Archaeal DnaG contains a conserved N-terminal RNA-binding domain and enables tailing of rRNA by the exosome

Abstract

The archaeal exosome is a phosphorolytic 3'-5' exoribonuclease complex. In a reverse reaction it synthesizes A-rich RNA tails. Its RNA-binding cap comprises the eukaryotic orthologs Rrp4 and Csl4, and an archaea-specific subunit annotated as DnaG. In Sulfolobus solfataricus DnaG and Rrp4 but not Csl4 show preference for poly(rA). Archaeal DnaG contains N- and C-terminal domains (NTD and CTD) of unknown function flanking a TOPRIM domain. We found that the NT and TOPRIM domains have comparable, high conservation in all archaea, while the CTD conservation correlates with the presence of exosome. We show that the NTD is a novel RNA-binding domain with poly(rA)-preference cooperating with the TOPRIM domain in binding of RNA. Consistently, a fusion protein containing full-length Csl4 and NTD of DnaG led to enhanced degradation of A-rich RNA by the exosome. We also found that DnaG strongly binds native and in vitro transcribed rRNA and enables its polynucleotidylation by the exosome. Furthermore, rRNA-derived transcripts with heteropolymeric tails were degraded faster by the exosome than their non-tailed variants. Based on our data, we propose that archaeal DnaG is an RNA-binding protein, which, in the context of the exosome, is involved in targeting of stable RNA for degradation.

Figure 1.

Comparison of bacterial and archaeal DnaG and composition of reconstituted Sulfolobus solfataricus exosomes. (A) Domain composition of DnaG from Escherichia coli and S. solfataricus. Toprim-N and Toprim-C are the N-and C-terminal parts of the crystallized part of E. coli DnaG (40), which do not show similarity to archaeal DnaG proteins. NTD, N-terminal domain; CTD, C-terminal domain. (B) Schematic illustration of different exosomal complexes which were reconstituted previously and/or in this work. 41, Rrp41; 42, Rrp42; N, T and C, NTD, TOPRIM domain and CTD of DnaG. The Toprim domain is in dark green. Top views based on crystal structures of the Rrp41/Rrp42 hexamer (4), nine-subunit exosomes with homotrimeric, Rrp4 or Csl4 containing caps (5,22) and biochemical data for DnaG-containing exosomes (33).The Csl4-NT-exosome contains a homotrimeric cap build of the fusion protein Csl4-NT, which comprises full-length Csl4 and the NTD of DnaG.

Figure 2.

Phylogenetic analysis of DnaG proteins in Archaea. Genes encoding DnaG homologs exist in all genome sequenced archaea. The neighbor-joining phylogenetic tree of DnaG proteins is based on full-length protein sequences obtained from NCBI (http://www.ncbi.nlm.nih.gov/). Archaea which do not harbor genes for the core exosomal subunits Rrp41 and Rrp42 are marked with ‘-exo’. In Methanomicrobia, exosome-less and exosome-containing species group separately. H. volcanii, Haloferax volcanii; H. borinquense, Halogeometricum borinquense; H. walsbyi, Haloquadratum walsbyi; N. magadii, Natrialba magadii; H. salinarum, Halobacterium salinarum; H. marismortui, Haloarcula marismortui; N. pharaonis, Natronomonas pharaonis; M. hungatei, Methanospirillum hungatei; M. labreanum, Methanocorpusculum labreanum; M. petrolearius, Methanoplanus petrolearius; M. palustris, Methanosphaerula palustris; M. barkeri, Methanosarcina barkeri; M. burtonii, Methanococcoides burtonii; M. thermophila, Methanosaeta thermophila; M. paludicola, Methanocella paludicola; F. placidus, Ferroglobus placidus; A. fulgidus, Archaeoglobus fulgidus; T. kodakarensis, Thermococcus kodakarensis; P. furiosus, Pyrococcus furiosus; M. thermautotrophicus, Methanothermobacter thermautotrophicus; M. smithii, Methanobrevibacter smithii; M. jannaschii, Methanocaldococcus jannaschii; M. maripaludis, Methanococcus maripaludis; M. okinawensis, Methanothermococcus okinawensis; P. torridus, Picrophilus torridus; T. acidophilum, Thermoplasma acidophilum; M. kandleri, Methanopyrus kandleri; A. saccharovorans, Acidilobus saccharovorans; A. pernix, Aeropyrum pernix; I. hospitalis, Ignicoccus hospitalis; P. fumarii, Pyrolobus fumarii; M. sedula, Metallosphaera sedula; S. solfataricus, Sulfolobus solfataricus; S. tokodaii, Sulfolobus tokodaii; T. tenax, Thermoproteus tenax; P. aerophilum, Pyrobaculum aerophilum; N. equitans, Nanoarchaeum equitans; K. cryptofilum, Candidatus Korarchaeum cryptofilum; N. maritimus, Nitrosopumilus maritimus; E. coli, Escherichia coli.

Figure 3.

Multiple sequence alignment of DnaG using Clustal X-2.1. The secondary structure of Sulfolobus solfataricus DnaG was modeled with Phyre2 (http://www.sbg.bio.ic.ac.uk/phyre2/html). The domains of DnaG are marked above the alignment. Mutated residues in the N-terminal (K6 and Y7, this work) and TOPRIM (E175, (35)) domains of S. solfataricus DnaG are marked with red triangles above the alingment. In the C-terminal domain, the S. solfataricus D329 residue conserved in exosome-containing archaea is marked with an orange triangle above the alignment. 59 aa of the C-terminal domain of Methanosarcina barkeri were omitted from the analysis (marked with red 59 in the M. barkeri sequence). DnaG regions showing similarities to other proteins in bacteria and/or eukarya are marked below the alignment: orange line, similarity to bacterial RNA helicase; red line, similarity to mammalian ribosomal protein L32; purple line, similarity to transcription elongation factor Spt4/5 (compare to Supplementary Table S2). The archaeal species framed in rectangle are exosome-containing. The archaeal species out of the rectangle are exosome-less. S. solfataricus, Sulfolobus solfataricus; P. furiosus, Pyrococcus furiosus; M. barkeri, Methanosarcina barkeri; M. thermautotrophicus, Methanothermobacter thermautotrophicus; H. volcanii, Haloferax volcanii; M. jannaschii, Methanocaldococcus jannaschii; H. marismortui, Haloarcula marismortui; M. hungatei, Methanospirillum hungatei.

Figure 4.

The CTD of DnaG is involved in the interaction with the Csl4 exosome. (A) Schematics of the used DnaG variants. (B)-(E), Upper panels show silver stained 12% SDS-PAA gels (SDS-PAGE) and lower panels show corresponding western blot analyses with polyclonal DnaG-directed antibodies. All recombinant proteins carry hexahistidine tags. The only exception is the Strep-tagged Csl4 in (E). Relevant proteins are marked on the right side of the panels. The size of marker proteins in kDa is given on the left side. (B) The DnaG-directed antibodies detect DnaG-ΔNT and DnaG-ΔCT but not Rrp41. The loaded, purified recombinant proteins are indicated above the upper panel. (C), (D) and (E) Results from coimmunoprecipitation experiments (CoIP) with polyclonal, Rrp41-directed antibodies or from a pull-down experiment with Strep-Tactin as indicated below the upper panels. In, input, the mixture of proteins used; FT, flow-through; W1, W5, the first and the last washing fractions; E, the elution fraction. (C) DnaG-ΔNT does not interact with the immobilized Rrp41-specific antibodies used for CoIP. Purified DnaG-ΔNT was used in the input fraction. (D) DnaG-ΔNT interacts with the Csl4 exosome. DnaG-ΔNT was mixed with reconstituted Csl4 exosome and subjected to CoIP. (E) No detectable interaction between DnaG-ΔCT and the Csl4 exosome. Cell-free extract of the Escherichia coli strain producing DnaG-ΔCT (DnaG-ΔCT lysate) was mixed with the reconstituted Strep-Csl4 exosome, and CoIP was performed. Dilutions of purified DnaG-ΔCT were loaded left to the elution fraction. The asterisk indicates an E. coli protein band which co-purifies with DnaG-ΔNT and DnaG-ΔCT.

Figure 5.

Conserved residues in the NTD are essential for the RNA binding activity of DnaG. (A) Coomassie stained SDS-PAA gel with purified wild-type DnaG (WT), mutated DnaG-E175Q with an amino acid exchange in the TOPRIM domain (E175Q) and mutated DnaG-K6AY7A protein with amino acid exchange in the NTD (K6AY7A). M, marker proteins, sizes in kDa are marked. (B) Circular dichroism analysis of WT DnaG and DnaG-K6AY7A at room temperature shows no change in the secondary structure elements of the mutant in comparison to the wild-type. (C), (D) and (E) show phosphorimages of EMSA in native 5% PAA gels with 25-fmol-labeled substrate and 2.5 pmol of WT DnaG or mutant proteins (marked above the panels). C, negative control without protein. Detected unbound substrates or complexes are marked on the right site of the panels. (C) EMSA with labeled poly(dA)₃₀ (DNA) or poly(rA)₃₀ (RNA). (D) EMSA with labeled poly(rA)₃₀ and 40 pmol of unlabeled RNA (30-nt MCS-RNA or poly(rA)₃₀) as competitor. The presence or absence of competitor is indicated above the panel. (E) EMSA with labeled poly(rA)₃₀ and 2.5 pmol unlabeled poly(rA)₃₀ (RNA) or 25 pmol unlabeled poly(dA)₃₀ (DNA) as competitor. The presence of competitor in the assays is indicated above the panel.

Figure 6.

The NTD of DnaG confers strong poly(A) RNA binding capability and poly(A) specificity to the fusion Csl4-NT protein. (A) A phosphorimage of EMSA in native 5% PAA gel with the indicated amounts of Csl4, Csl4-NT or DnaG and 25 fmol of labeled poly(rA)₃₀. The presence of unlabeled competitors (30-nt MCS-RNA or poly(rA)₃₀) in the assays and their amounts is indicated. C, negative control without protein. (B) A phosphorimage of a denaturing 16% PAA gel with degradation assays containing 2 pmol of 5′-labeled poly(rA)₃₀ and 0.3 pmol of the Csl4 exosome, the DnaG/Csl4-exosome, the Csl4-NT exosome or Csl4-NT. The time of incubation (min) is indicated. The poly(rA)₃₀ substrate and the degradation (degr.) products are marked on the right side. (C) Graphical representation of the results shown in (B) and of two further independent RNA degradation assays. (D) A phosphorimage of a denaturing 16% PAA gel with degradation assays containing 50 fmol internally labeled transcript of 59 nt, which corresponds to a native RNA tail in Sulfolobus solfataricus (native tail RNA), and 0.3 pmol of the indicated exosome complexes. The other descriptions are like in (B).

Figure 7.

DnaG-K6AY7A binds to the Csl4 exosome but does not influence the degradation of poly(rA)₃₀. (A) Strep-Csl4 exosome was mixed with DnaG-K6AY7A containing cell-free extract and a pull-down assay with Strep-Tactin Sepharose beads was performed (lanes 7–11). As a negative control, the assay was performed with the cell-free extract only (lanes 2–6). M, marker; In, input, the mixture of proteins used; FT, flow-through; W1, W5, the first and the last washing fractions; E, the elution fraction. The protein fractions were analyzed by 12% SDS-PAA gel and silver stained. Relevant proteins are marked on the right side of the panel. The size of marker proteins in kDa is given on the left side. A protein co-purifying with Strep-Csl4 is marked by an asterisk. (B) A phosphorimage of a denaturing 16% PAA gel with degradation assays with 8 pmol radioactively labeled poly(rA)₃₀ and 0.3 pmol of the Csl4 exosome, DnaG/Csl4 exosome or the DnaG-K6AY7A/Csl4 exosome. The time of incubation (min) is indicated. The poly(rA)₃₀ substrate and the degradation (degr.) products are marked on the right side. (C) Graphical representation of the results from (B) and two further independent RNA degradation assays.

Figure 8.

DnaG enables the polyadenylation of rRNA by the exosome. Shown are phosphorimages of assays with an internally labeled in vitro transcript (i.v.Tr.) corresponding to the 3′-end of 16S rRNA in Sulfolobus solfataricus (3′ 16S rRNA; panels A and B), 5′-labeled native 5S rRNA (panels C and D) and internally labeled in vitro transcript corresponding to 5S rRNA (panels E and F). Substrate and reaction products are marked on the right side of the panels. Proteins and protein complexes are indicated above the panels, along with the reaction time (min). C, negative control without protein. (A) Denaturing 10% PAA gel with polyadenylation assays containing 60 fmol of the radioactively labeled substrate and 0.3 pmol of the indicated exosome complexes. (B) EMSA in native 5% PAA gel with 120 fmol of the radioactively labeled transcript and 5 pmol of each of the indicated proteins and protein complexes. (C) Denaturing 10% PAA gel with polyadenylation assays with 70 fmol native 5S rRNA and 0.3 pmol of the indicated exosome complexes. (D) EMSA in native 5% PAA gel with 140 fmol of the radioactively labeled substrate and 5 pmol of each of the indicated proteins and protein complexes. (E) and (F) Denaturing 10% PAA gels with polyadenylation assays. 70 fmol of the 5S rRNA in vitro transcript were present in each assay together with 0.3 pmol of the indicated exosome complexes. The incubation time in (F) was 30 min.

Figure 9.

A heteropolymeric tail increases the degradation of rRNA derived transcripts by the exosome. Shown are phosphorimages of degradation assays with 0.3 pmol of the DnaG/Csl4/Rrp4 and 60 fmol of each of the indicated, internally labeled in vitro transcripts. (A) Denaturing 16% PAA gel with assays containing 3′ 16S rRNA, 3′ 16S rRNA-A₂₀ (with a poly(A) tail of 20 nt) or 3′ 16S rRNA-hetero₂₀ (with an A-rich, heteropolymeric tail of 20 nt), as indicated above the panels. The reaction time at 60°C is also indicated. Upper panel, short time exposure allowing the differentiation between tailed and non-tailed transcripts. Lower panel, long time exposure allowing the detection of degradation products. Substrates and degradation products are marked on the right side. (B) Graphical representation of the results from (A) and two further independent RNA degradation assays. (C) Denaturing 16% PAA gel with assays containing 5S rRNA or 5S rRNA-hetero₂₀ transcripts. For conditions and descriptions, see (A). (D) Graphical representation of the results from (C) and two further independent RNA degradation assays.