Human ERAL1 is a mitochondrial RNA chaperone involved in the assembly of the 28S small mitochondrial ribosomal subunit

Abstract

The bacterial Ras-like protein Era has been reported previously to bind 16S rRNA within the 30S ribosomal subunit and to play a crucial role in ribosome assembly. An orthologue of this essential GTPase ERAL1 (Era G-protein-like 1) exists in higher eukaryotes and although its exact molecular function and cellular localization is unknown, its absence has been linked to apoptosis. In the present study we show that human ERAL1 is a mitochondrial protein important for the formation of the 28S small mitoribosomal subunit. We also show that ERAL1 binds in vivo to the rRNA component of the small subunit [12S mt (mitochondrial)-rRNA]. Bacterial Era associates with a 3' unstructured nonanucleotide immediately downstream of the terminal stem-loop (helix 45) of 16S rRNA. This site contains an AUCA sequence highly conserved across all domains of life, immediately upstream of the anti-Shine-Dalgarno sequence, which is conserved in bacteria. Strikingly, this entire region is absent from 12S mt-rRNA. We have mapped the ERAL1-binding site to a 33 nucleotide section delineating the 3' terminal stem-loop region of 12S mt-rRNA. This loop contains two adenine residues that are reported to be dimethylated on mitoribosome maturation. Furthermore, and also in contrast with the bacterial orthologue, loss of ERAL1 leads to rapid decay of nascent 12S mt-rRNA, consistent with a role as a mitochondrial RNA chaperone. Finally, whereas depletion of ERAL1 leads to apoptosis, cell death occurs prior to any appreciable loss of mitochondrial protein synthesis or reduction in the stability of mitochondrial mRNA.

Figure 1. ERAL1 associates with the 28S mt-SSU

(A) Left-hand panel: Western blot analysis, with the indicated antibodies, of HEK-293T cells subfractionated into nuclear (N) or cytoplasmic (Cp) compartments, with the latter fraction further divided into mitochondrial (M) and cytosolic (Cs) fractions. Equivalent volumes of each fraction are loaded. Right-hand panel: mitochondria (30 μg) were pretreated with the indicated amounts of proteinase K and solubilized with Triton X-100 or subjected directly to Western blot analysis with the indicated antibodies. (B) HeLa cell lysate (0.7 mg) was separated by isokinetic gradient centrifugation as detailed in the Experimental section, prior to fractionation and Western blot analysis to indicate the position of mitoribosomal subunits (DAP3 for 28S mt-SSU; MRPL3 for 39S mt-LSU). The blot is representative of three independent repeats. (C) Eluted immunoprecipitate from mitochondria of cells expressing MRPS27–FLAG were separated by isokinetic gradient centrifugation and fractions were silver-stained (upper panel) or subjected to Western blotting with the indicated antibodies (lower panels). 28S mt-SSU, grey circle; 39S mt-LSU, black hexagon. Gels are representative of two independent repeats.

Figure 2. ERAL1 is required for assembly of the 28S small mitoribosomal subunit

(A) Western blot of HEK-293T cell lysate (5, 10 or 15 μg) pretreated with NT or three independent ERAL1-targeted siRNAs for 3 days and probed with anti-ERAL1 or anti-β-actin antibodies. UTR, siRNA targeting the the 3′-UTR of Eral1; ORF1 and ORF2, siRNAs targeting regions of the open reading frame. In all further siRNA experiments, depletion of ERAL1 was confirmed by Western blotting. (B) Western blot of immunoprecipitate from cells expressing MRPS26–FLAG following 3 days of non-targeting (NT) or ERAL1-depletion (ORF1) siRNA. Position of mitoribosomal subunits are indicated with the markers DAP3, MRPS18B and MRPS25 for the28S mt-SSU, and MRPL3 and MRPL12 for the 39S mt-LSU. Quantification was performed on three independent experiments. On each occasion, signals were normalized to levels of MRPS26 in the immunoprecipitate. Results are means+S.D. (C) Western blot of immunoprecipitate from cells expressing ICT1–FLAG. Experimental details and data analysis as described for panel (A) (n=3). (D) Northern blot of 4 μg of RNA isolated from HEK-293T cells treated with si-NT or following ERAL1 depletion (si-ORF1) for 4 days. RNA from four independent experiments is shown. Probes highlight transcripts encoding components of complex I (MTND2), complex III (MTCYB; mitochondrially encoded cytochrome b) and complex IV (MTCO3, COX3) as well as both 16S (MT-RNR2) and 12S (MT-RNR1) mt-rRNAs. A probe to human 18S rRNA is shown as a loading and quality control (18S). The quantification is shown for 14 independent transcripts from four repeats. Results are means+S.D. ***P<0.001, **P<0.01, *P<0.05.

Figure 3. ERAL1 binds in vivo to the predicted 3′ terminal stem–loop of 12S mt-rRNA

Short RNA species bound by ERAL1 in vivo were identified following the CLIP assay as described in the Experimental section. (A) The sequence of the 50 nt 3′ terminal residues of 12S mt-rDNA, equivalent to nt 1551–1601 of the mitochondrial genome [1] is listed, along with the first four 5′ residues of mt-tRNA^Val, immediately downstream. The entire insert of each clone is shown as a filled line spanning the corresponding part of the reference sequence, with the number of clones with an identical sequence indicated. The minimal ERAL1-binding site is boxed. The bold AA highlights the dimethylated adenine residues at nt 1582–1583. The sequences of all 31 clones are given in Supplementary Table S1 (at http://www.BiochemJ.org/bj/430/bj4300551add.htm). *, two clones were deleted for two A residues of the A triplet at nt 1582–1584. (B) The highly conserved 3′ terminal stem–loop structure in the mitochondrial 12S (left-hand side) or bacterial 16S (helix 45, right-hand side) [22] are shown, along with the terminal unstructured nucleotides. The minimal 33 nt binding site for ERAL1, along with the delineated nonamer for Aquifex aeolicius Era [12] is highlighted as a solid line. The highly conserved AUCA is shown in bold; the anti-SD region, CCUCC, is italicized. The two dimethylated adenines are shown in outline [23,24].

Figure 4. Depletion of ERAL1 leads to apoptosis prior to appreciable loss of mitochondrial protein synthesis

(A) Counts of HEK-293T or HeLa cells were taken after siRNA treatment (3 or 4 days respectively) with an NT control or each of the three independent siRNAs targeted either to the EraL1 open reading frame (ORF1; ORF2) or to the corresponding 3′-UTR. Counts were performed on three (HeLa) or six (HEK-293T) independent repeats. ***P<0.001. (B) Metabolic radiolabelling of mitochondrial gene products was performed for 30 min in HeLa cells treated for three days with siRNA (NT or against ERAL1). Aliquots (50 μg) were separated by SDS/PAGE and visualized with a PhosphorImage as described in the Experimental section. Mitochondrial gene products are assigned from [20]. A small section of the gel is shown following exposure, rehydration and Coomassie Blue (CB)-staining to confirm equal loading. ATP8, mitochondrially encoded ATP synthase 8; Cytb, cytochrome b; ND, NADH dehydrogenase. (C) De novo synthesis (left-hand panel) and steady-state levels (right-hand panel) of mt-proteins in HEK-293T cell lysates following 3 days of siRNA treatment (siORF1 or NT control). Quantification of metabolic labelling is presented as a percentage of NT control for each translation product and is from three independent experiments. COX1/ND4, COX2/ATP6 and ND4L/ATP8 were quantified together as they could not confidently be quantified independently. The Western blot analysis visualized mitoribosomal proteins with antibodies as described in the text; markers used are: for complex I, NDUFB8 and NDUFA9; for complex IV, COX2; for the mitochondrial matrix, chaperone HSP70, and for the mitochondrial outer membrane, porin. Results are means+S.D.