Model for 5.8S rRNA processing. Initial processing of the 7S precursor is performed by Rrp44 in association with the core exosome, leaving a 5.8S+30 nt RNA. Final processing is performed by Rrp6. Mtr4 is proposed to help make the 5.8S+30 substrate accessible to Rrp6, possibly by removal of the RNA from the core exosome. Although Rrp6 is depicted in association with the core exosome during processing of the 5.8S+30 substrate, such an interaction may not be required, as previously suggested (Callahan and Butler, 2008).

Surface conservation. Evolutionary conservation scores were calculated and mapped onto the protein surface using Consurf (Landau et al, 2005). Conservation is depicted by colour gradient ranging from variable (white) to highly conserved (orange). (A) Mtr4 conservation mapped onto the Mtr4 structure. (B) Ski2 conservation mapped onto the Mtr4 structure. (C) Hel308 conservation mapped onto the Hel308 structure (PDB 2p6r). The strong conservation observed along the base of Mtr4 and Ski2 suggests that this is a potential interaction surface. See Supplementary Figure S3 for a more detailed alignment of Mtr4 and Ski2 sequences.

Structural overview of Mtr4. (A) Stereo view of the Mtr4 structure. A Cα trace is shown with rainbow colouring from N- (blue) to C-termini (red). Residue numbers are indicated. (B) Representative stereo view of the 2f_of_c electron density map, contoured at 1 σ. The (^*) corresponds to the same position marked in (A). (C) Ribbon and surface representations of the Mtr4 structure, coloured by structural domains (see also Supplementary Figure S1). (D) Mtr4 primary sequence and secondary structure. Helices and strands are numbered as indicated. Dashed lines denote gaps in the model due to lack of electron density. The first strand in domain 1 (β1) is not indicated, but presumably comes from part of the 85 residues at the N-terminus of Mtr4; the precise sequence could not be determined because no electron density is observed connecting it to the rest of the domain.

RNA-binding model. (A) Nucleic acid from the DNA-bound Hel308 structure (PDB 2p6r) was superimposed on the Mtr4 structure to model the likely RNA-binding site. Surface (top) and cartoon (middle) representations are shown, coloured by domain as in Figure 1. Inset images highlight specific structural features observed in Mtr4 that are associated with strand separation and translocation in Hel308, including the β-hairpin (from domain 2) and ratchet helix (domain 4). In addition, extensive interactions are predicted along the fist of the arch. (B) Sequence conservation mapped on the surface of the arch. The colour scheme of conservation is the same as in Figure 2A. The position of three strictly conserved arginine residues that potentially interact with RNA is indicated (⧫). (C) Stereo view of domain 4. Domain 4 of Mtr4 is an eight-helix bundle and is designated as a DSHCT domain (purple). The domain from Mtr4 is superimposed with the DNA-bound seven-helix bundle of Hel308 (white) to highlight the structural similarity between these two domains. An 18-residue disordered loop in Mtr4 is drawn as a dashed line. Helix H38 of Mtr4 corresponds to the ratchet helix of Hel308.

Domain 3a—the ‘arch' domain. (A) Sequence alignment of the arch domain. Conservation was calculated for Mtr4 sequences using Consurf (Landau et al, 2005), based on 108 Mtr4 sequences (orange, strictly conserved residues; yellow, similar residues). Conservation was similarly calculated for 42 Ski2 sequences and is displayed below the Mtr4 sequences. Observed (Mtr4) and predicted (Ski2) secondary structure for the arch is indicated. The colouring of the Mtr4 secondary structure corresponds to the stereo figure of the arch shown above. Two types of conservation are observed: residues that have a role in maintaining the overall fold or structure of the arch (▾), and those that have a potential functional role (⧫). Proline residues that are predicted to influence the bend at the elbow are underlined. (S.cer, Saccharomyces cerevisiae; S.pom, Schizosaccharomyces pombe; N.cra, Neurospora crassa; H.sap, Homo sapiens; M.mus, Mus musculus; D.rer, Danio rerio; D.mel, Drosophila melanogaster; C.ele, Caenorhabditis elegans; M.bre, Monosiga brevicollis; A.tha, Arabidopsis thaliana) (B) Comparison of the Mtr4 fist with the L14e ribosomal protein. Structures (top) and topology diagrams (bottom) are shown. Similar features between the structures are coloured blue. The central fold of the L14e protein from S. solfataricus (PDB 2joy) is structurally similar to the fist of Mtr4 and was used to guide model building in the fist. A more extensive comparison is shown in Supplementary Figure S2. (C) Domain arrangement of Ski2-like RNA and DNA helicases. Domains 1, 2, 3, and 4 are characteristic of all identified Ski2-like helicases. Sequence and secondary structure analysis indicates that the arch domain is unique to the Mtr4/Ski2 sub-family. The function of DDX60 as an RNA or DNA helicase has not been demonstrated.

The arch domain of Mtr4 is required for RNA processing and degradation, but not for helicase or ATPase activity. (A) Displacement of a radiolabelled single-stranded RNA from a 16-bp duplex with a 3′-single-stranded overhang by full-length Mtr4 and the Mtr4-archless mutant, as observed on a non-denaturing polyacrylamide gel. Illustrations on the left describe the mobility of duplex and single-stranded RNA through the gel; the asterisk represents the ³²P label. Aliquots were taken at the time points indicated after the addition of ATP. Lane 8 displays the complete dissociation of the duplex after heating an aliquot to 95°C, and lanes 7 and 15 show a reaction without ATP after 60 min. (B) Time courses for the fraction of displaced (unwound) RNA compared with total RNA is shown. Closed circles indicate Mtr4 activity; open triangles indicate Mtr4-archless activity. A solid line representing Mtr4 and a dashed line representing Mtr4-archless were fit to the data points as a first-order reaction (Wang et al, 2008), allowing for the determination of reaction amplitudes (A=0.427±0.010, Mtr4; A=0.364±0.016, Mtr4-archless) and observed unwinding rate constants (k′_unw=0.066±0.005/min, Mtr4; k′_unw=0.058±0.008/min, Mtr4-archless). (C) Enhancement of ATPase activity after introduction of RNA for Mtr4 and Mtr4-archless. The rate was determined using a malachite green assay that monitors the rate of release of inorganic phosphate [Pi] from ATP over time. Background values (activity in the absence of RNA) have been subtracted from the total rate. (D) The arch domain is required for optimal growth rate. The indicated plasmids were introduced into a yeast strain that had the MTR4 gene deleted from the chromosome, and that also contained a plasmid encoding wild-type Mtr4 with a URA3 selectable marker. Growth on 5-FOA plates (left) selects for cells that have lost the URA3 plasmid, and thus shows that Mtr4-archless confers a slow growth phenotype. (E) A western blot probed with anti-Mtr4 antibodies (top) and re-probed with anti- Pgk1 antibodies as a loading control shows that archless and wild-type Mtr4 are expressed at similar levels. (F) The 5′ ETS and 5.8S rRNA are degraded and processed through actions of the nuclear exosome and Mtr4. The 5′ ETS and 7S rRNA precursor are downstream products of 35S rRNA precursor processing events. 5.8S rRNA is generated through processing of 7S and 5.8S+30 intermediates and the 5′ ETS is completely degraded. (G) The 5.8S+30 intermediate accumulates in an Mtr4-archless mutant and an rrp6 knockout strain but not in strains that lack Rrp44 exonuclease or endonuclease activity. Shown is a northern blot probed with a ³²P-radiolabelled oligonucleotide specific for the 5.8S rRNA (top) and re-probed for the RNA subunit of the signal recognition particle (bottom). (H) The 5′ ETS signal is four-fold greater in the Mtr4-archless, rrp6 knockout, and rrp44-exo⁻ strains, but only the rrp44-exo⁻ strain shows decay intermediates. Same northern blot, as shown in (G), re-probed with a ³²P-radiolabelled oligonucleotide specific for the 5′ ETS is shown. Supplementary Figure S4 provides the structural justification for design of the Mtr4-archless mutant used in these studies.