Transcriptome analysis of Gossypium hirsutum flower buds infested by cotton boll weevil (Anthonomus grandis) larvae

Abstract

Background: Cotton is a major fibre crop grown worldwide that suffers extensive damage from chewing insects, including the cotton boll weevil larvae (Anthonomus grandis). Transcriptome analysis was performed to understand the molecular interactions between Gossypium hirsutum L. and cotton boll weevil larvae. The Illumina HiSeq 2000 platform was used to sequence the transcriptome of cotton flower buds infested with boll weevil larvae.

Results: The analysis generated a total of 327,489,418 sequence reads that were aligned to the G. hirsutum reference transcriptome. The total number of expressed genes was over 21,697 per sample with an average length of 1,063 bp. The DEGseq analysis identified 443 differentially expressed genes (DEG) in cotton flower buds infected with boll weevil larvae. Among them, 402 (90.7%) were up-regulated, 41 (9.3%) were down-regulated and 432 (97.5%) were identified as orthologues of A. thaliana genes using Blastx. Mapman analysis of DEG indicated that many genes were involved in the biotic stress response spanning a range of functions, from a gene encoding a receptor-like kinase to genes involved in triggering defensive responses such as MAPK, transcription factors (WRKY and ERF) and signalling by ethylene (ET) and jasmonic acid (JA) hormones. Furthermore, the spatial expression pattern of 32 of the genes responsive to boll weevil larvae feeding was determined by "in situ" qPCR analysis from RNA isolated from two flower structures, the stamen and the carpel, by laser microdissection (LMD).

Conclusion: A large number of cotton transcripts were significantly altered upon infestation by larvae. Among the changes in gene expression, we highlighted the transcription of receptors/sensors that recognise chitin or insect oral secretions; the altered regulation of transcripts encoding enzymes related to kinase cascades, transcription factors, Ca2+ influxes, and reactive oxygen species; and the modulation of transcripts encoding enzymes from phytohormone signalling pathways. These data will aid in the selection of target genes to genetically engineer cotton to control the cotton boll weevil.

Figure 1

Distribution of differentially expressed genes (DEGs) (x-axis) into Gene Ontology (GO) categories (biological process) (y-axis) according to Gene Set Enrichment Analysis (GSEA). Only biological processes (BPs) discussed in the results are presented here. A complete list of BPs can be found in Additional file 6.

Figure 2

Representative overview of DEG involved in biotic stress response in Gossypium hirsutum 48 h after infection with cotton boll weevil larvae. Log2-fold change of gene expression (cotton boll weevil compared to mock-inoculated control) was analysed by MapMan software. Yellow squares represent up-regulated genes and blue squares represent down-regulated genes. The colour saturation indicates log fold change > 4 and < 4. The figure shows that the molecular recognition of pathogens and herbivores by plants to trigger a defence response requires initial recognition including the following: 1. Microbe-, pathogen- and damage-associated molecular patterns (MAMP, PAMP and DAMP) are recognised by pattern recognition receptors (PRR) and lead to PAMP-triggered immunity (PTI). 2. Oviposition-associated compounds are recognised by unknown receptors and trigger defensive responses. 3. Putative herbivore-associated molecular patterns (HAMP) are recognised by receptors and lead to herbivore-triggered immunity (HTI). 4. Wounding by herbivores leads to the release of DAMP and to wound-induced resistance (WIR).

Figure 3

Phylogenetic tree of WRKY domains between cotton and Arabidopsis . The amino acid sequences of the WRKY domain of cotton and Arabidopsis were aligned with MUSCLE, and the phylogenetic tree was constructed using the JTT model with an estimated γ-distribution parameter (G). The maximum-likelihood analyses were performed with the program PhyML version 3.0, and assessment of node confidence was performed using 1,000 bootstrap replicates. The members of group I, II (a-e) and III are labelled according to the classifications of AtWRKY domains by Eulgem et al. [33]. The triangles indicate cotton WRKY genes, and filled triangles represent the genes analysed by qPCR. Contig9787 was named GhWRKY70-like1; was named contig5359 GhWRKY64-like1; and contig16334 was named GhWRKY72-like1 after calculating the p-distance to determine the closest relationship with Arabidopsis members.

Figure 4

Isolation of cotton tissues from paraffin-embedded sections by laser microdissection (LMD). Isolation of cotton tissues from paraffin-embedded sections by laser microdissection (LMD). Sections before LMD (a, c, e), and sections after LM (b, d, f). The area selected for laser microdissection is outlined in green (a region near the damage caused by larvae feeding, which comprised the stamen tissue, viewed at a, c and d) or blue (a region farther from the injured area, which comprised the carpel tissue, viewed at a, e, and f). The assessment of extracted RNA integrity from the stamen and carpel are shown in g and h, respectively. Electropherograms were obtained with an Agilent 2100 Bioanalyser. Open and closed arrowheads indicate the 18S and 28S ribosomal RNA peaks, respectively. RNA quality is expressed as the RNA integrity number (RIN). Scale bars = 100 μm.

Figure 5

Comparison of expression levels by “ in situ ” qPCR of a subset of 32 DEG. These genes were examined from two different areas (stamen and carpel) isolated from 6 mm cotton flower buds infested by cotton boll weevil larvae by laser microdissection (LMD) in relation to the control. The reference genes GhACT4 and GhFBX6 were used to normalise the qPCR data. The relative expression level was calculated using the relative expression software tool (REST^©), and a subsequent statistical test of the analysed C_P values by a Pair-Wise Fixed Reallocation Randomization Test was performed.

Figure 6

Venn diagram comparing the cotton boll weevil ( Anthonomus grandis ) induced transcriptome with the response to another tissue-chewing pest, Plutella xylostella . The number of induced (up-regulated, left column) and repressed (down-regulated, right column) genes after 48 h of cotton boll weevil feeding (A. grandis) was compared to the response of another 24 h herbivory-treatment previously published [39]. Selected genes have a p-value ≤ 0.05 and |logFC| ≥ 2.0.