explore the molecular signatures of clinical ART res-
istance in Cambodia, where this phenotype was first reported
1,8
.
A candidate molecular marker of ART resistance
The ART-resistant F32-ART5 parasite line was selected by culturing
the ART-sensitive F32-Tanzania clone under a dose-escalating, 125-cycle
*These authors contributed equally to this work.
1
Institut Pasteur, Parasite Molecular Immunology Unit, 75724 Paris Cedex 15, France.
2
Centre National de la Recherche Scientifique, Unite
´
de Recherche Associe
´
e 2581, 75724 Paris Cedex 15, France.
3
Institut Pasteur du Cambodge, Malaria Molecular Epidemiology Unit, Phnom Penh, Cambodia.
4
Laboratory of Malaria and Vector Research, National Institute of Allergy and Infectious Diseases, National
Institutes of Health, Bethesda, Maryland 20892, USA.
5
Institut Pasteur, Plate-forme Ge
´
nomique, De
´
partement Ge
´
nomes et Ge
´
ne
´
tique, 75724 Paris Cedex 15, France.
6
National Center for Parasitology,
Entomology and Malaria Control, Phnom Penh, Cambodia.
7
SSA WHO, Drug Monitoring in Cambodia, National Center for Parasitology, Entomology and Malaria Control, Phnom Penh, Cambodia.
8
Service
de Parasitologie et Mycologie, Centre Hospitalier Universitaire de Toulouse, 31059 Toulouse Cedex 9, France.
9
Naval Medical Research Unit #2 Detachment, Phnom Penh, Cambodia.
10
Swiss Tropical and
Public Health Institute, 4051 Basel, Switzerland.
11
MRC Centre for Genomics and Global Health, University of Oxford, Oxford OX3 7BN, UK.
12
Mahidol-Oxford Tropical Medicine Research Unit, Mahidol
University, Bangkok 10400, Thailand.
13
Wellcome Trust Sanger Institute, Hinxton, Cambridge CB10 1SA, UK.
14
Global Malaria Program, World Health Organization, 1211 Geneva, Switzerland.
15
Centre
National de Re
´
fe
´
rence du Paludisme, CHU Bichat-Claude Bernard, APHP, PRES Sorbonne Paris Cite
´
, 75018 Paris, France.
16
Centre National de la Recherche Scientifique, Laboratoire de Chimie de
Coordination UPR8241, 31077 Toulouse Cedex 4, France.
17
Universite
´
de Toulouse, UPS, Institut National Polytechnique de Toulouse, 31077 Toulouse Cedex 4, France. {Present addresses: Institut
Pasteur, Genetics and Genomics of Insect Vectors Unit, 75724 Paris Cedex 15, France (F.A.); Institut Pasteur, Functional Genetics of Infectious Diseases Unit, 75724 Paris Cedex 15, France (J.B.); Centre de
Physiopathologie de Toulouse-Purpan, Institut National de la Sante
´
et de la Recherche Me
´
dicale UMR1043, Centre National de la Recherche Scientifique UMR5282, Universite
´
Toulouse III, 31024 Toulouse
Cedex 3, France (S.M. & A.B.); Institut Pasteur, Unite
´
de Biologie et Ge
´
ne
´
tique du Paludisme, Team Malaria Targets and Drug Development, 75724 Paris Cedex 15, France (J.-C.B.).
00 MONTH 2014 | VOL 000 | NATURE | 1
Macmillan Publishers Limited. All rights reserved
©2013
regimen of artemisinin for 5 years
18
. Whole-genome sequences were
obtained for both F32-ART5 and F32-TEM (its sibling clone cultured
without artemisinin) at 4603 and 5003 average nucleotide coverage,
respectively. Compared to F32-TEM, no deleted genes were identified
in F32-ART5. The exomes of F32-ART5 and F32-TEM were compared
after excluding (1) genes from highly variable, multi-gene families (var,
rifin and stevor), (2) positions with coverage lower than 25% of the
mean coverage of the parasite line, (3) single-nucleotide polymorph-
isms (SNPs) found to be mixed in F32-ART5, given that acquired
ART-resistance mutation(s) could be expected to be fixed in the sam-
ple after 5 years of continuous pressure, (4) SNPs shared between F32-
ART5 and the ART-sensitive 3D7 parasite strain and (5) synonymous
SNPs (Extended Data Fig. 1).
This analysis identified eight mutations in seven genes that were
subsequently confirmed by Sanger sequencing of PCR products (Extended
Data Table 1). Each gene harbours one mutant codon in F32-ART5
compared to F32-TEM, F32-Tanzania or 3D7 (Extended Data Table 2).
Information on the expression of the genes and the biological function
of the proteins are listed in Extended Data Table 3. Only one of these
genes, cysteine protease falcipain 2a (PF3D7_1115700), has previously
been associated with in vitro responses to ART
19
. To determine when
each mutation arose in the F32-ART5 lineage, we analysed the whole-
genome sequences of parasites at various drug-pressure cycles (Fig. 1).
This analysis showed that the PF3D7_0110400 D56V and PF3D7_
1343700 M476I mutations were acquired first, during the steep increase
of ART resistance, and remained stable thereafter. Importantly, the
appearance of these two mutations is associated with an increase in
the RSA
0–3 h
survival rate, from less than 0.01% to 12.8%. Subsequent
PCR analysis of the PF3D7_1343700 locus detected the M476I muta-
tion after 30 drug-pressure cycles, consistent with the sharp increase in
RSA
0–3 h
survival rate observed thereafter. The other SNPs appeared
stepwise at la ter stages of selection: PF3D7_021 3400 (68 cycles); PF3D 7_
1115700 (98 cycles); PF3D7_1302100, PF3D7_1459600 and PF3D7_
1464500 (120 cycles) (Extended Data Table 2). These data indicate that
the PF3D7_1343700 M476I mutation increased the resistance of F32-
Tanzania to DHA in the RSA
0–3 h
.
To explore whether these mutations are associated with ART resis-
tance in Cambodia, we investigated sequence polymorphism in all seven
genes by mining whole-genome or Sanger sequences for 49 culture-
adapted parasite isolates collected in 2010–2011 (see Methods). We chose
these isolates based on their differential RSA
0–3 h
survival rates (Sup-
plementary Table 1) and their sequences were compared to those of
control parasites lines 3D7, 89F5
20
and K1992 (see Methods). Three
genes (PF3D7_0110400, PF3D7_0213400 and PF3D7_1302100) encode
a wild-type sequence for all parasite isolates. The other four genes show
intra-population diversity, with previously reported or novel SNPs
(Supplementary Table 1). PF3D7_1115700 has 11 SNPs that are not
associated with RSA
0–3 h
survival rates (P 5 0.06, Kruskal–Wallis test).
PF3D7_1459600 has 6 SNPs that are not associated with survival rates
(P 5 0.65). PF3D7_1464500 has 12 SNPs previously reported in older
isolates from southeast Asia, including the ART-susceptible Dd2 line
21
,
probably reflecting a geographic signature. These SNPs also show no
significant association with survival rates (P 5 0.42). Therefore, these
six genes were not studied further.
In contrast, PF3D7_1343700 polymorphism shows a significant
association with RSA
0–3 h
survival rates (Fig. 2). Indeed, RSA
0–3 h
survival
rates differ substantially between parasite isolates with wild-type (median
0.17%, range 0.06–0.51%, n 5 16) or mutant (18.8%, 3.8–58%, n 5 33)
K13-propeller alleles (P , 10
24
, Mann–Whitney U test) (Supplementary
Table 1). Four mutant alleles are observed, each harbouring a single
non-synonymous SNP within a kelch repeat of the C-terminal K13-
propeller domain, namely Y493H, R539T, I543T and C580Y located
within repeats no. 2, 3, 3 and 4, respectively. Both the K1992 and the
ART-susceptible 89F5 lines carry a wild-type K13-propeller. There are
no associations betweenpolymorphisms in the K13-propellerand those
in the other candidate genes (Supplementary Table 1). Based on these
observations and the acquisition of M476I in kelch repeat no. 2 by
F32-ART5, we investigated whether K13-propeller polymorphism is a
molecular signature of ART resistance in Cambodia.
Emergence and spread of K13-propellermutants in Cambodia
Over the last decade, the prevalence of ART resistance has steadily
increased in the western provinces of Cambodia, but not elsewhere in
†
*
‡
PF3D7_0110400
PF3D7_1343700
PF3D7_0213400
PF3D7_1115700
PF3D7_1302100
PF3D7_1459600
PF3D7_1464500
24-h drug pressure
Number of drug-pressure cycles
0
60 120
Artemisinin concentration (μM)
RSA
0–3 h
survival rate (%)
100
10
1
0.1
0.01
10
1
0.1
0.01
48-h drug pressure
20 40
80
100
Figure 1
|
Temporal acquisition of mutations in F32-ART5. F32-Tanzania
parasites exposed to increasing artemisinin concentrations for 120 consecutive
cycles
18
were analysed by whole-genome sequencing at five time-points
(red arrows). Loci mutated after a given number of drug-pressure cycles
are shown (red boxes). The earliest time-points where three mutations were
detected by PCR (black arrows) are indicated by { for PF3D7_1343700, * for
PF3D7_0213400 and { for PF3D7_1115700. Orange and green circles indicate
RSA
0–3 h
survival rates for F32-ART5 and F32-TEM parasites, respectively
(mean of 3 experiments each).
0.001
0.1
1
10
100
RSA
0–3 h
survival rate (%)
Wild type C580Y R539T Y493H I543T
PF3D7_1343700 pol
y
morphisms
*
Figure 2
|
Survival rates of Cambodian parasite isolates in the RSA
0–3 h
,
stratified by K13-propeller allele. Genotypes were obtained by mining whole-
genome sequence data (n 5 21) or sequencing PCR products (n 5 28). Mutant
parasites have significantly higher RSA
0–3 h
survival rates than wild-type
parasites: wild type (n 5 17, median 0.16%, IQR 0.09–0.24, range 0.04–0.51);
C580Y (n 5 26, median 14.1%, IQR 11.3–19.6, range 3.8–27.3, P , 10
26
for
wild type versus C580Y, Mann–Whitney U test); R539T (n 5 5, median 24.2%,
IQR 12.6–29.5, range 5.8–31.3, P , 10
23
for wild type versus R539T); Y493H
(51.4%); and I543T (58.0%). The RSA
0–3 h
survival rate (0.04%) of control 3D7
parasites is indicated by an asterisk.
RESEARCH ARTICLE
2 | NATURE | VOL 000 | 00 MONTH 2014
Macmillan Publishers Limited. All rights reserved
©2013
the country
2
. To test whether the spatiotemporal distribution of K13-
propeller mutations correlates with that of ART resistance, we sequenced
the K1 3-propeller of archived parasite isolates from Cambodian patients
with malaria in 2001–2012 (Extended Data Table 4). Data from six
provinces were compared (n 5 886): Pailin, Battambang and Pursat in
the west where ART resistance is established
1,6,8,22
, Kratie in the south-
east where ART resistance has increased in recent years
2
, and Preah
Vihear in the north and Ratanakiri in the northeast where there was
virtually no evidence of ART resistance during this time period
2
. This
analysis reveals overall 17 mutant alle les, including three high-frequency
(. 5%) alleles (C580Y, R539T and Y493H). The frequency of wild-type
sequence decreased significantly over time in all three western pro-
vinces, but not in Preah Vihear or Ratanakiri. The frequency of the
C580Y allele increased significantly from 2001–2002 to 2011–2012 in
Pailin and Battambang, indicating its rapid invasion of the population
and near fixation in these areas (Fig. 3).
To further investigate the geographic diversity of K13-propeller
polymorphism in Cambodia, we extended our sequence analysis to
include data from four additional provinces (n 5 55, Kampong Som,
Kampot, Mondulkiri and Oddar Meanchey) in 2011–2012 (Extended
Data Table 4). Although a large number of mutations are observed (Sup-
plementary Fig. 1 and Ex tended Dat a Tabl e 5), the C580Y alleleaccounts
for 85% (189/222) of all mutant all eles observed in 2011–2012 (Extend ed
Data Fig. 2). This mapping outlines the elevated frequency (74%, 222/
300) of parasites harbouring a single non-synonymous mutation in
the K13-propeller and the geographic disparity of their distribution.
Importantly, the frequency distribution of mutant alleles over the vari-
ous provinces matches that of day 3 positivity in patients treated for
malaria with an ACT (Spearman’s r 5 0.99, 95% confidence interval
0.96–0.99, P , 0.0001), considered a suggestive sign of clinical ART
resistance (Extended Data Fig. 3).
K13-propeller polymorphisms and clinical AR T resistance
To confirm that K13-propeller polymorphism is a molecular marker
of clinical ART resistance, we first identified 163 patients from Pursat
and Ratanakiri in whom we measured parasite clearance half-lives
(range 1.58–11.53 h)
6
in 2009–2010 and for which parasites were prev-
iously assigned to a KH subpopulation (KH1, KH2,KH3, KH4 or KHA)
on the basis of ancestry analysis of whole-genome sequence data
15
.
Thirteen patients with mixed genotypes (a wild-type and one or more
mutant K13-propeller alleles) were excluded. Of the remaining 150
patients, 72 carried parasites with a wild-type allele and the others carried
parasites with onlya single non-synonymous SNP in the K13-propeller:
C580Y (n 5 51), R539T (n 5 6) and Y493H (n 5 21) (Extended Data
Table 6). The para site clearance ha lf-life in patients with wild -type para-
sites is significantly shorter (median 3.30 h, interquartile range (IQR)
2.59–3.95) than those with C580Y (7.19 h, 6.47–8.31, P , 10
26
,Mann–
Whitney U test), R539T (6.64 h, 6.00–6.72, P , 10
24
) or Y493H(6.28 h,
5.37–7.14, P , 10
26
) parasites (Fig. 4a). Also, the parasite clearance
half-life in patients carrying C580Y parasites is significantly longer than
those with Y493H parasites (P 5 0.007, Mann–Whitney U test). These
data indicate that C580Y, R539T and Y493H identify slow-clearing
parasites in malaria patients treated with ART.
Because KH2, KH3, KH4 and KHA parasites have longer half-lives
than KH1 parasites
15
, we proposed that allelic variation in the K13-
propeller accounts for these differences. Among 150 parasites, 55, 26,
14, 12 and 43 are classified as KH1, KH2, KH3, KH4 and KHA, respectively.
Three K13-propeller alleles strongly associate with KH groups: 96%
(53/55) of KH1, 96% (25/26) of KH2 and 100% (12/12) of KH4 para-
sites carry the wild-type, C580Y and Y493H alleles, respectively (Extended
Data Table 6). Whereas KH3 parasites (n 5 14) carry the wild-type,
C580Y and R539T alleles, R539T is not observed in KH1, KH2 or KH4
Pailin
(western Cambodia)
Battambang
(western Cambodia)
Pursat
(western Cambodia)
Kratie
(southeastern Cambodia)
Preah Vihear
(northern Cambodia)
Ratanakiri
(northeastern Cambodia)
Wild type
C580Y
R539T
Y493H
G449A
I543T
N537I
T474I
P574L
R561H
A481V
D584V
S623C
P553L
V568G
N458Y
G533S
T508N
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
2001–2002
n = 40
2003–2004
n = 43
2005–2006
n = 46
2007–2008
n = 95
2009–2010
n = 66
2011–2012
n = 66
2001–2002
n = 64
2003–2004
No data
2005–2006
No data
2007–2008
No data
2009–2010
No data
2011–2012
n = 71
2001–2002
No data
2003–2004
n = 10
2005–2006
No data
2007–2008
No data
2009–2010
n = 43
2011–2012
n = 19
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
2001–2002
n = 15
2003–2004
No data
2005–2006
No data
2007–2008
No data
2009–2010
No data
2011–2012
n = 17
2001–2002
n = 27
2003–2004
n = 27
2005–2006
n = 25
2007–2008
n = 24
2009–2010
No data
2011–2012
n = 19
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
2001–2002
n = 56
2003–2004
n = 30
2005–2006
n = 22
2007–2008
No data
2009–2010
n = 8
2011–2012
n = 35
Figure 3
|
Frequency of K13-propeller alleles in 886 parasite isolates in
six Cambodian provinces in 2001–2012. Genotypes were obtained by
sequencing PCR products from archived blood samples. All mutant alleles
carry a single non-synonymous SNP (colour-coded, same colour codes as in
Fig. 2 for wild type, C580Y, R539T, Y493H and I543T). Significant reductions
(Fisher’s exact test) in wild-type allele frequencies were observed in Pailin,
Battambang, Pursat and Kratie over time (see Methods).
ARTICLE RESEARCH
00 MONTH 2014 | VOL 000 | NATURE | 3
Macmillan Publishers Limited. All rights reserved
©2013
parasites. As expected, KHA parasites have a mixed allele composition.
Importantly, K13-propeller mutations more accurately identify slow-
clearing parasites than KH group (Fig. 4b), demonstrating that the
association of K13-propeller polymorphism with clinical ART resist-
ance in Cambodia is partially independent of the genetic background
of KH subpopulations. Within the KH1 group (n 5 55), the parasite
clearance half-life in patients with wild-type parasites is significantly
shorter (n 5 53, median 2.88 h, IQR 2.52–3.79) than those with Y493H
parasites (n 5 2, median 6.77 h, P 5 0.02, Mann–Whitney U test). Within
the KH3 subpopulation (n 5 14), the half-life in patients with wild-type
parasites is shorter (n 5 3, median 3.65 h) than those with C580Y
(n 5 7, median 8.73 h, IQR 7.35–9.06, P 5 0.02) or R539T (n 5 4,
6.65 h, 6.29–6.80, P 5 0.03) parasites.
Discussion
The F32-ART5 lineage acquired a K13-propeller mutation as it developed
ART resistance, as indicated by its ability to survive a pharmacologi-
cally relevant exposure to DHA in the RSA
0–3 h
. Genes putativelyassociate d
with ART resistance (Pfcrt
23,24
, Pftctp
25,26
, Pfmdr1
8,27,28
, Pfmrp1
27–29
and
ABC transporters
30
) or enc oding putative targets of ART (PfATPase6
31,32
and Pfubcth—the orthologue of Plasmodium chabaudi ubp1
33,34
) were
not mutated during the 5-year selection of F32-ART5, and Pfmdr1
amplification was not observed
35–40
. In addition, all candidate ART-
resistance genes recently identified using population genetics appro-
aches
14,40,41
remained unaltered in F32-ART5, except for PF3D7_1343700
and PF3D7_1459600 located in the linkage-disequilibrium windows
identified in ref. 16. These findings led us to identify another 17 single
K13-propeller mutations in naturally circulating parasites in Cambodia.
Several of these mutations associate strongly with the spatiotemporal
distribution of ART resistance in Cambodia, increased parasite sur-
vival rates in response to DHA in vitro, and long parasite clearance
half-lives in response to ART treatment in vivo. None of the six other
genes mutated in F32-ART5 associate with RSA
0–3 h
survival rates in
parasite isolates from Cambodia.
K13-propeller polymorphism fulfils the definition of a molecular
marker of ART resistance for several reasons: (1) there has been a pro-
gressive loss of wild-type parasites in western Cambodia during the
decade of emerging ART resistance in this region; (2) mutant para-
sites cluster in Cambodian provinces where ART resistance is well
established and are less prevalent where ART resistance is uncommon;
(3) PF3D7_1343700 is located 5.9 kilobases upstream of the 35-kb locus
identified in ref. 14 as being under recent positive selection, and within
the region of top-ranked signatures of selection outlined in ref. 16;
(4) multiple mutations, all non-synonymous, are present in the K13-
propeller, reflecting positive selection rather than a hitchhiking effect
or genetic drift; (5) mutations occur in a domain that is hi ghly conserved
in P. falciparum, with only one non-synonymous SNP being documen-
ted in a single parasite isolate from Africa
42
; (6) all polymorphisms
we observe in Cambodia are novel and all but one (V568G) occur at
positions strictly conserved between Plasmodium species (Supplemen-
tary Fig. 1 and Supplementary Fig. 2), suggesting strong structural and
functional constraints on the protein; (7) the three most-prevalent
K13-propeller mutations correlate strongly with RSA
0–3 h
survival
rates in vitro and parasite clearance half-lives in vivo at the level of
individual parasite isol ates and ma laria patients, re spectively; and (8) the
frequency of mutant alleles correlates strongly with the prevalence of
day 3 positivity after ACT treatment at the level of human populations
in Cambodia.
On the basis of homology with other kelch propeller domains, we
anticipate that the observed K13-propeller mutations destabilize the
domain scaffold and alter its function. The carboxy-terminal portion
of PF3D7_1343700 encodes six kelch motifs, which are found in a large
number of proteins with diverse cellular functions
43,44
. Given that the
toxicity of ART derivatives depends principally on their pro-oxidant
activity, the reported role of some kelch-containing proteins in regu-
lating cytoprotective and protein degradation responses to external
stress is particularly intriguing. The K13-propeller shows homology
with human KLHL12 and KLHL2, involved in ubiquitin-based protein
degradation, and KEAP1, involved in cell adaptation to oxidative stress
(Extended Data Fig. 4). More work is needed to delineate the normal
function of K13 and the effect of various mutations. Allele exchange
studies in mutant and wild-type parasites may help to define the con-
tribution of K13-propeller polymorphisms on different genetic back-
grounds to the RSA
0–3 h
survival rate. Indeed, it is particularly worrying
that as few as two mutations, that is, the K13-propeller M476I and
PF3D7_0110400 D56V, were sufficient to confer ART resistance to F32-
Tanzania, which has a typical African genetic background. Cambodian
parasites with mutant K13-propellers display a wide range of RSA
0–3 h
survival rates (3.8–58%) and parasite clearance half-lives (4.5-11.5 h).
Further studies are therefore required to identify additional genetic
determinants of ART resistance, which may reside in the strongly
selected regions recently identified
14,16
. In this context, analysing the
RSA
0–3 h
survival rates as a quantitative trait among parasites harbour-
ing the same K13-propeller mutation could help to identify additional
genetic loci involved in ART resistance.
12
10
8
6
4
2
0
12
10
8
6
4
2
0
Parasite clearance half-life (h)Parasite clearance half-life (h)
Wild type C580Y R539T Y493H
PF3D7_1343700 polymorphisms
KH1 KH2 KH3 KH4 KHA
KH subpopulations
a
b
Figure 4
|
Parasite clearance half-lives. a, Correlation of parasite clearance
half-lives and K13-propeller alleles for parasite isolates in Pursat and Ratanakiri
in 2009–2010. Wild-type parasites have shorter half-lives (median 3.30 h, IQR
2.59–3.95, n 5 72) than C580Y (7.19 h, 6.47–8.31, n 5 51, P , 10
26
, Mann–
Whitney U test), R539T (6.64 h, 6.00–6.72, n 5 6, P , 10
26
) or Y493H (6.28 h,
5.37–7.14, n 5 21, P , 10
26
) parasites. The half-life of C580Y parasites is
significantly longer than that of Y493H parasites (P 5 0.007). b, Correlation of
parasite clearance half-lives, KH subpopulations
15
and K13-propeller alleles for
the same 150 parasite isolates. Half-lives are shown for Pursat (squares) and
Ratanakiri (triangles) parasites, stratified by KH group and K13-propeller allele
(colour-coded as in a). Median half-lives stratified by K13-propeller allele are
KH1: wild type (2.88) and Y493H (6.77); KH2: C580Y (7.13) and Y493H (4.71);
KH3: wild type (3.65), C580Y (8.73) and R539T (6.65); KH4: Y493H (6.37); and
KHA: wild type (4.01), C580Y (7.09), Y493H (6.18) and R539T (5.73).
RESEARCH ARTICLE
4 | NATURE | VOL 000 | 00 MONTH 2014
Macmillan Publishers Limited. All rights reserved
©2013