Transcriptome profiling reveals differential gene expression of detoxification enzymes in a hemimetabolous tobacco pest after feeding on jasmonate-silenced Nicotiana attenuata plants

Abstract

Background: The evolutionary arms race between plants and insects has driven the co-evolution of sophisticated defense mechanisms used by plants to deter herbivores and equally sophisticated strategies that enable phytophagous insects to rapidly detoxify the plant's defense metabolites. In this study, we identify the genetic determinants that enable the mirid, Tupiocoris notatus, to feed on its well-defended host plant, Nicotiana attenuata, an outstanding model for plant-insect interaction studies.

Results: We used an RNAseq approach to evaluate the global gene expression of T. notatus after feeding on a transgenic N. attenuata line which does not accumulate jasmonic acid (JA) after herbivory, and consequently accumulates very low levels of defense metabolites. Using Illumina sequencing, we generated a de novo assembled transcriptome which resulted in 63,062 contigs (putative transcript isoforms) contained in 42,610 isotigs (putative identified genes). Differential expression analysis based on RSEM-estimated transcript abundances identified 82 differentially expressed (DE) transcripts between T. notatus fed on wild-type and the defenseless plants. The same analysis conducted with Corset-estimated transcript abundances identified 59 DE clusters containing 85 transcripts. In both analyses, a larger number of DE transcripts were found down-regulated in mirids feeding on JA-silenced plants (around 70%). Among these down-regulated transcripts we identified seven transcripts possibly involved in the detoxification of N. attenuata defense metabolite, specifically, one glutathione-S-transferase (GST), one UDP-glucosyltransferase (UGT), five cytochrome P450 (P450s), and six serine proteases. Real-time quantitative PCR confirmed the down-regulation for six transcripts (encoding GST, UGT and four P450s) and revealed that their expression was only slightly decreased in mirids feeding on another N. attenuata transgenic line specifically silenced in the accumulation of diterpene glycosides, one of the many classes of JA-mediated defenses in N. attenuata.

Conclusions: The results provide a transcriptional overview of the changes in a specialist hemimetabolous insect associated with feeding on host plants depleted in chemical defenses. Overall, the analysis reveals that T. notatus responses to host plant defenses are narrow and engages P450 detoxification pathways. It further identifies candidate genes which can be tested in future experiments to understand their role in shaping the T. notatus-N. attenuata interaction.

Keywords: Coyote tobacco; Detoxification; Heteroptera; Piercing-sucking herbivory; Trophic interactions.

Fig. 1

Overview of the transformed plants used to dissect the mechanisms used by Tupiocoris notatus to tolerate toxic metabolites produced by its host plant, Nicotiana attenuata. Schematic of the signaling and biosynthetic pathways of jasmonate-induced defenses in N. attenuata. The enzymes silenced in the transgenic RNAi lines are highlighted in red font. Abbreviations: OPDA, 12-oxophytodienoic acid; AOC, allene oxide cyclase; JA, jasmonic acid, JA-Ile, jasmonic acid-isoleucine; GGPPS, geranyl diphosphate synthase; DTGs, diterpene glycosides; PI, proteinase inhibitors; TPIs, trypsin proteinase inhibitors

Fig. 2

Quality assessment of the two de novo assembled Tupiocoris notatus transcriptomes. Each transcriptome was assembled using the full set of HQ reads (ALL) or using two thirds of HQ reads (TWO). a Frequency distribution of the number of T. notatus contigs that hit a single Acyrthosiphon pisum sequence after Blastx; b Venn diagram showing the number of A. pisum sequences hit by at least one T. notatus contig from each transcriptome; c Overall distribution of ortholog hit ratio (OHR) calculated using Blastx annotation against the A. pisum proteome

Fig. 3

Annotation of Tupiocoris notatus transcriptome. a Species distribution of the top Blastx hit performed against NCBI nr database; b Venn diagram showing the number of orthologous genes shared between T. notatus, Acyrthosiphon pisum, Diaphorina citri and Drosophila melanogaster; c and d) Gene Onthology (GO) assignments as predicted by Blast2GO at GO level 3 and 2 for the categories Molecular Function and Biological Process

Fig. 4

Comparison between results obtained with Corset- or RSEM-estimated transcript abundances, and RT-qPCR. a Venn diagrams depicting the number of up- and down-regulated contigs in mirids feeding on JA-silenced Nicotiana attenuata plants identified by RNAseq analysis. Among the down-regulated genes, a list of differentially-expressed (DE) contigs annotated as detoxification enzymes or serine proteases (ser. prot) is shown; b) and c) Scatter plots showing the correlation between log2 fold-change (log2FC) estimated by RNAseq or by RT-qPCR for putatively unique transcripts (PUTs) likely involved in N. attenuata metabolite detoxification. Asterisks indicate contigs that have been identified as DE by RNAseq analysis using both Corset- or RSEM-estimated transcript abundances

Fig. 5

Expression of putatively unique transcripts (PUTs) involved in response to Nicotiana attenuata 's jasmonic acid-mediated defenses in Tupiocoris notatus feeding on different transgenic lines with reduced degrees of defenses. Expression was checked by RT-qPCR. Mirids were allowed to feed for 3 days on empty vector (EV) and transformed plants silenced in the expression of: JA biosynthesis (irAOC), trypsin proteinase inhibitors (irPI) and diterpene glycosides (irGGPPS). Statistical significance compared to control (EV) was analyzed by ANOVA followed by post-hoc Dunnett test, n = 3

Fig. 6

Phylogenetic relationships of Tupiocoris notatus, Acyrthosiphon pisum and Rhodnius prolixus P450s. Maximum-likelihood tree was built in Mega 6 based on MUSCLE amino acid alignment. Black dots indicate branch support >70 bootstrap. Black stars indicate T. notatus P450s down-regulated in mirids feeding on irAOC plants. Genes from A. pisum (black) were taken from [53] and genes from R. prolixus (green) from [54]. T. notatus contigs are in red

Fig. 7

Phylogenetic relationships of Tupiocoris notatus, Acyrthosiphon pisum and Rhodnius prolixus GSTs. Maximum-likelihood tree was built in Mega 6 based on MUSCLE amino acid alignment. Black dots indicate branch support >70 bootstrap. Black star indicates the T. notatus GST PUT down-regulated in mirids feeding on irAOC plants. Genes from A. pisum (black) were taken from [53] and genes from R. prolixus (green) from [54]. T. notatus contigs are in red

Fig. 8

Phylogenetic relationships of Tupiocoris notatus, Drosophila melanogaster and Bombyx mori UGTs. Maximum-likelihood tree was built in Mega 6 based on MUSCLE amino acid alignment. Black dots indicate branch support >70 bootstrap. Black star indicates the T. notatus UGT PUT down-regulated in mirids feeding on irAOC plants. Genes from B. mori (orange) were taken from [19] and genes from D. melanogaster (blue) from [55]. T. notatus contigs are in red