Comparison of mass spectrometry data and bioinformatics predictions to assess the bona fide localization of proteins identified in cell wall proteomics studies.

ABSTRACT

Plant cell walls have complex architectures made of polysaccharides among which cellulose, hemicelluloses, pectins and cell wall proteins (CWPs). Some CWPs are anchored in the plasma membrane through a glycosylphosphatidylinositol (GPI)-anchor. The secretion pathway is the classical route to reach the extracellular space. Based on experimental data, a canonical signal peptide (SP) has been defined, and bioinformatics tools allowing the prediction of the sub-cellular localization of proteins have been designed. In the same way, the presence of GPI-anchor attachment sites can be predicted using bioinformatics programs. This article aims at comparing the bioinformatics predictions of the sub-cellular localization of proteins assumed to be CWPs to mass spectrometry (MS) data. The sub-cellular localization of a few CWPs exhibiting particular features has been checked by cell biology approaches. Although the prediction of SP length is confirmed in most cases, it is less conclusive for GPI-anchors. Three main observations were done (i) the variability observed at the N-terminus of a few mature CWPs could play a role in the regulation of their biological activity; (ii) one protein was shown to have a double sub-cellular localization in the cell wall and the chloroplasts; and (iii) peptides were found to be located at the C-terminus of several CWPs previously identified in GPI-anchored proteomes, thus raising the issue of their actual anchoring to the plasma membrane.