Diversity of Plasmodium falciparum chloroquine resistance transporter (pfcrt) exon 2 haplotypes in the Pacific from 1959 to 1979

Abstract

Nearly one million deaths are attributed to malaria every year. Recent reports of multi-drug treatment failure of falciparum malaria underscore the need to understand the molecular basis of drug resistance. Multiple mutations in the Plasmodium falciparum chloroquine resistance transporter (pfcrt) are involved in chloroquine resistance, but the evolution of complex haplotypes is not yet well understood. Using over 4,500 archival human serum specimens collected from 19 Pacific populations between 1959 and 1979, the period including and just prior to the appearance of chloroquine treatment failure in the Pacific, we PCR-amplified and sequenced a portion of the pfcrt exon 2 from 771 P. falciparum-infected individuals to explore the spatial and temporal variation in falciparum malaria prevalence and the evolution of chloroquine resistance. In the Pacific, the prevalence of P. falciparum varied considerably across ecological zones. On the island of New Guinea, the decreases in prevalence of P. falciparum in coastal, high-transmission areas over time were contrasted by the increase in prevalence during the same period in the highlands, where transmission was intermittent. We found 78 unique pfcrt haplotypes consisting of 34 amino acid substitutions and 28 synonymous mutations. More importantly, two pfcrt mutations (N75D and K76T) implicated in chloroquine resistance were present in parasites from New Hebrides (now Vanuatu) eight years before the first report of treatment failure. Our results also revealed unexpectedly high levels of genetic diversity in pfcrt exon 2 prior to the historical chloroquine resistance selective sweep, particularly in areas where disease burden was relatively low. In the Pacific, parasite genetic isolation, as well as host acquired immune status and genetic resistance to malaria, were important contributors to the evolution of chloroquine resistance in P. falciparum.

Figure 1. Approximate locations of populations.

Populations are denoted by their sample codes (Table S1) and assigned to one of five ecological zones as indicated by colors as followed: WNG coast (purple), PNG coasts (green), PNG Papuan Plateau (blue), PNG eastern highlands (red), and Island Melanesia (orange). Samples collected from the same locations in different years (e.g. SWC60, SWC62, and SWC69) did not represent repeated samplings from the same populations.

Figure 2. Temporal variation in P. falciparum infection rates (bar) and pfcrt haplotype diversities (line).

Ecological zones are indicated by colors as followed: PNG eastern highlands (red), WNG coast (purple), and PNG coasts (green).

Figure 3. Median-joining network of pfcrt haplotypes (A) and alignment of translated sequences (B) from 1959–61.

In the network diagram, each unique pfcrt haplotype is represented by a node, and the size of each node is proportional to the frequency of the haplotype. Substitutions are represented by branches, with the branch length proportional to the number of substitutions. Non-synonymous substitutions are represented by branches in bold. Pfcrt haplotypes that appeared in more than one time period are noted by their haplotype numbers. In the amino acid sequence alignment, the shaded region represents transmembrane domain 1. Synonymous substitutions are represented by dashes (-) and synonymous substitutions by asterisks (*). Populations are denoted by their sample codes (Table S1).

Figure 4. Median-joining network of pfcrt haplotypes (A) and alignment of translated sequences (B) from 1962–5.

Figure 5. Median-joining network of pfcrt haplotypes (A) and alignment of translated sequences (B) from 1969–72.

Hypothetical nodes are indicated by arrows.

Figure 6. Median-joining network of pfcrt haplotypes (A) and alignment of translated sequences (B) from 1979.

Figure 7. Gene flow among P. falciparum populations from different ecological zones.

Gene flow among parasite populations from (A) 1959–61, (B) 1962–5, (C) 1969–72, and (D) 1979, is represented by solid lines. Gene flow is inferred from pairwise F_ST distances determined to be non-statistically significant (p>0.05).