Integrated analyses of the rice secretome

Abstract

The plant cell wall contains proteins secreted from the cell. A subset of these proteins is called the secretome, which plays important roles in biological and physiological processes. To gain insight into the secretome for monocot species, we performed proteomic analysis of the rice secretome. In this addendum, we combined the results of two independent studies. For this, 154 rice secreted proteins from our study were compared to 42 non-redundant rice secreted proteins from another group. Surprisingly, only 20 proteins were commonly found in the two groups, indicating that the materials and methods are very important factors finding secreted proteins. Finally, a total of 172 rice secreted proteins were assigned molecular functions and biological processes according to gene ontology annotation. These comparative and integrative analyses of the rice secretome provide a large number of rice secreted proteins that can be used in subsequent investigations.

Plant cells are surrounded by rigid cell walls that provide structural support, protection from pathogens, and maintenance of the cellular shape in response to various stresses.^1–3 The cell wall possesses a large number of secreted proteins that play critical roles in a range of cellular processes. Typically secreted proteins contain an N-terminal signal peptide that is cleaved during transport into the lumen of the endoplasmic reticulum (ER).⁴ Recently, several approaches have been carried out to extract secreted proteins from different materials while minimizing contamination, and these isolated proteins were then validated by bioinformatics tools for the secretory pathway.^5–10 A subset of those proteins that are secreted from the cell can be called the secretome.¹¹

Recently, two independent proteomic analyses of the rice secretome have been reported. One report identified a large number of secreted proteins from rice calli using multi-dimensional protein identification technology (MudPIT),¹² while the other employed systematic analyses of the rice secretome from leaves and calli using two-dimensional electrophoresis (2-DE)-based approaches.⁷ In this addendum, we provide integrated datasets and analysis from two recent publications to build a rich resource for the rice secretome.

Comparative Analysis of Two Rice Secretome Datasets

Recently, we identified 154 typical rice secreted proteins from 555 redundant proteins identified by MudPIT followed by several bioinformatics analyses.¹² In addition, 222 redundant secreted proteins from rice leaves and calli were identified using a 2-DE-based approach by another research group.⁷ However, the dataset of Jung et al.⁷ must be reprocessed in order to obtain typically secreted proteins. Although these authors demonstrated that the isolated secretome includes few or no intracellular proteins, it seems that contamination of intracellular proteins in the plant system was highly prevalent. For example, five chloroplast proteins, such as the RuBisCO large subunit-binding protein subunits a and b, were identified.⁷ Therefore, it is necessary to use several bioinformatics tools to validate such contaminants. The 222 redundant rice proteins were converted into their corresponding rice loci based on the rice genome annotation (http://rice.plantbiology.msu.edu/).¹³ This process identified 142 non-redundant rice proteins. Using the SignalP 3.0 program, only 44 proteins were predicted to have a signal peptide by two different algorithms (SignalP-NN and SignalP-HMM).⁴ The TMHMM v. 2.0 program was used to exclude transmembrane proteins, and two proteins (LOC_Os01g71350 and LOC_Os07g35560) were identified to posses three and two transmembrane domains, respectively.¹⁴ In total, 42 proteins were predicted to be pure rice secreted proteins (See Suppl. Table 1).

In Figure 1A, materials, methods, and the number of identified proteins for each experimental system are depicted. In comparison to the number of identified proteins between the two publications, MudPIT analysis led to the identification of 555 proteins whose number is 2.5 times greater than that of Jung et al. who found 222 protein spots on 2-DE gels (Fig. 1A). Finally, we present 154 typically secreted proteins whose number is three times greater than the 42 total proteins identified by Jung et al. (Fig. 1B). This result indicates the technical benefit of MudPIT analysis to detect a large number of proteins, including low abundance proteins, which are not visible on 2-DE gels.^15,16 While 2-DE gel and MudPIT are quite different technologies for the purposes of separating independent protein components, it is highly recommended to integrate both methods for increased protein coverage.^15,16

Figure 1.

The comparison of two independent data sets of the rice secretome. (A) Materials, experimental methods, and number of identified proteins from two different publications for the rice secretome are depicted. To obtain typical secreted rice proteins, several bioinformatics tools were applied. (B) A Venn diagram displays the number of typically secreted proteins from two different datasets. Cho et al.¹² identified a total of 154 proteins (a + c) and Jung et al.⁷ found a total of 42 proteins (b + c).

In the present study, we considered only proteins that were freely secreted into the extracellular space based on SignalP prediction. Recently, it has been suggested that non-classical secreted proteins can also be predicted by a prediction program, such as SecretomeP.¹⁷ However, this program is currently only available in bacteria and mammalians, but not in plants.¹⁸ A Venn diagram displays the number of secreted proteins from the two publications considered here (Fig. 1B). About half of the proteins from the study by Jung et al.⁷ overlapped in our protein list. Surprisingly, 22 rice secreted proteins from leaves could not be found in our study. This result suggests that the components of secreted proteins are highly dependent on plant materials, such as suspension cells, calli, and plant leaves as well as environmental and experimental conditions. Therefore, it is desirable to establish various experimental systems in order to determine the complete rice secretome.

Gene Ontology Analysis of the Rice Secretome

To get a comprehensive overview of how the rice secretome is associated with its various molecular functions and biological processes, we performed gene ontology (GO) annotation implemented with the Blast2GO analysis tool with default parameters.¹⁹ A total of 172 rice secreted proteins were blasted against NCBI's non-redundant protein database and their functions were annotated. During this process, enzyme codes (EC) and conserved domains of each protein were also analyzed (See Suppl. Tables 1–3).^20,21 These analyses found that 83% of the secreted proteins include one or more GO term assignments.

According to GO annotation, highly enriched molecular functions of the rice secretome were revealed as proteins with hydrolase activity (67 proteins) followed by 43 proteins with binding functions. Protein binding (22 proteins), catalytic activity (19 proteins), kinase activity (11 proteins), carbohydrate binding (7 proteins), lipid binding (7 proteins), receptor activity (7 proteins), and enzyme regulator activity (6 proteins) were also highly represented molecular functions (Fig. 2A). According to the biological processes of the rice secretome, 59 proteins with cellular processes, 42 proteins with carbohydrate metabolic processes, and 41 proteins with responses to stress were highly enriched in the rice secretome. Interestingly, electron transport (4 proteins), flower development (3 proteins) and transport (6 proteins) were revealed as biological processes of the rice secretome (Fig. 2B). Although there are still numerous unknown functions for these secreted proteins, the identified proteins can be grouped into three main functional classes: cell wall modification, structural maintenance and cellular defense. In this addendum, we will not discuss in detail the functions of those individual proteins, which were already well described in two previous studies.^7,12 Taken together, the large number of rice secreted proteins in the present study provides useful information for studying the functions of individual proteins.

Figure 2.

Highly presented GO assignments in the rice secretome. The identified rice secreted proteins were assigned according to their molecular function (A) and biological process (B). Non-redundant functions and processes are displayed. Bars indicate the number of proteins assigned in each function and process according to GO terms.

Addendum to: Cho WK, Chen XY, Chu H, Rim Y, Kim S, Kim ST, Kim SW, Park ZY, Kim JY. The proteomic analysis of the secretome of rice calli. Physiologia Plantarum. 2009 doi: 10.1111/j.1399-3054.2009.01198.x. In press.