Comparative transcriptomics of New Zealand Weta

Abstract

Weta from the Orthopteran family Anostostomatidae are an iconic component of the New Zealand fauna. Within this family there are approximately 56 species belonging to five genera: Hemideina (tree weta), Deinacrida (giant weta), Hemiandrus (ground weta) and Anisoura/Motuweta (tusked weta). A significant proportion of the 18 tree and giant weta species are classified as threatened. Despite widespread interest in these species, current genomic resources are lacking. This thesis aimed to address this issue by using next generation sequencing (Roche 454) to assemble and annotate reproductive transcriptomes (male testis and accessory gland tissues) from two tree weta species (H. thoracica and H. crassidens). The resulting databases could then be used to identify genes involved in reproduction and speciation, and to compare the patterns of evolution and selection acting on reproductive proteins with others that exhibit general patterns of expression. De novo assembly of transcripts resulted in putative gene sets of 41,884 and 25,523 nonredundant sequences for H. thoracica and H. crassidens, respectively. From these datasets a number of genes and conserved protein domains were identified, that illustrate the wide variety of diverse transcripts expressed within tree weta reproductive tissues. These datasets represent a significant increase in the current genomic resources for weta. Candidate genes encompassing both reproductive and general metabolic roles were chosen from the 454 data for further evolutionary analysis. Of these genes, 11 candidates were Sanger sequenced across 27 individuals from the two tree weta species and one giant weta species (D. mahoenui), with the aim of investigating the patterns of nucleotide substitution across the genes. These analyses revealed that among the reproductive candidates a mixture of adaptive evolution and functional constraint is occurring, whereas in comparison the general metabolic genes exhibit evolutionary constraint, with little or no nonsynonymous substitutions observed. From among these candidates, three genes showed evidence of selection, including a serine protease, a lectin-like gene and a contig of unknown function, and within them sites under positive selection.