Induced Structural Disorder as a Molecular Mechanism for Enzyme Dysfunction in Phosphoglucomutase 1 Deficiency

Abstract

Human phosphoglucomutase 1 (PGM1) plays a central role in cellular glucose homeostasis, mediating the switch between glycolysis and gluconeogenesis through the conversion of glucose 1-phosphate and glucose 6-phosphate. Recent clinical studies have identified mutations in this enzyme as the cause of PGM1 deficiency, an inborn error of metabolism classified as both a glycogen storage disease and a congenital disorder of glycosylation. Reported here are the first crystal structures of two disease-related missense variants of PGM1, along with the structure of the wild-type enzyme. Two independent glycine-to-arginine substitutions (G121R and G291R), both affecting key active site loops of PGM1, are found to induce regions of structural disorder, as evidenced by a nearly complete loss of electron density for as many as 23 aa. The disordered regions are not contiguous in sequence to the site of mutation, and even cross domain boundaries. Other structural rearrangements include changes in the conformations of loops and side chains, some of which occur nearly 20 Å away from the site of mutation. The induced structural disorder is correlated with increased sensitivity to proteolysis and lower-resolution diffraction, particularly for the G291R variant. Examination of the multi-domain effects of these G➔R mutations establishes a correlation between interdomain interfaces of the enzyme and missense variants of PGM1 associated with disease. These crystal structures provide the first insights into the structural basis of enzyme dysfunction in PGM1 deficiency and highlight a growing role for biophysical characterization of proteins in the field of precision medicine.

Fig. 1

A schematic of the catalytic reaction of PGM1, showing the reversible conversion of glucose 6-phosphate to glucose 1-phosphate. The bisphosphorylated intermediate undergoes a 180° reorientation (indicated by blue arrow) in the active site prior to the second phosphoryl transfer.

Fig. 2

The crystal structure of wild-type human PGM1. (a) A ribbon diagram of PGM1 colored by domain. Domain 1 (residues 1–191) is shown in blue, domain 2 (residues 192–304) in green, domain 3 (residues 305–421) in gold, and domain 4 (residues 422–562) in pink. Phosphoserine 117 is highlighted with dots, and Mg²⁺ is shown as an orange sphere. Glycines 121 and 291 are highlighted in magenta. For clarity, the three metal-binding aspartates (residues 288, 290 and 292) are not labeled. Missing residues in the D4 acNve site loop are shown with a doPed line. N- and C-termini are shown with red lePers. (b) A close-up view of the acNve site WT PGM1 showing the four regions (i-iv) discussed in text; side chains of key acNve site residues are shown as sNcks. Colors as in panel (a). The sulfate ion from the crystallizaNon buffer that binds in the phosphate-binding site is shown with spheres. A model for the proposed binding of substrate glucose 1-phosphate is shown in thin lines (yellow), based on a structural superposiNon with a related enzyme-ligand complex (PDB code: 1P5D).

Fig. 3

Overall structure and detailed views of the G121R crystal structure. (a) A superposiNon of the G121R missense variant (colored by domain as in Fig. 2) and WT PGM1 (white) showing their overall structural similarity. Arg121 is shown in magenta; the region of the WT enzyme corresponding to the disordered residues in the G121R structure is in yellow. (b) The G121R mutant shown with a surface, and superimposed with WT enzyme (yellow), to highlight the patch of disordered residues in the mutant. (c) A close-up view comparing the vicinity of residue 121 in the WT and the G121R structures. The side chain of Phe257 (yellow), which lies in the interface between D1 and D2 in the WT structure, is displaced by Arg121 (magenta), causing disorder from residues 258–264 (between red circles). (d) AdvenNNous hydrogen bonds made between Arg121 and the backbone carbonyl groups of several residues in the D1-D2 interface.

Fig. 4

Overall and detailed views of the G291R crystal structure. (a) A superposition of the G291R missense variant (colored by domain as in Fig. 2) and WT PGM1 (white) showing their overall structural similarity. Arg291 is shown in magenta; regions of the WT enzyme corresponding to the disordered residues in the mutant structure are highlighted in bright colors: residues 64–65 in orange; residues 116–125 in green; and residues 253–268 in blue. (b) The G291R mutant shown with a surface, and superimposed with WT enzyme as in (a). (c) The vicinity of Arg291 showing its location in the 3-way interface of domains 1, 2, and 3, and locations of the three disordered loops in the mutant structure (dashed lines). The metal binding loop from WT PGM1 is shown in yellow; the position of Trp359 in the WT and G291R structures shows its rearrangement in packing. (d) The metal-binding loop of WT PGM1, showing the bound Mg²⁺ ion, its coordinating aspartates, and location of Ser117. Hydrogen bonds made between Asp288 and other residues in the loop are in dashed red lines; coordinating interactions to the Mg²⁺ ion are in black dotted lines. (e) The region corresponding to (d) in the G291R structure, showing the rearrangement of the three aspartates and loss of the bound Mg²⁺. Hydrogen bond interactions of Asp288 and Asp292 are shown with red dashed lines.

Fig. 5

Relationship of the known PGM1 missense variants to domain interfaces of the enzyme. A backbone trace of PGM1 showing the interface (blue surface) between the D2/3 unit (cyan) and the rest of the enzyme, i.e., domains 1 and 4 (gray). Most of the known disease-related missense variants cluster in (magenta spheres) or very near (orange spheres) to this extensive interface. The four missense variants that fall outside of this region are shown in green.