Population genetics of Anopheles coluzzii immune pathways and genes

Abstract

Natural selection is expected to drive adaptive evolution in genes involved in host-pathogen interactions. In this study, we use molecular population genetic analyses to understand how natural selection operates on the immune system of Anopheles coluzzii (formerly A. gambiae "M form"). We analyzed patterns of intraspecific and interspecific genetic variation in 20 immune-related genes and 17 nonimmune genes from a wild population of A. coluzzii and asked if patterns of genetic variation in the immune genes are consistent with pathogen-driven selection shaping the evolution of defense. We found evidence of a balanced polymorphism in CTLMA2, which encodes a C-type lectin involved in regulation of the melanization response. The two CTLMA2 haplotypes, which are distinguished by fixed amino acid differences near the predicted peptide cleavage site, are also segregating in the sister species A. gambiae ("S form") and A. arabiensis. Comparison of the two haplotypes between species indicates that they were not shared among the species through introgression, but rather that they arose before the species divergence and have been adaptively maintained as a balanced polymorphism in all three species. We additionally found that STAT-B, a retroduplicate of STAT-A, shows strong evidence of adaptive evolution that is consistent with neofunctionalization after duplication. In contrast to the striking patterns of adaptive evolution observed in these Anopheles-specific immune genes, we found no evidence of adaptive evolution in the Toll and Imd innate immune pathways that are orthologously conserved throughout insects. Genes encoding the Imd pathway exhibit high rates of amino acid divergence between Anopheles species but also display elevated amino acid diversity that is consistent with relaxed purifying selection. These results indicate that adaptive coevolution between A. coluzzii and its pathogens is more likely to involve novel or lineage-specific molecular mechanisms than the canonical humoral immune pathways.

Keywords: C-type lectin; JAK-STAT; balancing selection; genetics of immunity; innate immunity; population genetics.

Figure 1

Distribution of Tajima’s D across all 37 genes. Tajima’s D was calculated for each gene using silent sites and plotted as a histogram showing deviation from neutral expectations. Loci are ordered based on the value of D to draw attention to the contrast between CTLMA2 and the rest of the dataset.

Figure 2

Sliding window analysis of Tajima’s D in CTL4 and CTLMA2. CTL4 and CTLMA2 are located directly adjacent on the chromosome and were PCR amplified and cloned as a single fragment. Tajima’s D was calculated using silent sites in a sliding window along the entire sequenced region using a 200-bp window with a 25-bp step size. The dashed line indicates the expected value of D under a neutral equilibrium model. The schematic below the plot shows the exon structure and direction of transcription for CTL4 and CTLMA2. Positive values of D in the 5′ region of CTLMA2 indicate an excess of intermediate frequency polymorphism in this region.

Figure 3

Amino acid alignment of CTLMA2 sequences of A. coluzzii collected in Burkina Faso, along with previously published sequences from A. gambiae and A. arabiensis collected in Kenya (Obbard et al. 2007). Sites matching the top sequence are indicated with a dot (•). Dashes (–) note unavailable sequence because we sequenced a different amplicon than was available for A. gambiae or A. arabiensis

Figure 4

Neighbor-joining tree of CTLMA2 sequences from A. coluzzii (blue circles), A. gambiae (black triangle), A. arabiensis (red squares), A. quadriannulatus (open circle), A. merus (open triangle), and A. christyi (open square). Bootstrap support was determined using 1000 bootstrap replicates, and nodes with >50% bootstrap support are shown.