Altered 2-thiouridylation impairs mitochondrial translation in reversible infantile respiratory chain deficiency

Abstract

Childhood-onset mitochondrial encephalomyopathies are severe, relentlessly progressive conditions. However, reversible infantile respiratory chain deficiency (RIRCD), due to a homoplasmic mt-tRNA(Glu) mutation, and reversible infantile hepatopathy, due to tRNA 5-methylaminomethyl-2-thiouridylate methyltransferase (TRMU) deficiency, stand out by showing spontaneous recovery, and provide the key to treatments of potential broader relevance. Modification of mt-tRNA(Glu) is a possible functional link between these two conditions, since TRMU is responsible for 2-thiouridylation of mt-tRNA(Glu), mt-tRNA(Lys) and mt-tRNA(Gln). Here we show that down-regulation of TRMU in RIRCD impairs 2-thiouridylation and exacerbates the effect of the mt-tRNA(Glu) mutation by triggering a mitochondrial translation defect in vitro. Skeletal muscle of RIRCD patients in the symptomatic phase showed significantly reduced 2-thiouridylation. Supplementation with l-cysteine, which is required for optimal TRMU function, rescued respiratory chain enzyme activities in human cell lines of patients with RIRCD as well as deficient TRMU. Our results show that l-cysteine supplementation is a potential treatment for RIRCD and for TRMU deficiency, and is likely to have broader application for the growing group of intra-mitochondrial translation disorders.

Figure 1.

Analyses of 2-thiouridine modification of mt-tRNA species in RIRCD, TRMU and control myoblasts. APM, (N-)acroylamino-phenyl-mercuric chloride); RIRCD, reversible infantile respiratory chain deficiency; TRMU, patient cells carrying the TRMU mutation; CTRL, control. (A) Northern blotting with adding APM to the gels to separate thiolated and unthiolated tRNA species was performed and probed for mt-tRNA^Glu, mt-tRNA^Lys, mt-trNA^Gln, cytoplasmic tRNA^Lys and 5S rRNA in immortalized human myoblasts of patients with RIRCD, TRMU deficiency and control cell lines. Results derive from two independent experiments, all representative blots were used for all tRNA probes following each other. (B) Quantification of the northern blots shows relative steady-state levels of the tRNAs and (C) the percentage of thiolated tRNA species compared with the whole amount of each tRNAs. For each sample the signal corresponding to the amount of tRNA was normalized to the signal corresponding to the amount of 5S RNA. The total levels of each of the four thio-modified tRNAs in the control cells were set arbitrarily to 100%. The values in the histogram are averages of two measurements, one corresponding to the signal from the gel without APM and the other to the total signal (thiolated plus unmodified) from the gel containing APM. The quantification of the modification is presented at the bottom panel and is expressed as a percentage of the thiolated signal from the thiolated + non-thiolated signals.

Figure 2.

Ablation of TRMU decreased 2-thiouridylation and steady-state level of mt-tRNA^Glu in RIRCD patient myoblasts. (A) Northern blotting with/without APM was performed in RIRCD and control cells after down-regulation of TRMU by siRNA or treatment by non-targeting siRNA (NT). Results derive from the same experiment; blots were used for all tRNA probes subsequently. Representative northern blots were quantified as described in Figure 1. (B) Relative steady-state levels of the tRNAs. (C) We show the percentage of thiolated tRNA species compared with the whole amount of each tRNAs in the studied cell lines.

Figure 3.

Down-regulation of TRMU hinders mitochondrial protein translation, protein synthesis and modifies the gene expression of other mt-tRNA modifier enzymes. (A) ³⁵S-Methionine pulse labelling for mitochondrial translation after down-regulation of TRMU resulted in a decreased mitochondrial translation in RIRCD cells, but not in controls. (B) Histogram of the representative translation assay. NT, non-targeting siRNA. (C) Immunoblotting detected very low mitochondrial protein levels for COX I, COX II and NDUFB8 after down-regulation of TRMU in RIRCD cells. TRMU depletion resulted in mildly decreased COX I and COX II and normal NDUFB8 in controls. β-Actin was used as a loading control. (D) Blue native PAGE detected decreased complex I and IV in RIRCD myoblasts, and a mild decrease of complex I and IV in controls after down-regulation of TRMU by siRNA. Complex II showed an additional band if TRMU was down-regulated both in RIRCD cells and controls. Complex III was normal, but we also detected some additional bands by complex V antibody. Immunoblotting with complex II antibodies was used as a loading control. (E) Real-time PCR analysis indicated elevated gene expression of TRMU in RIRCD myoblasts compared with control cells and following TRMU siRNA transfection the gene expression, as expected, decreased in both cell lines. The expression of EARS2 and the MTO1 (another tRNA modifying enzyme) seemed to be lower in the patient cell line when comparing to control and this further decreased after TRMU down-regulation. The Cystathionase (CST) expression, however, increased after the siRNA transfection. (n = 3). Data are represented as the mean ± SD. (F) Immunoblotting for MTO1 and EARS2 in the same cell lines detected no significant change in protein expressions.

Figure 4.

2-Thiouridylation is decreased in RIRCD skeletal muscle. We performed northern blotting in skeletal muscle and probed for mt-tRNA^Glu, mt-tRNA^Lys, mt-tRNA^Gln, cytoplasmic tRNA^Lys and 5S rRNA. Quantification of the northern blot results shows the relative steady state and percentage of thiolated tRNA species compared with the whole amount of each tRNAs. (A) Northern blotting with and without APM has been performed in skeletal muscle of control individuals of different age. 3 m, 3 month; 1y 6 m, 1 year 6 months. Representative blots were used for all tRNA probes following each other. (B and C) Quantification of the representative northern blot results shows the relative steady state and percentage of thiolated tRNA species compared with the whole amount of each tRNAs. (D and E) Northern blotting with and without APM has also been performed in skeletal muscle of other control individuals of different age and in a TRMU patient (only quantification is shown). (F–K) Quantification of the northern blot results (with and without APM) was performed in skeletal muscle of follow-up biopsies of two RIRCD patients in the symptomatic phase and after recovery. (F–H) Relative steady-state and percentage of thiolated tRNA species compared with the whole amount of each tRNAs in Patient 1. Quantification of the northern blot results shows the relative steady state and percentage of thiolated tRNA species compared with the whole amount of each tRNAs. (I–K) Relative steady state and percentage of thiolated tRNA species compared with the whole amount of each tRNAs in Patient 2.

Figure 4.

Continued

Figure 5.

In vitro l-cysteine supplementation improved the mitochondrial translation defect and increased complex activities. (A and B) Blue native PAGE indicated higher level of OXPHOS complexes after l-cysteine supplementation in both control and RIRCD myoblasts (n = 5). Data were normalized to the complex II and are presented as the mean ± SD. (C and D) ‘In gel’ activity measurements also demonstrated significantly increased complex activities in both cell lines after l-cysteine supplementation. SDS–PAGE detection of complex II was used as loading control. (n = 5). Data are presented as the mean ± SD. (E) Blue native PAGE after TRMU down-regulation in a control and a RIRCD cell line. (F) Immunoblotting after TRMU down-regulation in an RIRCD cell line, 5 mm l-cysteine prevented the mitochondrial translation defect of mitochondrial proteins (Complex I and Complex IV) when TRMU was down-regulated in RIRCD cells.

Figure 6.

In vitro l-cysteine supplementation increased the level of mitochondrial complexes in both TRMU- and MTO1-deficient fibroblasts. (A) Blue native PAGE indicated low level of complex I and IV in both patient cell lines compared to control fibroblasts. l-cysteine treatment improved the low level of these RC complexes. (B) Silver-stained mitochondrial complexes shown as loading control before and after treatment. (C) The relative level of RC complexes compared with control cells. Data were normalized to the complex II and are presented as the mean ± SD (n = 3). The control value obtained for the control untreated fibroblasts was represented as 100% and the value from the l-cysteine treated cells was expressed as a percentage of the control value. The asterisk denotes that the level of complex IV was significantly lower in the MTO1 and TRMU patient cells compared to control (P ≤ 0.004, ANOVA). The triangle indicates the significance after l-cysteine supplementation (P ≤ 0.006, ANOVA).

Figure 7.

Schematic representation of cysteine sources for functional TRMU enzyme. The cystathionase enzyme or also called cystathionine gamma-lyase plays an essential role in cysteine production. However, in the early months of life the activity of this enzyme is low. Metallothionein which represents another cysteine source, although presents at high levels at birth, dramatically decreases in the neonatal period. Therefore, the production of this amino acid is limited. Dietary cysteine intake might play a crucial role for the normal TRMU enzyme activity within the first few months of life when combined with underlying genetic diseases.