Large Scale Profiling of Protein Isoforms Using Label-Free Quantitative Proteomics Revealed the Regulation of Nonsense-Mediated Decay in Moso Bamboo (Phyllostachys edulis)

Abstract

Moso bamboo is an important forest species with a variety of ecological, economic, and cultural values. However, the gene annotation information of moso bamboo is only based on the transcriptome sequencing, lacking the evidence of proteome. The lignification and fiber in moso bamboo leads to a difficulty in the extraction of protein using conventional methods, which seriously hinders research on the proteomics of moso bamboo. The purpose of this study is to establish efficient methods for extracting the total proteins from moso bamboo for following mass spectrometry-based quantitative proteome identification. Here, we have successfully established a set of efficient methods for extracting total proteins of moso bamboo followed by mass spectrometry-based label-free quantitative proteome identification, which further improved the protein annotation of moso bamboo genes. In this study, 10,376 predicted coding genes were confirmed by quantitative proteomics, accounting for 35.8% of all annotated protein-coding genes. Proteome analysis also revealed the protein-coding potential of 1015 predicted long noncoding RNA (lncRNA), accounting for 51.03% of annotated lncRNAs. Thus, mass spectrometry-based proteomics provides a reliable method for gene annotation. Especially, quantitative proteomics revealed the translation patterns of proteins in moso bamboo. In addition, the 3284 transcript isoforms from 2663 genes identified by Pacific BioSciences (PacBio) single-molecule real-time long-read isoform sequencing (Iso-Seq) was confirmed on the protein level by mass spectrometry. Furthermore, domain analysis of mass spectrometry-identified proteins encoded in the same genomic locus revealed variations in domain composition pointing towards a functional diversification of protein isoform. Finally, we found that part transcripts targeted by nonsense-mediated mRNA decay (NMD) could also be translated into proteins. In summary, proteomic analysis in this study improves the proteomics-assisted genome annotation of moso bamboo and is valuable to the large-scale research of functional genomics in moso bamboo. In summary, this study provided a theoretical basis and technical support for directional gene function analysis at the proteomics level in moso bamboo.

Keywords: Phyllostachys edulis; alternative splicing; label-free quantitative proteomics; long noncoding RNA; nonsense-mediated mRNA decay; protein isoform.

Figure 1

The flowchart of high-throughput label-free quantitative proteomics in moso bamboo. (A) The fully developed leaves of 30-days-old seedlings and 60-days-old moso bamboo were collected to detect the concentration and efficiency of protein extraction among different periods; (B) The pipeline of label-free quantitative proteomics; (C) Venn diagram of moso bamboo proteins and the mRNA transcription; (D) The distribution of molecular weight of moso bamboo proteins; (E) GO functional annotation of total proteins.

Figure 2

Profiles of proteins in seedlings and leaves of moso bamboo. (A) Comparison of protein numbers in seedlings and leaves of moso bamboo; (B) The classification of enzymes in seedlings and leaves of moso bamboo; (C) Pathway analysis, according to the Kyoto Encyclopedia of Genes and Genomes (KEGG) database.

Figure 3

The gene ontology (GO) enrichment analysis of differential expression proteins in seedlings and leaves of moso bamboo based on mass spectrometry-based label-free quantitative proteomics. (A) The correlation analysis of protein abundance between seedlings and leaves of moso bamboo; (B) The biological process for down-regulated proteins; (C) The biological process for up-regulated proteins.

Figure 4

Revision of gene annotation in moso bamboo and discovery of new genes. Current annotation and translation frame (black frames) and revised gene models of PH01150953G0010 (A) and PH0100300G0250 (B) supported by matching peptide spectra. Stop codon was marked by an asterisk.

Figure 5

The re-annotation of lncRNA through mass spectrometry data. (A) Overview of pipeline for lncRNA re-annotation. All lncRNAs were translated into amino acid sequences. Selected Amino acid (aa) sequences (> 100 aa) were compared to mass spectrometry data and sequences with. more than 2 peptide-matches were considered to be authentic; (B) Venn diagram illustrating the number of lncRNA re-annotated by mass spectrometry; (C) The biological process for lncRNA re-annotated by mass spectrometry; (D) The molecular function for lncRNA re-annotated by mass spectrometry; (E) The cellular component for lncRNA re-annotated by mass spectrometry. Explanatory information on the functional enrichments and numbers of involved proteins in terms were all listed on the left. The cutoff of the p value is set at 0.05. Different colors (such as red, green and yellow) stand for –log10 (p value) in the GO analysis; (F) The plot of differentially expressed lncRNA encoded proteins.

Figure 6

Different protein isoforms from the same locus were identified by mass spectrometry data. (A) The pipeline for the verification of different protein isoforms from the same locus for moso bamboo through protein profiling data; (B) The Venn diagram illustrating the number of different protein isoforms verified by mass spectrometry in all AS/APA isoform; (C) Domain analysis of different protein isoforms from the same locus.