Isolation and structural analysis of a peptide containing the novel tyrosyl-glucose linkage in glycogenin.

The glucosylation site on glycogenin, the protein primer required for de novo glycogen synthesis, has been identified. The glucose is attached at position C1 in a glycosidic linkage with a unique tyrosine, and the sequence surrounding this residue was found to be: His-Leu-Pro-Phe-Ile-Tyr-Asn-Leu-Ser-Ser-Ile-Ser-Ile-Tyr(Glc)-Ser-Tyr-Leu -Pro- Ala-Phe-Lys. The same tyrosine residue is glycosylated whether glycogenin is isolated as a complex with the catalytic subunit of glycogen synthase, or covalently attached to glycogen. The possibility that insulin and growth factors may enhance glycogen synthesis via stimulation of the priming reaction is discussed.     

The discovery of glycogenin and the priming mechanism for glycogen biogenesis

The biogenesis of glycogen in skeletal muscle requires a priming mechanism that has recently been elucidated. The first step is catalysed by a protein tyrosine glucosyltransferase and involves the formation of a novel glycosidic linkage, namely the covalent attachment of glucose to a single tyrosine residue (Tyr194) on a priming protein, termed glycogenin. The next stage is the extension of the glucan chain from Tyr194 and involves the sequential addition of up to seven further glucosyl residues. This reaction is brought about autocatalytically by glycogenin itself, which is a Mn2+/Mg(2+)-dependent UDP-Glc-requiring glucosyltransferase. The glucan primer is elongated by glycogen synthase, but only when glycogenin and glycogen synthase are complexed together. Glycogen synthase dissociates from glycogenin during the synthesis of a glycogen molecule, enabling glycogen molecules to reach their maximum theoretical size. Each mature glycogen beta particle in muscle contains one molecule of glycogenin attached covalently, and an average one glycogen synthase catalytic subunit bound non-covalently. As evidence accumulates that a priming protein may be a fundamental property of polysaccharide synthesis in general, the molecular details of mammalian glycogen biogenesis may serve as a useful model for other systems.     

Synergistic phosphorylation of rabbit muscle glycogen synthase by cyclic AMP-dependent protein kinase and casein kinase I. Implications for hormonal regulation of glycogen synthase

The phosphorylation of rabbit skeletal muscle glycogen synthase by casein kinase I is markedly enhanced if the enzyme has previously been phosphorylated by cAMP-dependent protein kinase. The presence of phosphate in the primary cAMP-dependent protein kinase sites, sites 1a, 1b, and 2 (serine 7), increases the activity of casein kinase I toward residues in the vicinity of these sites. This synergistic phosphorylation correlates with potent inactivation of the glycogen synthase. Analysis of the NH2 terminus of the enzyme subunit indicated that phosphorylation at serine 7 caused serine 10 to become a preferred casein kinase I site and that phosphoserine can be an important recognition determinant for casein kinase I. This finding can also explain how epinephrine stimulation of skeletal muscle provokes significant increases in the phosphorylation state of serine residues, in particular serine 10, not recognized by cAMP-dependent protein kinase.     

Sites phosphorylated in bovine cardiac troponin T and I. Characterization by 31P-NMR spectroscopy and phosphorylation by protein kinases

Bovine cardiac troponin isolated in a highly phosphorylated form shows four 31P-NMR signals [Beier, N., Jaquet, K., Schnackerz, K. & Heilmeyer, L.M.G. Jr (1988) Eur. J. Biochem. 176, 327-334]. Troponin I, which contains phosphate covalently linked to serine-23 and/or -24 [Swiderek, K., Jaquet, K., Meyer, H. E. & Heilmeyer, L. M. G. Jr (1988) Eur. J. Biochem. 176, 335-342], shows three resonances. Mg2(+)-saturation of holotroponin shifts these troponin I resonances to higher fields. Direct binding of Mg2+ to the phosphate groups can be excluded. Both these serine residues of troponin I, 23 and 24, are substrates for cAMP- and cGMP-dependent protein kinases as well as for protein kinase C. Isolated bovine cardiac troponin T contains 1.5 mol phosphoserine/mol protein, indicating that minimally two serine residues are phosphorylated. One phosphoserine residue is located at the N-terminus. An additional phosphoserine is located in the C-terminal cyanogen bromide fragment, CN4, which contains covalently bound phosphate. Protein kinase C phosphorylates serine-194, thus demonstrating exposure of this residue on the surface of holotoponin.     

Glycogen synthesis in the absence of glycogenin in the yeast Saccharomyces cerevisiae

In eukaryotic cells, glycogenin is a self-glucosylating protein that primes glycogen synthesis. In yeast, the loss of function of GLG1 and GLG2, which encode glycogenin, normally leads to the inability of cells to synthesize glycogen. In this report, we show that a small fraction of colonies from glg1glg2 mutants can switch on glycogen synthesis to levels comparable to wild-type strain. The occurrence of glycogen positive glg1glg2 colonies is strongly enhanced by the presence of a hyperactive glycogen synthase and increased even more upon deletion of TPS1. In all cases, this phenotype is reversible, indicating the stochastic nature of this synthesis, which is furthermore illustrated by colour-sectoring of colonies upon iodine-staining. Altogether, these data suggest that glycogen synthesis in the absence of glycogenin relies on a combination of several factors, including an activated glycogen synthase and as yet unknown alternative primers whose synthesis and/or distribution may be controlled by TPS1 or under epigenetic silencing.     

Effects of phosphorylation of avian retrovirus nucleocapsid protein pp12 on binding of viral RNA

The major nucleocapsid protein of avian retroviruses, pp12, preferentially binds to the single-stranded regions of 60 S viral RNA with a apparent binding constant (Kapp) of 1.2 X 10(11) M-1. If the phosphate associated with serine residues of pp12 is hydrolyzed by either alkali treatment or with partially purified phosphoprotein phosphatase activities isolated from virions, the Kapp for binding to 60 S RNA decreases 100-fold. The high affinity binding of pp12 to viral RNA can be restored by phosphorylation of the protein with a protein kinase, protease-activated kinase I. The same serine residues phosphorylated in vivo are phosphorylated by protease kinase I in vitro. These residues have been identified as serine residues 40 and either 76 or 77. The protein purified from virions is phosphorylated primarily at serine residue 40 (greater than 90%). This suggests that phosphoserine residue 40 is responsible for modulating the binding of the protein to RNA. Thus, the phosphorylation state of pp12 can be reversibly altered in vitro resulting in the interconversion of the protein between a state of high and low affinity for single-stranded viral RNA.     

Glycogenin-dependent organization of Ascaris suum muscle glycogen

In Ascaris suum, muscle glycogen is synthesized during host feeding intervals and degraded during nonfeeding intervals. Glycogen accumulation is up to 12-fold greater than that observed in mammalian muscle. Previous studies have established that many aspects of the parasite glycogen metabolism are comparable with the host, but a novel form of glycogen synthase designated GSII also occurs in the parasite. In this report glycogenin has been identified as the core protein in both mature glycogen and the GSII complex. Digestion of GSII complex glycogen generates discreet intermediates that may correspond to a proglycogen pool, whereas digestion of mature glycogen does not generate these intermediates. Because both GSII complex glycogen and mature glycogen serve as GSII substrates, the GSII complex likely represents an intermediate between glycogenin and mature glycogen. The regulation of glycogenin synthesis or the regulation of GSII activity that converts glycogenin to proglycogen, or both, may account for high levels of polysaccharide accumulation that are essential for A. suum survival.     

Structural and functional studies on rabbit liver glycogenin

Glycogenin, the protein primer required for the biogenesis of muscle glycogen, has been isolated from rabbit liver glycogen. The protein comprised 0.0025% of liver glycogen by mass, 200-fold lower than the glycogenin content of muscle glycogen. Structural analyses, including determination of the amino acid sequence surrounding the glucosylated-tyrosine residue, showed identity with muscle glycogenin. Catalytically active liver glycogenin was partially purified and, like the skeletal muscle protein, catalysed an intramolecular, Mn2+- and UDP-Glc-dependent autoglucosylation reaction, forming a primer on which glycogen synthase could act. The results demonstrate that hepatic and muscle glycogenins are almost certainly identical proteins and that liver and skeletal muscle share a common mechanism for the biogenesis of glycogen molecules. The results also indicate that there is about one glycogenin molecule/liver glycogen alpha particle.     

Analysis of the in vivo phosphorylation state of protein phosphatase inhibitor-2 from rabbit skeletal muscle by fast-atom bombardment mass spectrometry

A new procedure has been developed for identifying phosphoserine residues in proteins, and is used to analyse the in vivo phosphorylation state of inhibitor-2. The method employs reverse-phase liquid chromatography to resolve phosphorylated and dephosphorylated forms of peptides and fast-atom bombardment mass spectrometry (FABMS) to identify phosphorylated derivatives. The positions of phosphorylation sites within peptides are located by gas-phase sequencer analysis after conversion of phosphoserine residues to S-ethylcysteine. The phosphorylation sites on inhibitor-2 were identified as serines-86, -120 and -121, the three residues phosphorylated in vitro by casein kinase-II. Serine-86 was phosphorylated to 0.7 mol/mol and serines-120 and -121 each to 0.3 mol/mol. These values were not altered significantly by intravenous injection of adrenalin or insulin. No phosphate was present in the region comprising residues 1-49, even after injection of adrenalin, demonstrating that inhibitor-2 is not a substrate for cyclic AMP-dependent protein kinase in vivo. The absence of phosphotyrosine also indicated that inhibitor-2 is not a physiological substrate for the insulin receptor. Surprisingly, no phosphate was present at threonine-72, the residue phosphorylated in vitro by glycogen synthase kinase-3, after injection of either propranolol, adrenalin or insulin. The implications of this finding for the in vivo activation of protein phosphatase 1I (the 1:1 complex between inhibitor-2 and the catalytic subunit of protein phosphatase-1) are discussed. FABMS analysis of inhibitor-2 confirmed the accuracy of the primary structure reported previously, and showed that the only post-translational modifications were an N-acetyl moiety and the three phosphoserine residues. FABMS also demonstrated the presence of an additional serine residue at the C-terminus, and showed that 50% of isolated inhibitor-2 molecules lack the C-terminal Ser-Ser dipeptide.     

Rabbit skeletal muscle glycogenin. Molecular cloning and production of fully functional protein in Escherichia coli.

Glycogenin is a self-glucosylating protein involved in the initiation reactions of glycogen synthesis. Initiation occurs in two stages, requiring first the covalent attachment of a glucose residue to Tyr-194 of glycogenin and then elongation to form an oligosaccharide chain. The latter reaction is known to be catalyzed by glycogenin itself. The glycogenin sequence determined from the protein by Campbell and Cohen (Campbell, D. G., and Cohen, P. (1989) Eur. J. Biochem. 185, 119-125) was used to design oligonucleotide probes to screen a rabbit muscle lambda gt11 library. A cDNA was isolated that predicted an amino acid sequence identical to that of Campbell and Cohen, except that Cys residues replaced Ser-88 and Leu-97. Northern analysis indicated a strongly hybridizing message of 1.8 kilobases, present in most tissues including skeletal muscle, but much weaker in kidney and scarcely detectable in liver. A much weaker 3-kilobase message was also detected in muscle. Polymerase chain reaction was used to isolate DNA fragments encoding a portion of glycogenin from rat and cow. The sequence of this segment was > 90% identical at the amino acid level across the three species, indicating that glycogenin is a highly conserved protein. Using the pET-8c vector, the glycogenin protein was expressed in Escherichia coli. Incubation of the recombinant glycogenin with UDP-[14C]glucose and Mn2+ resulted in labeling of the glycogenin protein, indicating that the recombinant glycogenin was enzymatically active and capable of self-glucosylation. Furthermore, after incubation with UDP-glucose, the recombinant glycogenin could serve as a substrate for glycogen synthase, leading to the production of high M(r) polysaccharide. Therefore, production of functional glycogenin did not require the intervention of any other mammalian protein.     

A self-glucosylating protein is the primer for rabbit muscle glycogen biosynthesis

In this paper we elucidate part of the mechanism of the early stages of the biosynthesis of glycogen. This macromolecule is constructed by covalent apposition of glucose units to a protein, glycogenin, which remains covalently attached to the mature glycogen molecule. We have now isolated, in a 3500-fold purification, a protein from rabbit muscle that has the same Mr as glycogenin, is immunologically similar, and proves to be a self-glucosylating protein (SGP). When incubated with UDP-[14C]glucose, an average of one molecular proportion of glucose is incorporated into the protein, which we conclude is the same as glycogenin isolated from native glycogen. The native SGP appears to exist as a high-molecular-weight species that contains many identical subunits. Because the glucose that is self-incorporated can be released almost completely from the acceptor by glycogenolytic enzymes, the indication is that it was added to a preformed chain or chains of 1,4-linked alpha-glucose residues. This implies that SGP already carries an existing maltosaccharide chain or chains to which the glucose is added, rather than glucose being added directly to protein. The putative role of SGP in glycogen synthesis is confirmed by the fact that glucosylated SGP acts as a primer for glycogen synthase and branching enzyme to form high-molecular-weight material. SGP itself is completely free from glycogen synthase. The quantity of SGP in muscle is calculated to be about one-half the amount of glycogenin bound in glycogen.     

Platelet glycoprotein Ib beta is phosphorylated on serine 166 by cyclic AMP-dependent protein kinase

Platelet responses are inhibited by agents such as prostaglandin E1 that increase the cytoplasmic concentration of cyclic AMP. Inhibition is thought to result from phosphorylation of specific proteins. One protein that becomes phosphorylated is glycoprotein (GP) Ib beta, a component of the GP Ib.IX complex. We have suggested that phosphorylation of GP Ib beta inhibits the collagen-induced polymerization of actin. The aim of the present study was to identify the amino acid(s) in GP Ib beta that is phosphorylated. Purified GP Ib.IX complex was phosphorylated by the catalytic subunit of purified bovine cyclic AMP-dependent protein kinase in the presence of [gamma-32P]ATP. Phosphoamino acid analysis showed that in GP Ib beta, [32P]phosphate was incorporated only into serine and was in a single tryptic peptide. Amino acid sequencing showed that this peptide was from the cytoplasmic domain of GP Ib beta and encompassed residues 161-175. A single serine residue, serine 166, contained the radiolabel. To determine whether the same residue was phosphorylated in intact platelets, GP Ib beta was isolated from 32P-labeled platelets before or after their exposure to prostaglandin E1. In both cases, radiolabel was present in phosphoserine and was in a single tryptic peptide. This peptide was the same as that which was phosphorylated in the purified GP Ib.IX complex, as shown by its identical mobility on two-dimensional tryptic maps, the presence of a positively charged residue in the fourth position, and the presence of the radiolabel in the sixth position of the peptide. This study shows that when cyclic AMP concentrations rise in platelets, the cytoplasmic domain of GP Ib beta is phosphorylated on serine 166, probably by cyclic AMP-dependent protein kinase. We suggest that phosphorylation of this residue may contribute to the inhibitory actions of cyclic AMP by inhibiting collagen-induced polymerization of actin.     

Proglycogen: a low-molecular-weight form of muscle glycogen.

We recently reported that muscle contains a trichloroacetic acid-precipitable component having Mr approx. 400 kDa that can be glucosylated by an endogenous enzyme acting on UDPglucose. This component contains within itself the autocatalytic, self-glucosylating protein glycogenin, the primer for glycogen synthesis. We now report that this substance, to which we give the name proglycogen, is a glycogen-like molecule constituting about 15% of total glycogen. It acts as a very efficient acceptor of glucose residues added from UDPglucose. Further, that the endogenous enzyme that adds the glucose to proglycogen is not the autocatalytic protein but a glycogen synthase-like enzyme. Proglycogen may be an intermediate in the synthesis and degradation of macromolecular glycogen and may exist and be metabolized as a separate entity. Consideration should now be given to the revival of the concept that tissue contains two forms of glycogen. One is proglycogen. The other is the 'classical', macromolecular glycogen. Additionally, proglycogen and glycogen may be glucosylated by different forms of synthase.     

Glycogenin activity in human skeletal muscle is proportional to muscle glycogen concentration

The de novo biosynthesis of glycogen is catalyzed by glycogenin, a self-glucosylating protein primer. To date, the role of glycogenin in regulating glycogen metabolism and the attainment of maximal glycogen levels in skeletal muscle are unknown. We measured glycogenin activity after enzymatic removal of glucose by alpha-amylase, an indirect measure of glycogenin amount. Seven male subjects performed an exercise and dietary protocol that resulted in one high-carbohydrate leg (HL) and one low-carbohydrate leg (LL) before testing. Resting muscle biopsies were obtained and analyzed for total glycogen, proglycogen (PG), macroglycogen (MG), and glycogenin activity. Results showed differences (P < 0.05) between HL and LL for total glycogen (438.0 +/- 69.5 vs. 305.7 +/- 57.4 mmol glucosyl units/kg dry wt) and PG (311.4 +/- 38.1 vs. 227.3 +/- 33.1 mmol glucosyl units/kg dry wt). A positive correlation between total muscle glycogen content and glycogenin activity (r = 0.84, P < 0.001) was observed. Similar positive correlations (P < 0.05) were also evident between both PG and MG concentration and glycogenin activity (PG, r = 0.82; MG, r = 0.84). It can be concluded that glycogenin does display activity in human skeletal muscle and is proportional to glycogen concentration. Thus it must be considered as a potential regulator of glycogen synthesis in human skeletal muscle.     

The structure of phosphorylated GSK-3beta complexed with a peptide, FRATtide, that inhibits beta-catenin phosphorylation.

BACKGROUND: Glycogen synthase kinase-3 (GSK-3) sequentially phosphorylates four serine residues on glycogen synthase (GS), in the sequence SxxxSxxxSxxx-SxxxS(p), by recognizing and phosphorylating the first serine in the sequence motif SxxxS(P) (where S(p) represents a phosphoserine). FRATtide (a peptide derived from a GSK-3 binding protein) binds to GSK-3 and blocks GSK-3 from interacting with Axin. This inhibits the Axin-dependent phosphorylation of beta-catenin by GSK-3. RESULTS: Structures of uncomplexed Tyr216 phosphorylated GSK-3beta and of its complex with a peptide and a sulfate ion both show the activation loop adopting a conformation similar to that in the phosphorylated and active forms of the related kinases CDK2 and ERK2. The sulfate ion, adjacent to Val214 on the activation loop, represents the binding site for the phosphoserine residue on 'primed' substrates. The peptide FRATtide forms a helix-turn-helix motif in binding to the C-terminal lobe of the kinase domain; the FRATtide binding site is close to, but does not obstruct, the substrate binding channel of GSK-3. FRATtide (and FRAT1) does not inhibit the activity of GSK-3 toward GS. CONCLUSIONS: The Axin binding site on GSK-3 presumably overlaps with that for FRATtide; its proximity to the active site explains how Axin may act as a scaffold protein promoting beta-catenin phosphorylation. Tyrosine 216 phosphorylation can induce an active conformation in the activation loop. Pre-phosphorylated substrate peptides can be modeled into the active site of the enzyme, with the P1 residue occupying a pocket partially formed by phosphotyrosine 216 and the P4 phosphoserine occupying the 'primed' binding site.     

GNIP, a novel protein that binds and activates glycogenin, the self-glucosylating initiator of glycogen biosynthesis

Glycogenin is a self-glucosylating protein involved in the initiation of glycogen biosynthesis. Self-glucosylation leads to the formation of an oligosaccharide chain, which, when long enough, supports the action of glycogen synthase to elongate it and form a mature glycogen molecule. To identify possible regulators of glycogenin, the yeast two-hybrid strategy was employed. By using rabbit skeletal muscle glycogenin as a bait, cDNAs encoding three different proteins were isolated from the human skeletal muscle cDNA library. Two of the cDNAs encoded glycogenin and glycogen synthase, respectively, proteins known to be interactors. The third cDNA encoded a polypeptide of unknown function and was designated GNIP (glycogenin interacting protein). Northern blot analysis revealed that GNIP mRNA is highly expressed in skeletal muscle. The gene for GNIP generates at least four isoforms by alternative splicing. The largest isoform GNIP1 contains, from NH(2)- to COOH-terminal, a RING finger, a B box, a putative coiled-coil region, and a B30.2-like motif. The previously identified protein TRIM7 (tripartite motif containing protein 7) is also derived from the GNIP gene and is composed of the RING finger, B box, and coiled-coil regions. The GNIP2 and GNIP3 isoforms consist of the coiled-coil region and B30.2-like domain. Physical interaction between GNIP2 and glycogenin was confirmed by co-immunoprecipitation, and in addition GNIP2 was shown to stimulate glycogenin self-glucosylation 3-4-fold. GNIPs may represent a novel participant in the initiation of glycogen synthesis.     

Multisite phosphorylation of the glycogen-binding subunit of protein phosphatase-1G by cyclic AMP-dependent protein kinase and glycogen synthase kinase-3

The glycogen-binding (G) subunit of protein phosphatase-1G is phosphorylated stoichiometrically by glycogen synthase kinase-3 (GSK3), and with a greater catalytic efficiency than glycogen synthase, but only after prior phosphorylation by cyclic AMP-dependent protein kinase (A-kinase) at site 1. The residues phosphorylated are the first two serines in the sequence AIFKPGFSPQPSRRGS-, while the C-terminal serine (site 1) is one of the two residues phosphorylated by A-kinase. These findings demonstrate that (i) the G subunit undergoes multisite phosphorylation in vitro; (ii) phosphorylation by GSK3 requires the presence of a C-terminal phosphoserine residue; (iii) GSK3 can synergise with protein kinases other than casein kinase-2.     

A 42,000-Da protein in rabbit tissues and in a glycogen synthase preparation cross-reacts with antibodies to glycogenin

Rabbit skeletal muscle glycogen previously has been shown to be covalently bound to a 40,000-Da protein ("glycogenin") via a novel glucosyl-tyrosine linkage [I.R. Rodriguez and W.J. Whelan (1985) Biochem. Biophys. Res. Commun. 132, 829-836]. Antibodies raised against rabbit skeletal muscle glycogenin cross-react with a similar protein present in rabbit heart and liver glycogens, as well as with a 42,000-Da "acceptor protein" present in high-speed supernatants of rabbit muscle, heart, retina, and liver. This 42,000-Da protein incorporates [U-14C]Glc when an ammonium sulfate fraction prepared from the tissue supernatants is incubated with UDP-[U-14C]Glc. The [U-14C]Glc incorporated can be removed quantitatively by treatment with amylolytic enzymes, indicating that the [U-14C]Glc incorporation represents elongation of a preexisting glucan attached to the acceptor protein. Furthermore, a commercial preparation of rabbit skeletal muscle glycogen synthase contains this 42,000-Da protein. We propose that the 42,000-Da protein represents the free form of glycogenin in tissues, with its covalently attached glucan chain(s) providing a "primed" elongation site for glycogen synthesis.     

No limiting role for glycogenin in determining maximal attainable glycogen levels in rat skeletal muscle

We examined whether the protein level and/or activity of glycogenin, the protein core upon which glycogen is synthesized, is limiting for maximal attainable glycogen levels in rat skeletal muscle. Glycogenin activity was 27.5 +/- 1.4, 34.7 +/- 1.7, and 39.7 +/- 1.3 mU/mg protein in white gastrocnemius, red gastrocnemius, and soleus muscles, respectively. A similar fiber type dependency of glycogenin protein levels was seen. Neither glycogenin protein level nor the activity of glycogenin correlated with previously determined maximal attainable glycogen levels, which were 69.3 +/- 5.8, 137.4 +/- 10.1, and 80.0 +/- 5.4 micromol/g wet wt in white gastrocnemius, red gastrocnemius, and soleus muscles, respectively. In additional experiments, rats were exercise trained by swimming, which resulted in a significant increase in the maximal attainable glycogen levels in soleus muscles ( approximately 25%). This increase in maximal glycogen levels was not accompanied by an increase in glycogenin protein level or activity. Furthermore, even in the presence of very high glycogen levels ( approximately 170 micromol/g wet wt), approximately 30% of the total glycogen pool continued to be present as unsaturated glycogen molecules (proglycogen). Therefore, it is concluded that glycogenin plays no limiting role for maximal attainable glycogen levels in rat skeletal muscle.     

GNN is a self-glucosylating protein involved in the initiation step of glycogen biosynthesis in Neurospora crassa

The initiation of glycogen synthesis requires the protein glycogenin, which incorporates glucose residues through a self-glucosylation reaction, and then acts as substrate for chain elongation by glycogen synthase and branching enzyme. Numerous sequences of glycogenin-like proteins are available in the databases but the enzymes from mammalian skeletal muscle and from Saccharomyces cerevisiae are the best characterized. We report the isolation of a cDNA from the fungus Neurospora crassa, which encodes a protein, GNN, which has properties characteristic of glycogenin. The protein is one of the largest glycogenins but shares several conserved domains common to other family members. Recombinant GNN produced in Escherichia coli was able to incorporate glucose in a self-glucosylation reaction, to trans-glucosylate exogenous substrates, and to act as substrate for chain elongation by glycogen synthase. Recombinant protein was sensitive to C-terminal proteolysis, leading to stable species of around 31kDa, which maintained all functional properties. The role of GNN as an initiator of glycogen metabolism was confirmed by its ability to complement the glycogen deficiency of a S. cerevisiae strain (glg1 glg2) lacking glycogenin and unable to accumulate glycogen. Disruption of the gnn gene of N. crassa by repeat induced point mutation (RIP) resulted in a strain that was unable to synthesize glycogen, even though the glycogen synthase activity was unchanged. Northern blot analysis showed that the gnn gene was induced during vegetative growth and was repressed upon carbon starvation.