Comparative Genomics Reveals the Origins and Diversity of Arthropod Immune Systems

Abstract

Insects are an important model for the study of innate immune systems, but remarkably little is known about the immune system of other arthropod groups despite their importance as disease vectors, pests, and components of biological diversity. Using comparative genomics, we have characterized the immune system of all the major groups of arthropods beyond insects for the first time--studying five chelicerates, a myriapod, and a crustacean. We found clear traces of an ancient origin of innate immunity, with some arthropods having Toll-like receptors and C3-complement factors that are more closely related in sequence or structure to vertebrates than other arthropods. Across the arthropods some components of the immune system, such as the Toll signaling pathway, are highly conserved. However, there is also remarkable diversity. The chelicerates apparently lack the Imd signaling pathway and beta-1,3 glucan binding proteins--a key class of pathogen recognition receptors. Many genes have large copy number variation across species, and this may sometimes be accompanied by changes in function. For example, we find that peptidoglycan recognition proteins have frequently lost their catalytic activity and switch between secreted and intracellular forms. We also find that there has been widespread and extensive duplication of the cellular immune receptor Dscam (Down syndrome cell adhesion molecule), which may be an alternative way to generate the high diversity produced by alternative splicing in insects. In the antiviral short interfering RNAi pathway Argonaute 2 evolves rapidly and is frequently duplicated, with a highly variable copy number. Our results provide a detailed analysis of the immune systems of several important groups of animals for the first time and lay the foundations for functional work on these groups.

Keywords: evolution; genomics; innate immunity.

Fig. 1.

Arthropod TLRs. (A) Phylogenetic tree of TLRs from seven species of arthropods, four chordates (human, mouse, chicken, and Ciona), and the nematode Caenorhabditis. The tree was reconstructed using the maximum-likelihood method from the TIR domains and is midpoint rooted. Myriapod and crustacean taxon labels are orange, chelicerates blue, and Drosophila black. Scale bar is substitutions per site. Bootstrap support values can be found in supplementary figure S1, Supplementary Material online. (B) The domain structure of arthropod TLRs. TLRs in the yellow vertebrate clade of panel (A) are shown in the black box. Red bars are LRRs, blue diamonds are cysteine clusters, magenta the TIR domain, and the gray line represents the plasma membrane. Domain locations are all to scale.

Fig. 2.

Presence or absence of Toll pathway members across the arthropods. Dashed gray symbols represent proteins where no homolog was detected. Only key domains are shown (Ank: ankyrin repeat). Topology of the phylogeny is taken from Sharma et al. (2014), although it should be noted that relationships within the chelicerata (Mesobuthus, Parasteatoda, Tetranychus, Metaseiulus, and Ixodes) remain poorly resolved. Divergence dates are in millions of years before present (Ma) and from Rota-Stabelli et al. (2013).

Fig. 3.

Presence or absence of Imd pathway members across the arthropods. Dashed gray symbols represent genes where no homolog was detected. Only key domains are shown (ANK: ankyrin repeats).

Fig. 4.

Gene tree, copy number, cellular location, and predicted catalytic activity of arthropod PGRPs. (A) PGRP tree was reconstructed using maximum likelihood from the PGRP domain sequences, and is midpoint rooted. Myriapod taxon labels are orange, chelicerates blue, and Drosophila black. Scale bar is substitutions per site. Bootstrap support values can be found in supplementary figure S2, Supplementary Material online. (B) Scale drawing and predicted cellular location of PGRPs. The PGRP domain is shown in lime green, signal peptides in white. Predicted catalytic PGRPs are denoted by an*.

Fig. 5.

(A) BGRPs and β-1,3-glucanases have variable copy number in the Mandibulata but are absent from the chelicerates. The tree is midpoint rooted and was reconstructed by maximum likelihood. Node labels are bootstrap support from 1000 replicates. Scale bar is substitutions per site. (B) Alignment of BGRPs and glucanases identified in this study together with those from Drosophila melanogaster (Waterhouse et al. 2007) and a midgut β-1,3 glucanase from Bombyx mori (NP_001159614.1). Two conserved Glu active site residues found in all insect glucanases are labeled E188 and E193 after their position in NP_001159614.1 following Hughes (2012). Proteins that have lost one or both Glu residues, and so are expected to have lost glucanase activity, are labeled with an asterisk in both (A) and (B). The complete alignment may be found in the supplementary material, Supplementary Material online.

Fig. 6.

Gene tree of the TEP family. Sequences include arthropod TEPs, MCR proteins, the vertebrate C3, C4 and C5 complement factors, and alpha-2 macroglobulins. Genes without a thioester motif are shown with an “*.” Myriapod and crustacean taxon labels are orange, chelicerates blue, and Drosophila black. Taxa in gray are all deuterostomes with the exception of the nematode Caenorhabditis, which is a protostome related to arthropods, and the coral Swiftia, which diverged before the split of vertebrates and arthropods. In addition to the chelicerate sequences we annotated, we included two arthropod sequences from horseshoe crabs (Limulus and Carcinoscorpius) and a sequence from the tick Ornithodoros. The tree is midpoint rooted and was reconstructed using the maximum-likelihood method. Scale bar is substitutions per site. The additional taxa are taken from the previous analyses of Zhu et al. (2005), Wu et al. (2012), and Sekiguchi et al. (2012). Bootstrap support values can be found in supplementary figure S3, Supplementary Material online.

Fig. 7.

The diversity of Dscam in the arthropods. The tree is reconstructed using the maximum-likelihood method based on a complete Dscam amino acid sequence alignment and is rooted using the mollusk Aplysia californica. Myriapod and crustacean taxon labels are orange, chelicerates blue, and Drosophila black. Scale bar is substitutions per site. Bootstrap support values can be found in supplementary figure S4, Supplementary Material online.

Fig. 8.

Proteins in the miRNA and antiviral siRNAi pathways. (A) Argonaute-family Argonaute proteins. (B) Dicer proteins. The trees are reconstructed by maximum likelihood and rooted following Schnettler et al. (2014). Node labels are bootstrap support from 1,000 replicates. Scale bar is substitutions per site.