Secretory bulk flow of soluble proteins is efficient and COPII dependent.

COPII-coated vesicles, first identified in yeast and later characterized in mammalian cells, mediate protein export from the endoplasmic reticulum (ER) to the Golgi apparatus within the secretory pathway. In these organisms, the mechanism of vesicle formation is well understood, but the process of soluble cargo sorting has yet to be resolved. In plants, functional complements of the COPII-dependent protein traffic machinery were identified almost a decade ago, but the selectivity of the ER export process has been subject to considerable debate. To study the selectivity of COPII-dependent protein traffic in plants, we have developed an in vivo assay in which COPII vesicle transport is disrupted at two distinct steps in the pathway. First, overexpression of the Sar1p-specific guanosine nucleotide exchange factor Sec12p was shown to result in the titration of the GTPase Sar1p, which is essential for COPII-coated vesicle formation. A second method to disrupt COPII transport at a later step in the pathway was based on coexpression of a dominant negative mutant of Sar1p (H74L), which is thought to interfere with the uncoating and subsequent membrane fusion of the vesicles because of the lack of GTPase activity. A quantitative assay to measure ER export under these conditions was achieved using the natural secretory protein barley alpha-amylase and a modified version carrying an ER retention motif. Most importantly, the manipulation of COPII transport in vivo using either of the two approaches allowed us to demonstrate that export of the ER resident protein calreticulin or the bulk flow marker phosphinothricin acetyl transferase is COPII dependent and occurs at a much higher rate than estimated previously. We also show that the instability of these proteins in post-ER compartments prevents the detection of the true rate of bulk flow using a standard secretion assay. The differences between the data on COPII transport obtained from these in vivo experiments and in vitro experiments conducted previously using yeast components are discussed.

Extensive N-glycosylation reduces the thermal stability of a recombinant alkalophilic bacillus alpha-amylase produced in Pichia pastoris.

Alkalophilic Bacillus alpha-amylase (ABA) was produced in the yeast Pichia pastoris with a yield of 50 mg L(-1) of culture supernatant. The recombinant protein, rABA, was glycosylated at seven of the nine sites for potential N-glycosylation as identified by automated peptide sequencing and MALDI-TOF MS of tryptic fragments. The number of hexose units within each glycan chain was found to vary from 8 to 18 as calculated from the masses of glycosylated peptide fragments. Temperature stability measurements in the absence of substrate showed that the T(50) of glycosylated rABA and its endoglycosidase H-deglycosylated form was 76 degrees C while that of ABA purified from Bacillus was 89 degrees C thus demonstrating that the original temperature stability of ABA was not retained by rABA. The relative thermoperformance, i.e., the activity at 80 degrees C relative to that at 37 degrees C was 0.9 +/- 0.3 for rABA. Removal of all seven N-linked glycans by endoglycosidase H increased the relative thermoperformance to 2.4 +/- 0.6, compared to the value of 3.5 +/- 1.1 for ABA. Thus, removal of the N-linked glycans did not improve the thermostability of rABA but modified its thermoperformance to approach that of the original Bacillus enzyme. rABA had the highest activity around pH 6. Treatment of rABA with endoglycosidase H shifted the pH activity profile in a more alkaline direction approaching the pH activity profile of ABA.