Characterization of plants expressing the human ß1,4-galactosyltrasferase gene.

Modification of the plant N-glycosylation pathway towards human type structures is an important strategy to implement plants as expression systems for therapeutic proteins. Nevertheless, relatively little is known about the overall impact of non-plant glycosylation enzymes in stable transformed plants. Here, we analyzed transgenic lines (Nicotiana benthamiana and Arabidopsis thaliana) that stably express a modified version of human ß1,4-galactosyltransferase ((ST)GalT). While some transgenic plants grew normally, other lines exhibited a severe phenotype associated with stunted growth and developmental retardation. The severity of the phenotype correlated with both increased (ST)GalT mRNA and protein levels but no differences were observed between N-glycosylation profiles of plants with and without the phenotype. In contrast to non-transgenic plants, all (ST)GalT expressing plants synthesized significant amounts of incompletely processed (largely depleted of core fucose) N-glycans with up to 40% terminally galactosylated structures. While transgenic plants showed no differences in nucleotide sugar composition and cell wall monosaccharide content, alterations in the reactivity of cell wall carbohydrate epitopes associated with arabinogalactan-proteins and pectic homogalacturonan were detected in (ST)GalT expressing plants. Notably, plants with phenotypic alterations showed increased levels of hydrogen peroxide, most probably a consequence of hypersensitive reactions. Our data demonstrate that unfavorable phenotypical modifications may occur upon stable in planta expression of non-native glycosyltransferases. Such important issues need to be taken into consideration in respect to stable glycan engineering in plants.

Post-translational regulation and trafficking of the granulin-containing protease RD21 of Arabidopsis thaliana.

RD21-like proteases are ubiquitous, plant-specific papain-like proteases typified by carrying a C-terminal granulin domain. RD21-like proteases are involved in immunity and associated with senescence and various types of biotic and abiotic stresses. Here, we interrogated Arabidopsis RD21 regulation and trafficking by site-directed mutagenesis, agroinfiltration, western blotting, protease activity profiling and protein degradation. Using an introduced N-glycan sensor, deglycosylation experiments and glyco-engineered N. benthamiana plants, we show that RD21 passes through the Golgi where it becomes fucosylated. Our studies demonstrate that RD21 is regulated at three post-translational levels. Prodomain removal is not blocked in the catalytic Cys mutant, indicating that RD21 is activated by a proteolytic cascade. However, RD21 activation in Arabidopsis does not require vacuolar processing enzymes (VPEs) or aleurain-like protease AALP. In contrast, granulin domain removal requires the catalytic Cys and His residues and is therefore autocatalytic. Furthermore, SDS can (re-)activate latent RD21 in Arabidopsis leaf extracts, indicating the existence of a third layer of post-translational regulation, possibly mediated by endogenous inhibitors. RD21 causes a dominant protease activity in Arabidopsis leaf extracts, responsible for SDS-induced proteome degradation.

Sialic acid concentrations in plants are in the range of inadvertent contamination.

The long held but challenged view that plants do not synthesize sialic acids was re-evaluated using two different procedures to isolate putative sialic acid containing material from plant tissues and cells. The extracts were reacted with 1,2-diamino-4,5-methylene dioxybenzene and the fluorescently labelled 2-keto sugar acids analysed by reversed phase and normal phase HPLC and by HPLC-electrospray tandem mass spectrometry. No N-glycolylneuraminic acid was found in the protein fraction from Arabidopsis thaliana MM2d cells. However, we did detect 3-deoxy-D: -manno-octulosonic acid and trace amounts (3-18 pmol/g fresh weight) of a compound indistinguishable from N-acetylneuraminic acid by its retention time and its mass spectral fragmentation pattern. Thus, plant cells and tissues contain five orders of magnitude less sialic acid than mammalian tissues such as porcine liver. Similar or lower amounts of N-acetylneuraminic acid were detected in tobacco cells, mung bean sprouts, apple and banana. Yet even yeast and buffer blanks, when subjected to the same isolation procedures, apparently contained the equivalent of 5 pmol of sialic acid per gram of material. Thus, we conclude that it is not possible to demonstrate unequivocally that plants synthesize sialic acids because the amounts of these sugars detected in plant cells and tissues are so small that they may originate from extraneous contaminants.