The genome of an underwater architect, the caddisfly Stenopsyche tienmushanensis Hwang (Insecta: Trichoptera)

Abstract

Background: Caddisflies (Insecta: Trichoptera) are a highly adapted freshwater group of insects split from a common ancestor with Lepidoptera. They are the most diverse (>16,000 species) of the strictly aquatic insect orders and are widely employed as bio-indicators in water quality assessment and monitoring. Among the numerous adaptations to aquatic habitats, caddisfly larvae use silk and materials from the environment (e.g., stones, sticks, leaf matter) to build composite structures such as fixed retreats and portable cases. Understanding how caddisflies have adapted to aquatic habitats will help explain the evolution and subsequent diversification of the group.

Findings: We sequenced a retreat-builder caddisfly Stenopsyche tienmushanensis Hwang and assembled a high-quality genome from both Illumina and Pacific Biosciences (PacBio) sequencing. In total, 601.2 M Illumina reads (90.2 Gb) and 16.9 M PacBio subreads (89.0 Gb) were generated. The 451.5 Mb assembled genome has a contig N50 of 1.29 M, has a longest contig of 4.76 Mb, and covers 97.65% of the 1,658 insect single-copy genes as assessed by Benchmarking Universal Single-Copy Orthologs. The genome comprises 36.76% repetitive elements. A total of 14,672 predicted protein-coding genes were identified. The genome revealed gene expansions in specific groups of the cytochrome P450 family and olfactory binding proteins, suggesting potential genomic features associated with pollutant tolerance and mate finding. In addition, the complete gene complex of the highly repetitive H-fibroin, the major protein component of caddisfly larval silk, was assembled.

Conclusions: We report the draft genome of Stenopsyche tienmushanensis, the highest-quality caddisfly genome so far. The genome information will be an important resource for the study of caddisflies and may shed light on the evolution of aquatic insects.

Figure 1:

An illustration of the adult caddisfly Stenopsyche tienmushanensis in its typical habitat.

Figure 2:

Functional gene annotations using four databases.

Figure 3:

The phylogenetic tree and gene expansion/contraction of 12 arthropod taxa. Multiple-copy orthologs represent the gene groups present in all species with a gene number >1 in at least one species. Species-specific paralogs represent genes uniquely present in only one species. Other types of orthologs represent the gene groups that are absent in some species and not species-specific paralogs. Numbers of expanded gene families are marked in green, while numbers of contracted gene families are marked in red. MRCA: most recent common ancestor. The number below MRCA is the total group number from the OrthoMCL analysis. Note that only some of the gene expansions/contractions are significant.

Figure 4:

The phylogenetic relationship of the significantly expanded gene groups of cytochrome P450 family in 10 insect species. The phylogeny was constructed using maximum likelihood, showing significant expansions in S. tienmushanensis. The bootstrap values are marked on the nodes.

Figure 5:

The maximum likelihood tree of odorant-binding proteins (OBPs) in five insect species. The bootstrap values are marked on the nodes. The expanded OBP groups in S. tienmushanensis are most closely related to those potentially responsible for pheromone detection in Drosophila.

Figure 6:

The H-fibroin gene complex in S. tienmushanensis. The sequences of H-fibroin gene fragments previously reported from S. marmorata are referred from [7, 12]. (a) The comparison of H-fibroin genes between S. tienmushanensis and S. marmorata. The depth of PacBio read coverage is shown in the line plot (smoothed by a sliding window average of 25 bps). The H-fibroin alignment of one representative tandem repetitive unit, non-repetitive 5′ end, and non-repetitive 3′ end between S. tienmushanensis and S. marmorata was shown in panels (b-d). Identical amino acids in alignment between S. tienmushanensis and S. marmorata were marked in gray shadow. The start and end positions of the nucleotides were shown in the alignment of the repetitive units. Amino acids in the black box represent the typical motifs of short repeat unit. S.tie: H-fibroin gene complex in S. tienmushanensis; S.mar5/S.mar3: the 5′/3′ end nucleotides of H-fibroin mRNA fragments in S. marmorata. The marked intron near the 5′ end of the gene complex (position: 43–124) was inferred from the alignment between the non-repetitive 5′ end between S. tienmushanensis and S. marmorata, positioned between sequences coding for the 14th and 15th amino acids of the N-terminus of the first predicted protein. The other marked intron (position: 10 643–10 729) was identified near the 5′ end of the second predicted gene, positioned between the second and third position in the codon for the 14th amino acid of the second predicted protein.