Structural and functional analysis of the Rpf2-Rrs1 complex in ribosome biogenesis

Abstract

Proteins Rpf2 and Rrs1 are required for 60S ribosomal subunit maturation. These proteins are necessary for the recruitment of three ribosomal components (5S ribosomal RNA [rRNA], RpL5 and RpL11) to the 90S ribosome precursor and subsequent 27SB pre-rRNA processing. Here we present the crystal structure of the Aspergillus nidulans (An) Rpf2-Rrs1 core complex. The core complex contains the tightly interlocked N-terminal domains of Rpf2 and Rrs1. The Rpf2 N-terminal domain includes a Brix domain characterized by similar N- and C-terminal architecture. The long α-helix of Rrs1 joins the C-terminal half of the Brix domain as if it were part of a single molecule. The conserved proline-rich linker connecting the N- and C-terminal domains of Rrs1 wrap around the side of Rpf2 and anchor the C-terminal domain of Rrs1 to a specific site on Rpf2. In addition, gel shift analysis revealed that the Rpf2-Rrs1 complex binds directly to 5S rRNA. Further analysis of Rpf2-Rrs1 mutants demonstrated that Saccharomyces cerevisiae Rpf2 R236 (corresponds to R238 of AnRpf2) plays a significant role in this binding. Based on these studies and previous reports, we have proposed a model for ribosomal component recruitment to the 90S ribosome precursor.

Figure 1.

Overall structure of the Rpf2-Rrs1 core complex. Rpf2 and Rrs1 are shown in cyan and pink, respectively. (A) Ribbon diagram of the Rpf2-Rrs1 core complex viewed from the front (left) and top (right) of the complex. (B) Schematic drawings indicating each domain. The proteolysis resistant complex (Rpf2-Rrs1 core complex) used for this experiment is indicated by color regions. (C) Topology diagram indicating the secondary structure. Helices and beta strands are numbered.

Figure 2.

Interactions between Rpf2 and Rrs1. (A and B) Dimerization interface between the long α-helix of Rrs1 and wide β-sheet of Rpf2. Residues of Rrs1 (purple) and Rpf2 (cyan) involved in the interactions are displayed as stick models and labeled. (A) Hydrophobic interactions. (B) Hydrogen-bonding interactions. (C) Close-up view of the C-terminal proline-rich loop of Rrs1. Prolines are labeled and displayed as stick models. Side chains of D32, S92 and T93 from Rrs1 form intra-molecular hydrogen bonds. These three residues are labeled and displayed as yellow stick models. (D) Sequences in the proline-rich loop after CLUSTALW alignment (30). The sequences are as follows: An; Aspergillus nidulans, Sc; Saccharomyces cerevisiae, Hs; Homo sapiens, Rn; Rattus norvegicus, Ca; Candida albicans, At; Arabidopsis thaliana. The most conserved sites are highlighted in black. Red arrow indicates the region for which the structure was determined. (E) Close-up view of hydrogen-bonding interactions at the C-terminal loop of Rrs1. Amino acid residues that form the interaction are displayed as stick models. Conserved residues are denoted in Italics.

Figure 3.

5S rRNA binding assay. (A) Schematic overview of the Rpf2 and Rrs1 variants. (B–E) Results of the gel shift assay. Sc5S rRNA (50 pmol) was incubated without factor or with 50, 100 and 200 pmol of ScRpf2-Rrs1 complex variants: ScRpf2 full-Rrs1 full (B), ScRpf2ΔC-Rrs1ΔC (C), ScRpf2ΔC-Rrs1 full (D), ScRpf2 full-Rrs1ΔC (E). Asterisk (*) indicates the 5S rRNA dimer, as confirmed by a gel-filtration analysis and urea-PAGE. Each gel was stained with ethidium bromide (Et-Br stain, left) and Coomassie Brilliant Blue (CBB stain, right).

Figure 4.

Evolutionarily conserved and electrically positive regions of ScRpf2 (region-1 to 4) were selected for mutagenesis experiments to test the significance of the interactions with 5S rRNA. (A) Sequence conservation is mapped onto the surface along with variable (cyan) and conserved (purple) residues using Consurf (28). (B) Electrostatic surface potential diagrams with positive (blue) and negative (red) electrostatic potentials are mapped onto a van der Waals surface diagram of the conserved surface patch using APBS (29) The color scale ranges between −3 k_BT (red) and +3 k_BT (blue), where k_B is Boltzmann's constant and T is temperature. (C) Ribbon diagram of the Rpf2-Rrs1 core complex in the same orientation as in A and B. Four regions containing seven residues (red) were selected for mutation analysis. Residue numbers of ScRpf2 are shown. Letters in parentheses correspond to the residue numbers for AnRpf2. (D) Results of the gel shift assay. Sc5S rRNA (50 pmol) was incubated without factor or with 50, 100, or 200 pmol of ScRpf2-Rrs1 complex mutated variants; the denoted numbers correspond to the mutant No. (Supplementary Table S1). Wild type indicates ScRpf2ΔC-Rrs1ΔC purified using the same method used to purify other point-mutated variants. Asterisk (*) indicates the 5S rRNA dimer, as confirmed by a gel-filtration analysis and urea-PAGE. Results are shown as an ethidium bromide-stained gel.

Figure 5.

Structural comparison of the N- and C-terminal halves of the Brix domains. (A and B) N- and C-terminal halves of Rpf2/Brix domain and (C) C-terminal half of Rpf2/Brix domain plus Rrs1. The long α-helix of Rrs1 appears to correspond with the C-terminal half of the Rpf2/Brix domain. (D and E) The N- and C-terminal halves of Mil/Brix, showing the duplicated architecture. (F and G) The superposition of the N-halves and C-halves of Rpf2/Brix and Mil/Brix. Rpf2, Rrs1 and Mil are colored in cyan, pink and yellow, respectively.

Figure 6.

Rpf2-subcomplex model and its location on the ribosomal precursor. (A) Rpf2-subcomplex model (left) and schematic representation (right). Red spheres represent the 5S rRNA binding region on Rpf2. Yellow spheres represent the RpL5 and RpL11 interaction region on Rrs1. (B) Rpf2-subcomplex model superposed on the ribosomal precursor (left); schematic representation (right). The molecules are colored as follows: cyan, Rpf2; pink, Rrs1; green, 5S rRNA; light orange, RpL11; blue, RpL5; gray, pre-60S particle.