Crystal and EM structures of human phosphoribosyl pyrophosphate synthase I (PRS1) provide novel insights into the disease-associated mutations

Abstract

Human PRS1, which is indispensable for the biosynthesis of nucleotides, deoxynucleotides and their derivatives, is associated directly with multiple human diseases because of single base mutation. However, a molecular understanding of the effect of these mutations is hampered by the lack of understanding of its catalytic mechanism. Here, we reconstruct the 3D EM structure of the PRS1 apo state. Together with the native stain EM structures of AMPNPP, AMPNPP and R5P, ADP and the apo states with distinct conformations, we suggest the hexamer is the enzymatically active form. Based on crystal structures, sequence analysis, mutagenesis, enzyme kinetics assays, and MD simulations, we reveal the conserved substrates binding motifs and make further analysis of all pathogenic mutants.

Fig 1. The weblogos of the conserved residues.

(A-D) The logos highlight the similarities and differences amongst the three PRS classes: class I, II and III. The logos demonstrate the overall conservation of the substrate-bound motifs.

Fig 2. Enzyme assays and 4 mutations located in the β9–10 strands.

(A) Enzyme assays of PRS1 and its mutants. (B) The enzyme assay of the H130A mutant displays no AMP product but produces ADP as its catalytic product. (C) The D183H, A190V, H193Q and H193L mutations are located in the antiparallel super flexible β9–10 strands. Error bars were calculated from the mean values, N = 3.

Fig 3. Structural differences between PRS1 and its mutants.

(A-E) Crystal structure superposition between PRS1 and its mutants. The mutant residue is colored green with the gray Fo-Fc map contoured at 2.5-σ in the image. The antiparallel β2–3 strands and the flexible loop are colored lemon.

Fig 4. MD simulation results of the five mutants and the WT PRS1.

(A) Contact numbers of both Asp⁶⁵ in WT (black line) and Gln⁶⁵ in D65N (red line) with the ATP binding motif β2–3. (B) Contact numbers of both Gln¹³³ in WT (black line) and Pro¹³³ in Q133P mutant (red line) with the catalytic flexible loop.(C) Contact numbers between the ATP binding motif β2–3 and the catalytic flexible loop for both the WT (black line) and the E43T mutant (red line). (D) Contact numbers of both Met¹¹⁵ in WT (black line) and Thr¹¹⁵ in M115T mutant (red line) with residues that are involved in the interactions with them in the crystal structures, respectively. (E) RMSF of all residues in both WT (black line) and A87T mutant (red line).

Fig 5. Structures of the PRS1 in different states.

(A) Negative stain EM of PRS1 and its complexes with AMPNPP, AMPNPP and R5P, and ADP. The four 2D averages (with scale bar, 50 Å) were identified through reference-free alignment and the classification of EM images of single particles preserved in the negative stain. (B) Hexamer subunit arrangement according to crystallography symmetry operation. (C) The 3D-EM structure of PRS1 with its hexamer crystal structure docked. The 3 different colors represent the three “head to head” dimmers like displayed in Fig. 1B. The figures shows different views of the PRS1 hexamer (PDB:3S5J), which was docked to the 3D EM structures. Scale bar, 50 Å.

Fig 6. The surface conservation among PRS proteins and the distribution of human PRS1 missense mutations associated with disorders.

(A) Only subunit A is shown for the surface conservation analysis of the PRS proteins. The dashed ovals represent different substrate pockets as indicated. The balls represent the different disorders associated with the residues. (B) The distribution of 16 different PRS1 missense mutations. The brown balls represent the residues distributed at the interfaces. The blue balls represented those near substrates or inhibitor binding pockets. The green balls represent those distributed in the inner parts of N- or C-terminal domains.