Amylose-defective mutants were selected after UV mutagenesis of Chlamydomonas reinhardtii cells. Two recessive nuclear alleles of the ST-2 gene led to the disappearance not only of amylose but also of a fraction of the amylopectin. Granule-bound starch synthase activities were markedly reduced in strains carrying either st-2-1 or st-2-2, as is the case for amylose-deficient (waxy) endosperm mutants of higher plants. The main 76-kDa protein associated with the starch granule was either missing or greatly diminished in both mutants, while st-2-1-carrying strains displayed a novel 56-kDa major protein. Methylation and nuclear magnetic resonance analysis of wild-type algal storage polysaccharide revealed a structure identical to that of higher-plant starch, while amylose-defective mutants retained a modified amylopectin fraction. We thus propose that the waxy gene product conditions not only the synthesis of amylose from endosperm storage tissue in higher-plant amyloplasts but also that of amylose and a fraction of amylopectin in all starch-accumulating plastids. The nature of the ST-2 (waxy) gene product with respect to the granule-bound starch synthase activities is discussed.Our knowledge of starch synthesis and degradation, while having developed mostly from observations made in higherplant storage tissues, has benefited from investigations performed on a number of model microbial systems. Chlorella pyrenoidosa, for instance, has been the subject of several of the pioneer studies dealing with the enzymology of starch anabolism (16,20). Prokaryotic organisms such as Escherichia coli have yielded a number of relevant genetic and biochemical studies, mostly because of the parallel that can be drawn between the regulation of plant and bacterial ADP-pyrophosphorylases (reviewed in reference 18). However, the very nature of the storage polysaccharide (glycogen) has prevented the use of model bacterial systems to investigate the biogenesis of the starch granule itself. Thus, our knowledge of the intricate pathways of amylose and amylopectin biosyntheses stems solely from those elegant genetic investigations performed on pea or cereal mutations which express themselves only in the endosperm (reviewed in references 15 and 22). Even in that case, we do not yet know precisely which of the starch synthases and starch branching enzymes are responsible for amylose or amylopectin biosynthesis, let alone how they act coordinately to produce a complex structure such as the starch granule. One of the best and most studied endosperm mutants, waxy maize, seemed until very recently to shed at least some light on the biosynthesis of amylose. Waxy mutations have been identified in most cereals (22, 32) and more recently in storage tissues of dicots such as potato (9) or in the perisperm of the amaranth (11). They all lead to the decrease or absence of both amylose and granule-bound starch synthase. While there is no doubt that one of the main proteins associated with the starch granule is the product of the waxy * Corresponding author...

ADP-glucose synthesis through ADP-glucose pyrophosphorylase defines the major rate-controlling step of storage polysaccharide synthesis in both bacteria and plants. We have isolated mutant strains defective in the STA6 locus of the monocellular green alga Chlamydomonas reinhardtii that fail to accumulate starch and lack ADP-glucose pyrophosphorylase activity. We show that this locus encodes a 514-amino-acid polypeptide corresponding to a mature 50-kDa protein with homology to vascular plant ADP-glucose pyrophosphorylase small-subunit sequences. This gene segregates independently from the previously characterized STA1 locus that encodes the large 53-kDa subunit of the same heterotetramer enzyme. Because STA1 locus mutants have retained an AGPase but exhibit lower sensitivity to 3-phosphoglyceric acid activation, we suggest that the small and large subunits of the enzyme define, respectively, the catalytic and regulatory subunits of AGPase in unicellular green algae. We provide preliminary evidence that both the small-subunit mRNA abundance and enzyme activity, and therefore also starch metabolism, may be controlled by the circadian clock.

It has been generally assumed that the [alpha]-(1->4)-linked and [alpha]-(1->6)-branched glucans of starch are generated by the coordinated action of elongation (starch synthases) and branching enzymes. We have identified a novel Chlamydomonas locus (STA7) that when defective leads to a wipeout of starch and its replacement by a small amount of glycogen-like material. Our efforts to understand the enzymological basis of this phenotype have led us to determine the selective disappearance of an 88-kD starch hydrolytic activity. We further demonstrate that this enzyme is a debranching enzyme. Cleavage of the [alpha]-(1->6) linkage in a branched precursor of amylopectin (preamylopectin) has provided us with the ground rules for understanding starch biosynthesis in plants. Therefore, we propose that amylopectin clusters are synthesized by a discontinuous mechanism involving a highly specific glucan trimming mechanism.

A low-starch mutant accumulating less than 5% of wild-type amounts was isolated after X-ray mutagenesis of Chlamydomonas reinhardtii cells. The recessive st-1-1 defect segregated as a single mendelian mutation through meiosis, and led to a severe decrease in starch accumulation under all culture conditions tested, whether in the light or in darkness. Adenosine 5'-diphosphoglucose pyrophosphorylase (in the absence of 3-phosphoglycerate), starch synthase, phosphoglucomutase, phosphorylase and starch-branching enzyme were all characterized and shown to be unaffected by the mutation. However, ADP-glucose pyrophosphorylase in the mutant had its sensitivity to activation by 3-phosphoglycerate lowered dramatically and became less responsive to orthophosphate. Our results are consistent both with a mutation in a structural gene of a multisubunit enzyme or in a regulatory gene responsible for switching ADP-glucose pyrophosphorylase from a 3-phosphoglycerate-insensitive to a 3-phosphoglycerate-sensitive form. These results provide definite proof of the in-vivo requirement for 3-phosphoglycerate activation to obtain substantial starch synthesis in plants. The conclusions hold both for synthesis from CO2 in the light or from exogenous organic carbon sources in darkness. A model is presented in which the existence of a 3-phosphoglycerate gradient explains localized starch synthesis around the pyrenoid of lower plants.

It has been generally assumed that the [alpha]-(1->4)-linked and [alpha]-(1->6)-branched glucans of starch are generated by the coordinated action of elongation (starch synthases) and branching enzymes. We have identified a novel Chlamydomonas locus (STA7) that when defective leads to a wipeout of starch and its replacement by a small amount of glycogen-like material. Our efforts to understand the enzymological basis of this phenotype have led us to determine the selective disappearance of an 88-kD starch hydrolytic activity. We further demonstrate that this enzyme is a debranching enzyme. Cleavage of the [alpha]-(1->6) linkage in a branched precursor of amylopectin (preamylopectin) has provided us with the ground rules for understanding starch biosynthesis in plants. Therefore, we propose that amylopectin clusters are synthesized by a discontinuous mechanism involving a highly specific glucan trimming mechanism.