We have identified and disrupted the gene coding for alpha-glucosidase II in Dictyostelium discoideum. This enzyme is responsible for removing two alpha 1,3-linked glucose residues from N-linked oligosaccharides on newly synthesized glycoproteins. Mutagenesis by restriction enzyme-mediated integration (REMI) generated a clone, DG1033, which grows well but forms abnormal fruiting bodies with short, thick stalks. The strain lacks alpha-glucosidase II activity and makes incompletely processed N-linked oligosaccharides that are abnormally large and have fewer sulfate and phosphate esters. The morphological, enzymatic, and oligosaccharide profile phenotypes of the disruption mutant are all recapitulated by a targeted disruption of the normal gene. Furthermore, all of these defects are corrected in cells transformed with a normal, full-length copy of the gene. The phenotypic characteristics of DG1033 as well as chromosomal mapping of the disrupted gene indicate that it is the site of the previously characterized modA mutation. The Dictyostelium gene is highly homologous to alpha-glucosidase II genes in the human and the pig, C. elegans, and yeast. Although various cell lines have been reported to be defective in alpha-glucosidase II activity, disruption of the Dictyostelium gene gives the first example of a clear developmental phenotype associated with loss of this enzyme.

Two cAMP-binding proteins, cbp1 and cbp2, were purified from the cytoplasm of the green alga Volvox carteri. Both proteins have a native molecular mass of 90 kDa as determined by gel filtration. cbp2 was purified to apparent electrophoretic homogeneity, having a subunit molecular mass of 42 kDa as determined by SDS/PAGE. The cbp1 preparation contains a 42-kDa and a 44-kDa band. The cAMP-binding activity is not associated with protein kinase activity. Tryptic peptides of cbp2 were sequenced by automated Edman degradation. Two pairs of peptides differ in one amino acid only, thus pointing to the presence of isoforms of cbp2. Both binding proteins differed from the cAMP-specific phosphodiesterases of V. carteri with respect to charge, molecular mass and binding affinity to N6-cAMP-agarose. Reverse-phase chromatography of the bound ligand revealed that the two binding proteins hydrolyse cAMP to 5' AMP. The binding specificity of purified cbp1 and cbp2 was probed by a set of modified cAMP derivatives. Both proteins bind cAMP strictly specifically in the anti conformation; position 1 and 6 of the adenine moiety and at least one of the exocyclic O atoms of the ribose cyclic phosphate moiety are essential. 3-Isobutyl-1-methylxanthine is an effective inhibitor of binding but the natural methylxyanthines are not. At present it is not clear whether cbp1 and cbp2 are individual proteins or isoforms of one another.

Many Dictyostelium lysosomal enzymes contain mannose-6-phosphate (Man-6-P) in their N-linked oligosaccharide chains. We have now characterized a new group of lysosomal proteins that contain N-acetylglucosamine-1-phosphate (GlcNAc-1-P) linked to serine residues. GlcNAc-1-P-containing proteins, which include papain-like cysteine proteinases, cofractionate with the lysosomal markers and are in functional vesicles of the endosomal/lysosomal pathway. Immunoblots probed with reagents specific for each carbohydrate modification indicate that the lysosomal proteins are modified either by Man-6-P or GlcNAc-1-P, but not by both. Confocal microscopy shows that the two sets of proteins reside in physically and functionally distinct compartments. Vesicles with GlcNAc-1-P fuse with nascent bacteria-loaded phagosomes less than 3 minutes after ingestion, while those with Man-6-P do not participate in bacterial digestion until about 15 minutes after phagocytosis. Even though both types of vesicles fuse with phagosomes, GlcNAc-1-P- and Man-6-P-bearing proteins rarely colocalize. Since both lysosomal enzymes and their bound carbohydrate modifications are stable in lysosomes, a targeting or retrieval mechanism based on these carbohydrate modifications probably establishes and/or maintains segregation.

No abstract