Genetic diversity of expressed Plasmodium falciparum var genes from Tanzanian children with severe malaria

Abstract

Background: Severe malaria has been attributed to the expression of a restricted subset of the var multi-gene family, which encodes for Plasmodium falciparum erythrocyte membrane protein 1 (PfEMP1). PfEMP1 mediates cytoadherence and sequestration of infected erythrocytes into the post-capillary venules of vital organs such as the brain, lung or placenta. var genes are highly diverse and can be classified in three major groups (ups A, B and C) and two intermediate groups (B/A and B/C) based on the genomic location, gene orientation and upstream sequences. The genetic diversity of expressed var genes in relation to severity of disease in Tanzanian children was analysed.

Methods: Children with defined severe (SM) and asymptomatic malaria (AM) were recruited. Full-length var mRNA was isolated and reversed transcribed into var cDNA. Subsequently, the DBL and N-terminal domains, and up-stream sequences were PCR amplified, cloned and sequenced. Sequences derived from SM and AM isolates were compared and analysed.

Results: The analysis confirmed that the var family is highly diverse in natural Plasmodium falciparum populations. Sequence diversity of amplified var DBL-1α and upstream regions showed minimal overlap among isolates, implying that the var gene repertoire is vast and most probably indefinite in endemic areas. var DBL-1α sequences from AM isolates were more diverse with more singletons found (p<0.05) than those from SM infections. Furthermore, few var DBL-1α sequences from SM patients were rare and restricted suggesting that certain PfEMP1 variants might induce severe disease.

Conclusions: The genetic sequence diversity of var genes of P. falciparum isolates from Tanzanian children is large and its relationship to disease severity has been studied. Observed differences suggest that different var genes might have fundamentally different roles in the host-parasite interaction. Further research is required to examine clear disease-associations of var gene subsets in different geographical settings. The importance of very strict clinical definitions and appropriate large control groups needs to be emphasized for future studies on disease associations of PfEMP1.

Figure 1

Distribution of unique sequence types (STs) of DBL1α in clinical isolates. Blue columns represent STs found in multiple samples from both AM and SM groups. Red columns represent STs found in multiple samples within the SM group. Pink columns represent STs found in multiple samples within the AM group. Yellow columns represent STs specific to each individual isolate.

Figure 2

Distribution of DBL-1α sequences into cys/PoLV groups by clinical status. SM DBL-1α sequences had more than 50% cys2 sequence tags (1–3 groups) compared to 27% in AM isolates.

Figure 3

Distribution of PoLVmotifs within clinical isolates SM (red bars) and in AM (white bars).

Figure 4

Cumulative diversity curves for DBL-1α sequences from Ifakara. The cumulative curve for DBL-1α was determined by simulation the number of unique sequences as a function of the number of patient samples. For each number of patient samples the statistics value was obtained from simulations of all possible sample combinations.

Figure 5

Phylogenetic network showing the comparison of 3 dominant DBL-1α sequence tags transcribed from clinical isolates, generated using Neighbour-Net [44]. Sequences transcribed by isolates with severe malaria (ISM, blue) and asymptomatic malaria (IAM, red) are compared. The sequences fall into two major clades separated with dotted lines, the upper cluster formed 2 subgroups of sequences isolated from severe patients, one with group A and the remaining var group B/A homology to 3D7. The majority of sequences from asymptomatic patient isolates clustered together and were homologous to group B, B/C and C of the 3D7 genome.

Figure 6

Phylogenetic comparison of 3 dominant DBLα sequence tags transcribed from each clinical isolates and selected tags from 3D7. A neighbor-joining tree was generated using amplified DBL-1α fragments. The bootstrap consensus tree inferred from 1000 replicates [45], using pairwise deletion of amino acid sequences with p-distance. Phylogenetic analyses were conducted in MEGA4 [39]. Sequences transcribed by isolates from children with severe malaria (ISM, pink), asymptomatic malaria (IAM,black) and 3D7 genes (group A, red; group B, blue; group C, green).

Figure 7

Phylogenetic comparison of var group A from 3D7 genome and 3 dominant var group A amplified from clinical isolates. A neighbor-joining tree was generated based upon the predicted protein start site in the N-terminal segment (NTS) domain to the first DBL-1α H block. The bootstrap consensus tree inferred from 1000 replicates [44], using pairwise deletion of amino acid sequences with p-distance. Phylogenetic analyses were conducted in MEGA4 [39]. Sequences transcribed by isolates from children with severe malaria (ISM, blue), asymptomatic (AS, green).