Loss of the RNA-binding protein TACO1 causes late-onset mitochondrial dysfunction in mice

Abstract

The recognition and translation of mammalian mitochondrial mRNAs are poorly understood. To gain further insights into these processes in vivo, we characterized mice with a missense mutation that causes loss of the translational activator of cytochrome oxidase subunit I (TACO1). We report that TACO1 is not required for embryonic survival, although the mutant mice have substantially reduced COXI protein, causing an isolated complex IV deficiency. We show that TACO1 specifically binds the mt-Co1 mRNA and is required for translation of COXI through its association with the mitochondrial ribosome. We determined the atomic structure of TACO1, revealing three domains in the shape of a hook with a tunnel between domains 1 and 3. Mutations in the positively charged domain 1 reduce RNA binding by TACO1. The Taco1 mutant mice develop a late-onset visual impairment, motor dysfunction and cardiac hypertrophy and thus provide a useful model for future treatment trials for mitochondrial disease.

Figure 1. A point mutation in the Taco1 gene causes destabilization and loss of TACO1 in mice.

(a) Schematic showing the location of the mutation in the Taco1 gene, mRNA and protein. The 5′- and 3′-untranslated regions are shown as grey boxes, the predicted mitochondrial targeting sequence is shown as a black box (predicted using MitoProtII43). Conservation of the protein sequence surrounding the mutation is highlighted with residues identical to those in the mouse sequence boxed (the mutated isoleucine is shown in red in the mouse sequence). Sequences used were obtained from GenBank at NCBI (human, Homo sapiens, NP_057444.2; mouse, Mus musculus, NP_081622.1; cow, Bos taurus, NP_001192540.1; zebrafish, Danio rerio, NP_001076342.1; fruit fly, Drosophila melanogaster, NP_609366.2; worm, Caenorhabditis elegans, NP_497183.1; yeast, Saccharomyces cerevisiae, NP_011535.1) and the alignment was produced using ClustalW2 (ref. 44). (b) The mutation was confirmed by Sanger sequencing of PCR amplicons from homozygous Taco1 mutant mice and matched wild-type littermates. (c) TACO1 protein levels were determined in mitochondria isolated from livers and hearts of Taco1^wt/wt and Taco1^mut/mut mice by immunoblotting. Porin was used as a loading control. The data are representative of results obtained from at least 8 mice from each strain.

Figure 2. Loss of TACO1 results in isolated complex IV deficiency and specific reduction in COXI translation.

(a) Mitochondrial proteins (50 μg) isolated from livers and hearts of adult Taco1^mut/mut and Taco1^wt/wt mice were analysed by BN-PAGE and immunoblotting. Specific antibodies representing proteins of each of the mitochondrial complexes were used to compare abundance of complexes in the Taco1^wt/wt and Taco1^mut/mut mice. The data are representative of results obtained from 8 mice from each strain and at least three independent biological experiments. *P<0.05 compared with control treatments by a two-tailed paired Student's t-test. (b) The Taco1 mutation does not affect mitochondrial RNA metabolism. The abundance of mature mitochondrial transcripts in mitochondria isolated from adult Taco1^wt/wt and Taco1^mut/mut livers and hearts was analysed by northern blotting. 18S rRNA was used as a loading control. The data are representative of results obtained from at least 8 mice from each strain and three independent biological experiments. *P<0.05 compared with control treatments by a two-tailed paired Student's t-test. (c) Mitochondrial proteins (20 μg) from adult Taco1^wt/wt and Taco1^mut/mut liver and heart were resolved on SDS–PAGE gels and immunoblotted against antibodies to investigate the steady-state levels of nuclear and mitochondrial-encoded proteins. Porin was used as a loading control. Representative blots are shown of three independent biological experiments. *P<0.05 compared with control treatments by a two-tailed paired Student's t-test. (d) Protein synthesis in the liver and heart of adult Taco1^wt/wt and Taco1^mut/mut was measured by pulse incorporation of ³⁵S-labelled methionine and cysteine. Equal amounts of mitochondrial protein (50 μg) were separated by SDS–PAGE, stained using Coomassie Brilliant Blue to show equal loading and visualized by autoradiography. Representative gels are shown of three independent biological experiments.

Figure 3. TACO1 is a mt-Co1 mRNA-binding protein.

(a) A set of partially overlapping RNA probes was used in RNA electrophoretic mobility shift assays (RNA EMSA). The locations of RNA probes within the mt-Co1 gene region in the mouse mtDNA are shown schematically. RNA EMSA analyses were performed with increasing concentrations of purified TACO1 protein. (b) Identification of TACO1-binding sites within the mt-Co1 region of mouse mtDNA using a custom RNA tiling microarray. The mtDNA region between nt 4,932–8,622 was tiled in 18 nt increments. The nuclear RNase P RNA (Rpph1), 5S rRNA (5S), a nuclear encoded initiator tRNA-Met (Mⁱ) and 500 nt of the 3′ end of the β-actin mRNA (Actb)-coding region and the RNA localization signal within its 3′-UTR were used as controls. (c) Single-nucleotide RNA tiling microarray analysis of the TACO1-binding sites within mt-Co1. (d) Part of the microarray showing binding of TACO1 to sequences sampled at 1-nt intervals through a TACO1-binding site within mt-Co1 (nt 5,649–5,721). The aligned sequences of the RNA probes for which TACO1 has the greatest affinity are shown to contain guanine (G) and adenine (A) residues, coloured in red and yellow, respectively. (e) The consensus binding site for TACO1 within mt-Co1 was identified using multiple EM for motif elicitation (MEME) analysis and is shown as a sequence logo. The 79 probes with signal above 2,000 A.U. from the single-nucleotide RNA tiling microarray were used for the analysis and the discovered motif occurred with P=1.6e-184.

Figure 4. Low levels of TACO1 associate with the mitoribosome.

(a) A continuous sucrose gradient was used to determine TACO1 distribution relative to mitochondrial ribosomal protein markers of the small (MRPS35) and large (MRPL37) ribosomal subunits by immunoblotting with specific antibodies. (b) Immunoprecipitation of FLAG-tagged TACO1, MRPS27 and MRPL44 from NIH-3T3 cells expressing either mitochondria targeted GFP, or GFP-tagged MRPS27, MRPL44 or TACO1, followed by immunoblotting with an anti-GFP antibody and an anti-FLAG antibody to detect the immunoprecipitated proteins. (c) Mitochondrial ribosomal protein abundance was measured by immunoblotting in mitochondria isolated from hearts and livers of young and old Taco1^mut/mut and Taco1^wt/wt mice. (d) The distributions of mitochondrial mRNAs, 12S and 16S rRNAs in sucrose gradients were analysed by northern blotting. The data are typical of results from at least three independent biological experiments.

Figure 5. Structure of TACO1 and similarity to bacterial DUF28 family members.

(a) Front view of TACO1 is in cartoon representation with domain 1 coloured in salmon, domain 2 in orange and domain 3 in wheat, respectively. The N- and C-terminals are indicated. (b) Interacting polar residues at the interface between domains are shown as sticks and hydrophobic residues as spheres. (c) Intra domain 2 interactions. Leu 164 and surrounding hydrophobic residues are shown as spheres. (d) The Asn 164 mutant amino acid and surrounding hydrophobic residues are shown as spheres. (e) Superimposition of TACO1 (grey) with Cbu1566 from Coxiella burnetii (yellow, PDB ID code 4F3Q). (f) Superimposition of TACO1 (grey) with Aq1575 from Aquifex aeolicus (magenta, PDB ID code 1LFP). The structural alignment was performed using the secondary-structure matching (SSM) method.

Figure 6. Surface charge distribution and RNA binding of TACO1.

(a) Front, side and back views of the molecular surface are coloured by the local electrostatic potential (blue, +5 kT; red −5 kT). (b) Front view of the surface of TACO1 with mutated residues shown as sticks. (c) RNA EMSA of TACO1 mutants. Analyses were performed with increasing concentrations of purified TACO1 protein. Free and bound RNAs are indicated.

Figure 7. Taco1^mut/mut mice have reduced complex IV enzyme activity, motor dysfunction and visual impairment.

(a) Respiratory complex activities were normalized to citrate synthase activity in mitochondria from Taco1^wt/wt (n=8) and Taco1^mut/mut mice (n=8). Data are means±s.e.m. of three to four separate experiments; *P<0.05 compared with control treatments by a two-tailed paired Student's t-test. (b) Reduced state 3 respiration in Taco1^mut/mut mice compared with Taco1^wt/wt and Taco1^mut/mut was measured using an OROBOROS oxygen electrode. Data are means±s.e.m. of three to four separate experiments; *P<0.05 compared with control treatments by a two-tailed paired Student's t-test. Quantification of behavioural studies evaluating number of times the box was reached in hanging wire experiments (c) and distance travelled along the wire in hanging wire experiments (d) comparing Taco1^wt/wt (n=5) and Taco1^mut/mut (n=5) mice. (e) Rotarod results measured in seconds spent on the rotarod over 4 days to show improvement and learning ability. (f) Time spent on the rotarod over 4 days to analyse motor function and learning ability. (g) Comparison of Taco1^wt/wt (n=5) and Taco1^mut/mut (n=5) tracking ability using optokinetic drum, measured in number of times spent tracking. (h) Comparison of Taco1^wt/wt and Taco1^mut/mut tracking ability measured in number of tracks performed. (i) Comparison of time spent in light versus dark in Taco1^wt/wt (n=5) and Taco1^mut/mut (n=5) mice. All behavioural data are means±s.e.m.; *P<0.01 compared with controls by a two-tailed Student's t-test. (j) Cresyl violet staining of 20 μm eye slices of adult Taco1^mut/mut mice (n=5) compared with Taco1^wt/wt mice (n=5) at × 40 magnification. (k) Quantitative analyses of the layers of the cresyl violet-stained slides were determined as a percentage of total μm using Image J. (l) Cresyl violet/toluidine blue staining of optic nerves from Taco1^wt/wt (n=5) and Taco1^mut/mut (n=5) mice visualized at × 100 magnification. (m) Scattered foci of mild cell loss and shrinkage of Purkinje cells are visible in the Taco1^mut/mut (white arrows) compared with the wild-type adult mice. Representative images of haematoxylin-and-eosin-stained brains from Taco1^wt/wt and Taco1^mut/mut mice (n=8).