SHORT REPORT: RARE PLASMODIUM FALCIPARUM MEROZOITE SURFACE PROTEIN 1 19-KDA (MSP-1₁₉) HAPLOTYPES IDENTIFIED IN MALI USING HIGH-THROUGHPUT GENOTYPING METHODS

Abstract

Genetic diversity in malaria vaccine antigens may compromise malaria vaccine efficacy, so it is important to understand this diversity and the processes that generate it. By applying new high-throughput genotyping methods to a large sample of infections from Mali (N = 1369), seven new 19-kDa merozoite surface protein 1 (MSP-1₁₉) haplotypes were identified. Herein we report the sequences of these new haplotypes and discuss their possible origins. Although they are present in < 1% of the samples examined, the existence of these rare haplotypes reveals a greater degree of diversity at this locus than previously reported and highlights the potential for Plasmodium to evolve under selective pressure from the immune system and from such interventions as vaccines and drugs.

INTRODUCTION

Progress toward a malaria vaccine has been slow due in part to the extensive genetic variability in Plasmodium. Such genetic variability is generated through mutation under selective pressure from the human immune system and sexual recombination in the mosquito vector, and is particularly prevalent in surface antigens being targeted for malaria vaccines.¹ Merozoite surface protein 1 is a candidate antigen for a blood-stage malaria vaccine. The 195-kDa precursor of this protein undergoes two rounds of proteolytic cleavage, leaving only the C-terminal 19 kDa on the surface of the merozoite as it invades the erythrocyte.² MSP-1₁₉ contains two epidermal growth factor (EGF)-like motifs that are thought to play a role in erythrocyte invasion.³ Antibodies to this region can block erythrocyte invasion in vitro² and are associated with protection from clinical malaria in field studies.^4–9 The sequence of MSP-1₁₉ is highly conserved¹⁰; however, six non-synonymous single nucleotide polymorphisms (SNPs) have been documented at amino-acid positions 1644, 1691, 1699, 1700, 1701, and 1716,^10–14 and it is unclear how these polymorphisms affect immunity. Intragenic recombination has been proposed as an important mechanism for generating novel genetic variants in MSP-1₁₉¹³; however, new variants can also be derived from single-nucleotide mutations that are maintained by positive natural selection.¹

If vaccine efficacy is allele-specific, then vaccination against polymorphic antigens could lead to selection for nontarget alleles in the parasite population, compromising vaccine efficacy. Such vaccine-induced selection has been suggested by theoretical studies^15–19 and has been observed in a clinical trial of a blood-stage vaccine.²⁰ We report the sequence of seven rare MSP-1₁₉ haplotypes identified at a malaria vaccine-testing site in Mali and discuss the possible origins of these new haplotypes and the potential implications of genetic diversity for the efficacy of MSP-1–based vaccines.

New MSP-1₁₉ haplotypes were identified from samples collected in a cohort study conducted in Bandiagara, Mali, during the years 1999–2001.²¹ From July to January of each year, individuals were visited weekly and contributed a filter paper blood sample at least monthly and at every clinical malaria episode. Among 629 study participants, 100 who had at least 2 years of follow up were randomly selected within three age strata: 30 children of age ≤ 5 years, 32 children of age 6–10 years, and 38 children of age ≥ 11 years.²¹ Samples were collected under protocols reviewed and approved by Institutional Review Boards of the University of Maryland School of Medicine and the University of Bamako Faculty of Medicine. Informed consent was obtained from all study participants or their guardians.

MSP-1₁₉ was amplified from samples collected at monthly surveys and clinical episodes occurring during the transmission season in the 3 years of the incidence study. A single PCR was used to amplify MSP-1₁₉ from samples with parasitemia > 1,000 parasites/μL, and a nested PCR was used to amplify MSP-1₁₉ from samples with parasitemia < 1,000 parasites/μL and microscopy-negative samples. Of the 2309 samples that underwent PCR (including microscopy-negative samples), 1375 were parasite-positive (by PCR).²¹

All PCR-positive samples underwent Pyrosequencing to determine allele frequencies at each of the six SNPs in MSP-1₁₉.²² Pyrosequencing (Biotage, Charlottesville, VA) is a high-throughput method that allows quantification of the proportions of alternative nucleotides at each SNP. Of the 1,375 PCR-positive samples, 1,369 gave successful MSP-1₁₉ genotyping results.²¹

A mathematical model was used to estimate the frequency of 14 confirmed MSP-1₁₉ haplotypes in each genotyped sample.^21,22 The haplotype-estimating algorithm uses maximum likelihood methods to determine the most probable combination of haplotypes given the allele frequencies for an infection, the haplotypes known to be circulating in the population, and a probability distribution of the measurement errors. Three of the 14 haplotypes included in the haplotype-estimating algorithm had not been observed previously and were confirmed by reamplification of MSP-1₁₉ followed by PCR cloning.²² When applied to the 1,369 genotyped samples, the algorithm was able to resolve haplotype frequencies for all but six samples given the list of 14 haplotypes.²¹

The six samples that were unable to be resolved by the algorithm had allele frequencies consistent with the presence of additional new MSP-1₁₉ haplotypes. The PCR and Pyrosequencing were repeated to rule out genotyping error as an explanation for the observed allele frequencies. Upon obtaining the same results, MSP-1₁₉ was reamplified from these six samples, using non-biotinylated MSP-1₁₉ primers, and cloned. Twelve clones were picked from each transformation. Pyrosequencing and direct sequencing of MSP-1₁₉ clones from the six samples revealed four additional new haplotypes: QKSNRF, QKSSGF, EKSNRL, and EKNNGF. Nucleotide and amino-acid alignments of the seven new haplotypes (three from the previous study²² and four from the current study) are shown in Figure 1, as well as haplotypes with substitutions at positions other than those at the six known SNPs (sequences available in GenBank, accession numbers DQ677569–DQ677579). Sequences for the 3D7 and FVO P. falciparum strains are also included for reference. As indicated in the figure, the QKSSGF haplotype also has nonsynonymous substitutions at positions 1674 (N→D) and 1690 (A→V). Additional non-synonymous substitutions were observed in other clones at residues 1647 (G→R, clones 0818c1–6 and 1–8), 1673 (E→G, clone 0818c1–6), and 1677 (P→L, clone 0818c1–7). The QKSSRL haplotype has a synonymous substitution at codon 1667. Synonymous substitutions were also observed in other clones at codons 1659 (0818c1–6), 1666 (0818c4–1), 1698 (0818c1–7, 4-1), and 1699 (0818c1–6, 1–8). It is possible that other polymorphic sites exist in the parasites infecting the cohort; however, because Pyrosequencing^™ genotypes the short regions surrounding known SNPs, only those samples giving unusual Pyrosequencing^™ results were flagged for cloning and direct sequencing.

Figure 1.

Alignment of novel MSP-1₁₉ haplotypes observed in Mali: (A) nucleotide alignment; (B) amino-acid alignment. Shaded text indicates SNPs at positions 1644, 1691, 1699, 1700, 1701, and 1716. Haplotypes based on these six positions are in parentheses.

Using the haplotype-estimating algorithm, QKSSGL, QKSSRL, and QTSSGL had prevalences of 0.07%, 1.3%, and 0.51%, respectively, in the cohort. QKSNRF, QKSSGF, EKSNRL, and EKNNGF were not included in the haplotype-estimating algorithm and were found in one sample each.

Table 1 contains a comprehensive list of MSP-1₁₉ haplotypes reported in the literature. Including the seven new haplotypes identified in Mali, 22 haplotypes have been documented, including one isolate from India that contained a Y allele at position 1700.²³ All but four of the reported haplotypes (EKSSGL, QTSSRF, ETSSRF, and EKSYGF) have been observed in the samples from Mali.

Table 1.

Comprehensive list of MSP-1₁₉ haplotypes reported in the literature to date

	Amino acid position
Haplotype	1644 (E/Q)	1691 (T/K)	1699 (S/N)	1700 (S/N)	1701 (R/G)	1716 (L/F)	Isolates/country	Reference no.
1	Q	K	S	N	G	L	FVO, Wellcome	29
2	E	T	S	S	R	L	3D7, MAD20	30
3	E	K	S	N	G	L	FUP, Uganda-PA	31
4	Q	K	S	N	G	F	T807 (Thai)	32
5	Q	T	S	S	R	L	Indo	33
6	E	K	S	S	R	L	Kenya-2	13
7	E	K	N	N	G	L	Kenya	13
8*	E	K	S	S	G	L	Kenya-1	13
9	E	K	S	N	G	F	Kenya-3	13
10	Q	K	N	N	G	L	Thai-Variant 2	14
11	E	T	S	S	G	L	India	34
12*	E	T	S	S	R	F	Brazil-1	11
13*	Q	T	S	S	R	F	Brazil-2	11
14	E	T	S	N	G	L	Vietnam	12
15*	E	K	S	Y†	G	F	India	23
16	Q	K	S	S	G	L	Mali-1
17	Q	T	S	S	G	L	Mali-2
18	Q	K	S	S	R	L	Mali-3
19	Q	K	S	N	R	F	Mali-4
20	Q	K	S	S	G	F	Mali-5
21	E	K	S	N	R	L	Mali-6
22	E	K	N	N	G	F	Mali-7

^*Haplotypes not observed in Bandiagara, Mali.

^†The Y allele at position 1700 was recently reported in one isolate from India.

The role of recombination in the generation of genetic diversity in Plasmodium has been debated,^24,25 but intragenic recombination has been implicated as a factor in generating diversity in MSP-1.^13,14 Three of the haplotypes observed in this study (QKSSRL, QKSSGL, and EKSNRL) were predicted to exist based on single and double crossover events between previously identified alleles¹³; however, until now they had not been identified in field isolates. Table 2 shows how the seven new haplotypes identified in this study could have arisen via recombination events. As indicated in the table, all but one of the seven new haplotypes (QKSNRF) could have arisen from single crossover events between haplotypes observed in Mali. If all known haplotypes are considered (including those not observed at the site), then all seven haplotypes could have been generated via single crossovers.

Table 2.

Possible recombination events that could have generated new MSP-1₁₉ haplotypes

Recombination events	Progeny
Single crossover
Q* KNNGL × E* TSSGL	EKNNGL, QTSSGL
Q* K^ NNGL × E* K^ SSRL	EKNNGL, QKSSRL
Q* KSNGL × E* TSSGL	EKSNGL, QTSSGL
QK* SNGL × ET* SSGL	ETSNGL, QKSSGL
Q* T^ SSRL × E* T^ SSGL	ETSSRL, QTSSGL
Q* K^ SNGL × E* K^ SSRL	EKSNGL, QKSSRL
Q* TSSRL × E* KSSRL	ETSSRL, QKSSRL
QKSNG* F × EKNNG* L	QKSNGL, EKNNGF
QKSSG* L × EKSNG* F	EKSNGL, QKSSGF
QKSSG* L × QKSNG* F	QKSNGL, QKSSGF
EKSN* GL × EKSS* RL	EKSSGL, EKSNRL
Q* K^ SNGL × E*K^SSGL	EKSNGL, QKSSGL
Q* TSSRL × E*KSSGL	ETSSRL, QKSSGL
Q* K^ NNGL × E*K^SSGL	EKNNGL, QKSSGL
QT* SSRL × EK*SSGL	EKSSRL, QTSSGL
QKSN* GL × ETSS*RF	ETSSGL, QKSNRF
QKSN* GL × QTSS*RF	QTSSGL, QKSNRF
QTSSR*F × EKNNG* L	QTSSRL, EKNNGF
Double crossover
Q1 KSNG2 F × E1 KSNR2 L	EKSNGL, QKSNRF
QKSN1 G2 F × ETSS1 R2 L	ETSSGL, QKSNRF
QK1 SNG2 F × ET1 SSG2 L	ETSNGL, QKSSGF
Q1 KSN2 GL × E1 KSS2 RL	QKSSGL, EKSNRL
Q1 KNNG2 L × E1 KSNG2 F	QKSNGL, EKNNGF

^*Position of crossover.

^{^}Alternative position for the crossover.

^1,2Crossover positions in a double-crossover event. Haplotypes in bold text indicate new haplotypes identified in this study. Haplotypes in underlined text indicate haplotypes that have not been observed at the Bandiagara, Mali site and therefore represent events that are less likely to have occurred in Bandiagara but that could have occurred elsewhere and been imported.

Although reshuffling of known polymorphisms via recombination could be responsible for the observed haplotypes, these new haplotypes could also be the result of convergence (i.e., selection for mutations that are identical but that do not share common ancestry).^1,26 For example, the haplotype QKSSRL could have arisen from a single nucleotide change from GAA to CAA in codon 1644 of the EKSSRL haplotype or from a single nucleotide change from ACA to AAA in codon 1691 of the QTSSRL haplotype. However, without additional sequence information from adjacent regions (e.g., neighboring microsatellites), it is difficult to distinguish recombination from convergent point mutations in this context.¹

In conclusion, by applying new high-throughput genotyping methods to a large sample of infections from Mali, seven new MSP-1₁₉ haplotypes have been identified, representing a 50% increase in the number of haplotypes previously reported in the literature. Few studies have examined MSP-1₁₉ genetic diversity in Africa,¹³ even though Africa carries the heaviest malaria burden and most MSP-1–based vaccines currently being developed and tested target this region of the protein.^27,28 Additional MSP-1₁₉ diversity may continue to be discovered as high-throughput methods are used to conduct large molecular epidemiology studies of this locus in other malaria endemic areas of Sub-Saharan Africa.

Although the impact of genetic diversity in MSP-1₁₉ on immunity and efficacy of MSP-1–based vaccines is not clear, it is possible that vaccination with one of the common MSP-1₁₉ haplotypes could give a competitive advantage to rare haplotypes such as those observed in this study, allowing them to increase in frequency in the parasite population.^16,17,19 Understanding the mechanisms by which diverse MSP-1₁₉ haplotypes arise may improve our ability to predict how Plasmodium will evolve in response to interventions such as vaccines and drugs.