关键词: 葡萄溃疡病, 可可毛色二孢, 生物信息学, 分泌蛋白, 信号肽, 表达模式

Abstract:

【Objective】Lasiodiplodia theobromae is an important phytopathogenic fungus with a worldwide distribution. This species causes severe Botryosphaeria dieback on a wide range of woody plants, which leads to reduced crop quality and tremendous economic losses. The objective of this study is to predict and analyze the candidate secreted proteins in the genome of L. theobromae, clarify their basic characteristics, so as to lay a foundation for the study of the pathogenic mechanism of secreted proteins in this pathogen.【Method】The signal peptide prediction algorithm SignalP v5.0 and subcellular localization prediction algorithm ProtComp v9.0, transmembrane helix prediction algorithm TMHMM v2.0, GPI-anchoring site prediction algorithm big-PI Fungal Predictor, and subcellular protein location distribution algorithm TargetP v2.0 were used to analyze 12 902 protein sequences of L. theobromae published in the previous study. The basic features including the length of the N-terminal signal peptide, the frequency of amino acid usage and cleavage site of the predicted secreted proteins were statistically analyzed. Based on the homology of the protein sequence, biological function annotation of the predicted secreted proteins was clarified by using BLASTP program. The activity of the signal peptide of the selected secreted proteins was detected by yeast secretion and cell translocation assays. Expression patterns of the selected secreted protein genes during L. theobromae infection were analyzed by qRT-PCR technology.【Result】In this study, 522 secreted proteins were verified, accounting for 4.3% of the total proteins present in the genome of L. theobromae. The lengths of amino acids of secreted proteins were ranged from 101 to 400 aa. The distribution length of signal peptides was from 18 to 20 aa and the largest number was 20 aa. The top frequent amino acid was alanine in the signal peptides, and the most frequently incorporated amino acids were non-polar and hydrophobic, accounting for 60.2% of the total amino acids. Further, the amino acids in the position -3 to -1 in the signal peptides were relatively conserved and the signal peptide cleavage site belonged to A-X-A type, which could be recognized and cleaved by Sp I type peptidase. Among them, 336 secreted proteins were identified with a predictive function, which was mostly enzymatic or virulence-associated protein. Besides, there are differences in terms of molecular weight, isoelectric point, the aliphatic index in the candidate secreted proteins. Finally, the predicted signal peptides of the 9 putative L. theobromae secreted proteins were confirmed to have secretory activity by using a yeast invertase secretion assay. qRT-PCR analysis demonstrated that the expression of selected protein genes was differentially regulated during host infection.【Conclusion】A total of 552 candidate secreted proteins of L. theobromae were predicted by a set of computer algorithms. Lengths of the signal peptides vary greatly and the most frequently are mainly non-polar and hydrophobic amino acids. Secreted proteins characterized in this study can be categorized under enzymes related to the degradation of cell wall components, necrosis induction proteins, and chitin-binding proteins which may play an important role in L. theobromae pathogenetic mechanism.

Key words: Botryosphaeria dieback, Lasiodiplodia theobromae, bioinformatics, secreted protein, signal peptide, expression pattern