Glycogen phosphorylase, the product of the glgP Gene, catalyzes glycogen breakdown by removing glucose units from the nonreducing ends in Escherichia coli

Abstract

To understand the biological function of bacterial glycogen phosphorylase (GlgP), we have produced and characterized Escherichia coli cells with null or altered glgP expression. glgP deletion mutants (DeltaglgP) totally lacked glycogen phosphorylase activity, indicating that all the enzymatic activity is dependent upon the glgP product. Moderate increases of glycogen phosphorylase activity were accompanied by marked reductions of the intracellular glycogen levels in cells cultured in the presence of glucose. In turn, both glycogen content and rates of glycogen accumulation in DeltaglgP cells were severalfold higher than those of wild-type cells. These defects correlated with the presence of longer external chains in the polysaccharide accumulated by DeltaglgP cells. The overall results thus show that GlgP catalyzes glycogen breakdown and affects glycogen structure by removing glucose units from the polysaccharide outer chains in E. coli.

FIG. 1.

GlgP catalyzes E. coli glycogen breakdown in vivo. (A) Iodine staining of (1) WT, (2) ΔglgP, and (3) ΔglgCAP cells. Bacteria were cultured overnight in solid Kornberg medium supplemented with 50 mM glucose. (B) Time course analyses of the glycogen content in WT (▵), ΔglgP (□), and ΔglgCAP (○) cells cultured in liquid Kornberg medium supplemented with 50 mM glucose. The bacterial strain employed in this study was B/r. Essentially similar results were obtained with other bacterial strains.

FIG. 2.

Glycogen accumulation in GlgP-overproducing cells and ΔglgP and ΔglgCAP E. coli mutants. Cells cultured in liquid Kornberg medium supplemented with 50 mM glucose were analyzed by EM. Upper left quadrant, WT cells transformed with the plasmid vector (pET-15b); upper right quadrant, GlgP-overproducing WT cells (pET-glgP); lower left quadrant, ΔglgCAP mutants; lower right quadrant, ΔglgP mutants. Arrows indicate the position of electron-transparent cytoplasmic glycogen granules resembling “holes” preferentially located at the periphery of E. coli cells (17). The bacterial strain employed in these studies was BL21(DE3)C43. Essentially similar results were obtained with other bacterial strains. Bars, 1 μm.

FIG. 3.

Increasing glycogen phosphorylase activity leads to a marked reduction of glycogen levels in E. coli. (A) Time course analyses of the glycogen content in cells transformed with either pET-15b (▵) or pET-glgP (○) cultured in liquid M9 minimal medium supplemented with 50 mM glucose. (B) Glycogen phosphorylase activity-dependent glycogen content in cells transformed with pBAD-glgP cultured in liquid M9 minimal medium supplemented with 50 mM glucose. glgP expression was regulated by adding the arabinose concentrations indicated in Fig. S4 in the supplemental material. At the end of the exponential growth phase, cells were harvested and both glycogen content and glycogen phosphorylase activities were measured. The bacterial strain employed in this study was BL21(DE3)C43. Essentially similar results were obtained with other bacterial strains.

FIG. 4.

Chain length distribution of glycogens, determined by FACE (see Materials and Methods). Isoamylase debranched glycogens from both WT and ΔglgP cells were produced as described in Materials and Methods and separated up to DP 30. (A) CLDs expressed on a DP scale (x axis) are displayed as the percentage of total chains, expressed on a molar basis (y scale). (B) Difference plot generated by subtracting the mol% value of the WT at each chain length from the corresponding mol% of the mutant at the corresponding chain length. Dark bars correspond to WT cells, and light bars correspond to ΔglgP cells. The bacterial strain used in this study was TG1. Essentially similar results were obtained with other bacterial strains.

FIG. 5.

CLD of glycogen β-limit dextrins determined by FACE. Isoamylase debranched glycogen β-limit dextrins from both WT and ΔglgP cells were produced as described in Materials and Methods and separated up to DP 20. Chains below DP 5 are not represented because of the very high background due to the presence of hyperabundant chains of DP 2 and 3 generated by the action of β-amylase. Dark bars correspond to WT cells, and light bars correspond to ΔglgP cells. The bacterial strain used in this study was TG1. Essentially similar results were obtained with other bacterial strains.