Biochemical Evidence for a Nuclear Modifier Allele (A10S) in TRMU (Methylaminomethyl-2-thiouridylate-methyltransferase) Related to Mitochondrial tRNA Modification in the Phenotypic Manifestation of Deafness-associated 12S rRNA Mutation

Abstract

Nuclear modifier gene(s) was proposed to modulate the phenotypic expression of mitochondrial DNA mutation(s). Our previous investigations revealed that a nuclear modifier allele (A10S) in TRMU (methylaminomethyl-2-thiouridylate-methyltransferase) related to tRNA modification interacts with 12S rRNA 1555A→G mutation to cause deafness. The A10S mutation resided at a highly conserved residue of the N-terminal sequence. It was hypothesized that the A10S mutation altered the structure and function of TRMU, thereby causing mitochondrial dysfunction. Using molecular dynamics simulations, we showed that the A10S mutation introduced the Ser¹⁰ dynamic electrostatic interaction with the Lys¹⁰⁶ residue of helix 4 within the catalytic domain of TRMU. The Western blotting analysis displayed the reduced levels of TRMU in mutant cells carrying the A10S mutation. The thermal shift assay revealed the T_m value of mutant TRMU protein, lower than that of the wild-type counterpart. The A10S mutation caused marked decreases in 2-thiouridine modification of U34 of tRNA^Lys, tRNA^Glu and tRNA^Gln. However, the A10S mutation mildly increased the aminoacylated efficiency of tRNAs. The altered 2-thiouridine modification worsened the impairment of mitochondrial translation associated with the m.1555A→G mutation. The defective translation resulted in the reduced activities of mitochondrial respiration chains. The respiratory deficiency caused the reduction of mitochondrial ATP production and elevated the production of reactive oxidative species. As a result, mutated TRMU worsened mitochondrial dysfunctions associated with m.1555A→G mutation, exceeding the threshold for expressing a deafness phenotype. Our findings provided new insights into the pathophysiology of maternally inherited deafness that was manifested by interaction between mtDNA mutation and nuclear modifier gene.

Introduction

The impairments in mitochondrial protein synthesis have been associated with both syndromic deafness (hearing loss with other medical problems such as diabetes) and nonsyndromic deafness (hearing loss is the only obvious medical problem) (1–5). Human mitochondrial translation machinery composed of 2 rRNAs and 22 tRNAs, encoded by mitochondrial DNA (mtDNA),⁴ and more than 150 proteins (ribosomal proteins, ribosomal assembly proteins, aminoacyl-tRNA synthetases, tRNA-modifying enzymes, tRNA methylating enzymes, and initiation, elongation, and termination factors), encoded by nuclear genes and imported into mitochondrion (6, 7). Mutations in LARS2, NARS2, and KARS encoding mitochondrial leucyl-tRNA synthetase, asparaginyl-tRNA synthetase, and lysyl-tRNA synthetase have been associated with deafness, respectively (8–10). The mitochondrial tRNA genes are the hot spots for deafness-associated mutations, including the tRNA^Leu(UUR) 3243A→G, tRNA^Ser(UCN) 7445A→G, 7511T→C, tRNA^His 12201T→C, tRNA^Asp 7551A→G, and tRNA^Glu 14692A→G mutations (11–17). The m.1555A→G and m.1494C→T mutations in the 12S rRNA gene have been associated with both aminoglycoside-induced and nonsyndromic deafness in many families worldwide (3, 4, 18–20). The m.1555A→G or m.1494C→T mutation is the primary causative evident, but modifier factors including aminoglycosides or nuclear modifier genes are required for the phenotypic manifestation of these mtDNA mutations (21–26). However, the role of these nuclear modifier genes remains poorly understood.

In the previous investigations, we showed that MTO1, MSS1 (GTPBP3), or MTO2 (TRMU) genes involved in the biosynthesis of the hypermodified nucleoside 5-methyl-aminomethyl-2-thio-uridine of several mitochondrial tRNAs were the potential modifier genes for the phenotypic expression of deafness-associated 12S rRNA 1555A→G mutation (25–30). These modified uridines at the wobble positions of tRNA^Lys, tRNA^Glu, and tRNA^Gln have a pivotal role in the structure and function of tRNAs, including structural stabilization, aminoacylation, and codon recognition at the decoding site of small rRNA (31–33). In particular, TRMU is the tRNA 5-methylaminomethyl-2-thiouridylate methyltransferase responsible for the 2-thiolation of tRNA^Lys, tRNA^Glu, and tRNA^Gln with mnm⁵s²U₃₄ in bacteria, mcm⁵s²U₃₄ in yeast, and m⁵s²U₃₄ in human mitochondria (34–37). Using TRMU as a candidate gene for genotyping analysis, in combination with functional assays, we identified a nuclear modifier allele (G28T and A10S) in the TRMU gene, which interacts with the m.1555A→G mutation to cause deafness (25). The A10S mutation resided at a highly conserved residue of the N-terminal sequence of this polypeptide but did not affect importation of TRMU precursors into mitochondria (25). However, the homozygous A10S mutation caused marked decreases in the steady state levels of mitochondrial tRNAs (25). Therefore, it was anticipated that the A10S mutation affected the structure and function of TRMU, thereby altering the mitochondrial function. The effect of A10S mutation on the stability of TRMU was assessed by the Western blotting and thermal shift assays. The primary defects by the mutation appeared to be the altered biosynthesis of 5-taurinomethyl-2-thiouridine (τm⁵s²U) nucleotides at the wobble position of mitochondrial tRNA^Gln, tRNA^Glu, and tRNA^Lys (25, 34, 38). Furthermore, the deficient synthesis of s²U₃₄ may alter the tRNA aminoacylation, because s²U₃₄ serves as a determinant for tRNA recognition by cognate aminoacyl-tRNA synthetases in bacteria (39). To further investigate the effect of the TRMU A10S mutation on mitochondrial function, we examined for the levels of tRNA modification, aminoacylation of tRNAs, translation, the rates of respiration, and the production of ATP and reactive oxygen species (ROS), through use of lymphoblastoid mutant cell lines derived from Arab-Israeli control subjects and from members of an Arab-Israeli family (two subjects carrying only m.1555A→G mutation (F12H and F6D), two individuals (F20C and F8A) harboring both m.1555A→G and heterozygous TRMU A10S mutations, and two individuals (F20A and F20D) carrying both m.1555A→G and homozygous TRMU A10S mutations) (25, 40).

Results

MD Simulation Analyses

We performed the molecular dynamics simulation to examine whether the A10S mutation alters the structure of TRMU (41). This method has been widely used for evaluating structural impact of diseasing-causing mutations (42). Based on the rational initial structure (43), both wild-type and mutated TRMU were evaluated by 300-ns all-atom molecular dynamics simulations. As shown in Fig. 1A, the A10S mutation did not affect the local structure around residue 10 and the overall structure of the TRMU protein. As shown in Fig. 1B, root mean square deviation curve of the mutated protein fluctuated more heavily than that of the wild-type protein, suggesting that the mutated protein exhibited unstable than its wild-type counterpart. Furthermore, we carried out root mean square fluctuation (RMSF) analysis on the two trajectories to analyze the mobility of different regions of the protein. As shown in Fig. 1C, the RMSF around the residue 10 in the two trajectories exhibited almost same value, indicating that the A10S mutation does not affect the stability of this region. However, the RMSF values of the helix 4 (residue 98–118) in mutated protein were much higher than that of the wild-type form (Fig. 1C), suggesting that the helix 4 in mutant protein was unstable than that of wild-type protein. Both the residue 10 and the helix 4 belong to the N-terminal catalytic domain, which is located on the opposite sides within the ATP binding pocket of the protein (43). In addition, it was anticipated that the A10S mutation could introduce new interactions between the residue 10 and helix 4, which may account for the less stability of the helix 4. As shown in Fig. 1D, the hydroxyl group of Ser¹⁰ in the mutant trajectory could form the dynamic electrostatic interaction with the Lys¹⁰⁶ residue of helix 4, whereas no interaction occurred between Ala¹⁰ and helix 4 in the wild-type trajectory. Thus, the electrostatic attraction between Ser¹⁰ and Lys¹⁰⁶ is quite dynamic, with H-bond occupancy of only around 1%, so this dynamic interaction caused the instability to the protein but did not affect its overall 3D structure.

FIGURE 1.

MD simulations on the wild-type and mutated TRMU proteins. A, superimposition of the crystal structure (gray) with the structures of wild-type (blue) and A10S mutant (orange) proteins at the end of the simulations. B, time evolution of the root mean square deviation (RMSD) values of all Cα atoms for the wild-type (black lines) and A10S mutant (red lines) proteins. C, RMSF curves were generated from the backbone atoms for the wild-type (black lines) and A10S mutant (red lines) proteins. D, electrostatic interactions formed between Ser¹⁰ and Lys¹⁰⁶ in the mutant protein.

The A10S Mutation Caused the Instability of TRMU

To experimentally test the predicted effect of A10S mutation for TRMU, we analyzed the levels of TRMU by Western blotting in these mutant cell lines carrying only m.1555A→G mutation, both m.1555A→G and heterozygous or homozygous A10S mutations and two control cell lines. These blots were then hybridized with other nuclear encoding mitochondrial proteins MTO1 and NDUFB8, as well as VADC as a loading control. As shown in Fig. 2, the levels of TRMU in mutant cell lines carrying only m.1555A→G, both m.1555A→G and heterozygous or homozygous A10S mutations were 94.1, 66.3, and 51.9%, relative to the average values of control cell lines, respectively. By contrast, the levels of MTO1 and NDUFB8 in mutant cell lines were comparable with those in control cell lines. These results strongly supported the deleterious effect of A10S mutation on TRMU structure.

FIGURE 2.

The A10S mutation caused the reduced levels of TRMU. A, scheme for the multiple sequence alignment of the TRMU homologues. The position of A10S mutation is marked with an arrow. B, Western blotting analysis of six mutant and two control cell lines. 20 μg of total cellular proteins from various cell lines were electrophoresed through a denaturing polyacrylamide gel, electroblotted and hybridized with TRMU, MTO1, and NDUFB8, respectively, and with VDAC as a loading control. Quantifications of TRMU levels were determined as described elsewhere (15). The values for the mutant cell lines are expressed as percentages of the average values for the control cell lines. Cell lines harboring homozygous (−/−), heterozygous (+/−), or wild-type (+/+) TRMU mutations are indicated. Cell lines carrying the m.1555A→G (−) or wild type (+) are indicated.

Analysis of TRMU Stability Using Differential Scanning Fluorimetry (DSF)

The thermal stability of mutant TRMU was also assessed using the DSF, a fluorescence method that was used to monitor solution phase protein stability (44). The technique involves subjecting a protein to heat denaturation under continuous fluorescence monitoring in the presence of the environmental sensitive fluorescent dye (Thermal-shift^TM dye). We used DSF to determine the melting temperature (T_m) of TRMU, the temperature at which the concentration of folded protein is equivalent to unfolded protein. The fluorescence changes of the dye orange occurred in the presence of 1 μg of wild-type and mutated TRMU over a temperature range from 25 to 95 °C. The thermal stability of mutant protein was compared with that of wild-type protein. As shown in Fig. 3, the T_m value of wild-type TRMU was 49.1 °C, whereas the T_m value of mutant TRMU was 45.9 °C. The lower thermal stability of mutant TRMU protein than wild-type protein further supported that the A10S mutation led to the instability of TRMU protein.

FIGURE 3.

Thermal stability of wild-type and mutant TRMU. The thermal denaturation was induced heating wild-type (solid line) and mutant (dashed line) TRMU proteins from 25 to 95 °C. Relative fluorescence curves were generated with the equation (F_T − F_min)/(F_max − F_min), where _T indicates fluorescence at temperature T, F_min indicates the minimum fluorescence, and F_max indicates the maximum fluorescence. ΔT_m indicates the difference of T_m value between wild-type and mutant TRMU. The calculations were based on three to four determinations.

Decreases in the Thiolation of tRNA^Gln, tRNA^Glu, and tRNA^Lys

To further investigate whether the TRMU A10S mutation affected the 2-thiouridine modification at position 34 in tRNAs, the 2-thiouridylation levels of tRNAs were determined by isolating total mitochondrial RNAs from eight lymphoblastoid cell lines, purifying tRNAs, qualifying the 2-thiouridine modification by the retardation of electrophoresis mobility in polyacrylamide gel containing 0.05 mg/ml ((N-acryloylamino)phenyl) mercuric chloride (APM) (45–47), and hybridizing digoxingenin (DIG)-labeled probes for tRNA^Gln, tRNA^Glu, and tRNA^Lys. In this system, the mercuric compound can specifically interact with the tRNAs containing the thiocarbonyl group-such as tRNA^Gln, tRNA^Glu, and tRNA^Lys, thereby retarding tRNA migration. As shown in Fig. 4, the 2-thiouridylation levels of tRNA^Gln, tRNA^Glu, and tRNA^Lys were reduced significantly in mutant cells carrying the homozygous TRMU A10S mutation, compared with control cells. In particular, the 2-thiouridylation levels of tRNA^Lys, tRNA^Glu, and tRNA^Gln in mutant cells carrying both homozygous A10S and m.1555A→G mutations were 52, 62, and 50%, relative to those of control cell lines, respectively. Furthermore, the 2-thiouridylation of tRNA^Lys, tRNA^Glu, and tRNA^Gln in mutant cells carrying both heterozygous A10S and m.1555A→G mutations were 76, 74, and 77%, relative to those of control cells, respectively. However, the levels of 2-thiouridylation of tRNA^Lys, tRNA^Glu, and tRNA^Gln in mutant cell lines carrying only m.1555A→G mutation were comparable with those in control cell lines.

FIGURE 4.

APM gel electrophoresis combined with Northern blotting of mitochondrial tRNAs. A, equal amounts (2 μg) of total mitochondrial RNAs were separated by polyacrylamide gel electrophoresis that contains 0.05 mg/ml APM and electroblotted onto a positively charged membrane and hybridized with DIG-labeled oligonucleotide probes specific for the tRNA^Lys. The blots were then stripped and rehybridized with DIG-labeled probes for tRNA^Glu and tRNA^Gln, respectively. The retarded bands of 2-thiolated tRNAs and non-retarded bands of tRNA without thiolation are marked by arrows. B, proportion in vivo of the 2-thiouridine modification levels of tRNAs. The proportion values for the mutant cell lines are expressed as percentages of the average values for the control cell lines. The calculations were based on three independent determinations of each tRNA in each cell line. The error bars indicate standard deviation; P indicates the significance, according to Student's t test, of the difference between mutant and control for each tRNA.

Analysis of Aminoacylation of tRNAs

We tested whether the deficient thiouridylation of tRNA^Lys, tRNA^Glu, and tRNA^Gln caused by the TRMU A10S mutation affects the aminoacylation of above tRNA as well as other tRNA. Indeed, our previous investigation also showed that the TRMU A10S mutation caused the reductions in the steady state levels of other tRNAs (25). The aminoacylation capacities of tRNA^Lys, tRNA^Tyr, tRNA^Leu(CUN), and tRNA^Ser(AGY) in control and mutant cell lines were examined by the use of electrophoresis in an acid polyacrylamide/urea gel system to separate uncharged tRNA species from the corresponding charged tRNA, electroblotting, and hybridizing with above tRNA probes (15, 48, 49). As shown in Fig. 5, the upper band represented the charged tRNA, and the lower band was uncharged tRNA. Electrophoretic patterns showed that either charged or uncharged tRNA^Ser(AGY) in all mutant cell lines migrated slower than control cell lines. The conformation change may be due to the presence of the m.12236G→A mutation in the tRNA^Ser(AGY) gene in the mutant cell lines (18) (supplemental Table S1). However, there were no obvious differences in electrophoretic mobility of tRNA^Lys, tRNA^Tyr, and tRNA^Leu(CUN) between mutant and control cell lines. Notably, the efficiencies of aminoacylated tRNA^Lys in cell lines carrying only m.1555A→G mutation, both m.1555A→G and heterozygous or homozygous TRMU A10S mutations were 103, 137, and 137% of those in control cell lines. Moreover, the efficiencies of aminoacylated tRNA^Leu(CUN) in cells harboring only m.1555A→G mutation, both m.1555A→G and heterozygous or homozygous TRMU A10S mutations were 117, 138, and 135% of those in control cell lines, whereas the efficiencies of aminoacylated tRNA^Ser(AGY) in those cell lines were 115, 126, and 129% of those in control cell lines. However, the levels of aminoacylation in tRNA^Tyr in mutant cell lines were comparable with those in the control cell lines.

FIGURE 5.

In vivo aminoacylation assays. A, 2 μg of total mitochondrial RNAs purified from eight cell lines under acid conditions were electrophoresed at 4 °C through an acid (pH 5.2) 10% polyacrylamide with 7 m urea gel, electroblotted, and hybridized with a DIG-labeled oligonucleotide probe-specific for the tRNA^Lys, tRNA^Tyr, tRNA^Leu(CUN), and tRNA^Ser(AGY), respectively. B, in vivo aminoacylated proportions of tRNA^Lys, tRNA^Tyr, tRNA^Leu(CUN), and tRNA^Ser(AGY) in the mutant and controls. The calculations were based on three independent determinations. Graph details and symbols are explained in the legend to Fig. 4.

Reductions in the Level of Mitochondrial Proteins

To further determine whether the TRMU A10S mutation alters mitochondrial translation, the Western blotting analysis was carried out to examine the levels of seven mtDNA encoding polypeptides in mutant and control cells with VDAC as a loading control. As shown in Fig. 6A, the levels of CO2 (subunit II of cytochrome c oxidase); ND1, ND4, ND5, and ND6 (subunits 1, 4, 5, and 6 of NADH dehydrogenase); A6 (subunit 6 of the H⁺-ATPase); and CYTB (apocytochrome b) were decreased in mutant cell lines, as compared with those control cell lines. As shown in Fig. 6B, the overall levels of seven mitochondrial translation products in mutant cell lines carrying both m.1555A→G and homozygous TRMU A10S mutations were 42%, relative to the mean value measured in the control cell lines. Notably, the average levels of ND1, ND4, ND5, ND6, CO2, A6, and CYTB in these mutant cells carrying both m.1555A→G and homozygous TRMU A10S mutations were 36, 40, 53, 24, 54, 51, and 37% of the average values of control cells, respectively. Furthermore, the overall levels of seven mitochondrial proteins in cell lines carrying both m.1555A→G and heterozygous TRMU A10S mutations were 55%, relative to those in controls. Moreover, the overall levels of seven mitochondrial translation products in mutant cell line carrying only m.1555A→G mutation were 73% of control cell lines. However, the levels of synthesis of polypeptides in mutants relative to that in controls did not correlate with either the number of codons or proportion of glutamic acid, glutamine, and lysine residues (supplemental Table S2).

FIGURE 6.

Western blotting analysis of mitochondrial proteins. A, 20 μg of total cellular proteins from lymphoblastoid cell lines were electrophoresed through a SDS-polyacrylamide gel, electroblotted, and hybridized with seven respiratory complex subunits in mutant and control cells with VADC as a loading control. CO2, subunit II of cytochrome c oxidase; ND1, ND4, ND5, and ND6, subunits 1, 4, 5, and 6 of the reduced nicotinamide-adenine dinucleotide dehydrogenase; A6, subunit 6 of the H⁺-ATPase; and CYTB, apocytochrome b. B, quantification of mitochondrial protein levels. Average content of CO2, ND1, ND4, ND5, ND6, A6 and CYTB per cell, normalized to the average content of VADC per cell in mutant cell lines and controls. The values for the mutant cell lines are expressed as percentages of the average values for the control cell lines. The horizontal dashed lines represent the average value for each group. The calculations were based on three independent determinations. Graph details and symbols are explained in the legend to Fig. 4.

Respiration Defects

To evaluate whether the TRMU A10S mutation affects cellular bioenergetics, we examined the oxygen consumption rates (OCRs) of mutant and control cell lines (50). As shown in Fig. 7, the average basal OCRs in mutant cell lines carrying both m.1555A→G and homozygous or heterozygous TRMU A10S mutations, and only m.1555A→G mutation were 54, 76, and 88%, relative to the mean value measured in the control cell lines, respectively. To investigate which of the enzyme complexes of the respiratory chain was affected in the mutant cell lines, OCR were measured after the sequential addition of oligomycin (inhibit the ATP synthase), FCCP (to uncouple the mitochondrial inner membrane and allow for maximum electron flux through the ETC), rotenone (to inhibit complex I), and antimycin A (to inhibit complex III). The difference between the basal OCR and the drug-insensitive OCR yields the amount of ATP-linked OCR, proton leak OCR, maximal OCR, reserve capacity, and non-mitochondrial OCR. As shown in Fig. 6, the ATP-linked OCR, proton leak OCR, maximal OCR, reserve capacity, and non-mitochondrial OCR were 51, 53, 53, 63, and 67% in the mutant cell lines carrying both m.1555A→G and homozygous TRMU A10S mutations and 68, 78, 78, 85, and 84% in the mutant cell lines carrying both m.1555A→G and heterozygous TRMU A10S mutations relative to the control cell lines, respectively. Moreover, the above five kinds of OCR levels in mutant cell line carrying only m.1555A→G mutation were 86, 92, 89, 91, and 88% of control cell lines.

FIGURE 7.

Respiration assays. A, an analysis of O₂ consumption in the various cell lines using different inhibitors. The OCRs were first measured on 5 × 10⁴ cells of each cell line under basal condition and then sequentially added to oligomycin (1.5 μm), carbonyl cyanide p-(trifluoromethoxy) phenylhydrazone (FCCP) (0.8 μm), rotenone (1 μm), and antimycin A (5 μm) at indicated times to determine different parameters of mitochondrial functions. B, graphs presented the ATP-linked OCR, proton leak OCR, maximal OCR, reserve capacity, and non-mitochondrial OCR in mutant and control cell lines. Non-mitochondrial OCR was determined as the OCR after rotenone/antimycin A treatment. Basal OCR was determined as OCR before oligomycin minus OCR after rotenone/antimycin. ATP-linked OCR was determined as OCR before oligomycin minus OCR after oligomycin. Proton leak was determined as basal OCR minus ATP-linked OCR. Maximal OCR was determined as the OCR after FCCP minus non-mitochondrial OCR. Reserve capacity was defined as the difference between maximal OCR after FCCP minus basal OCR. The average of four determinations for each cell line is shown. The horizontal dashed lines represent the average value for each group. Graph details and symbols are explained in the legend to Fig. 4.

Reduced Levels in Mitochondrial ATP Production

The capacity of oxidative phosphorylation in mutant and wild-type cells was examined by measuring the levels of cellular and mitochondrial ATP using a luciferin/luciferase assay. Populations of cells were incubated in the media in the presence of glucose, and 2-deoxy-d-glucose with pyruvate (15). As shown in Fig. 8A, in the presence of glucose (total cellular levels of ATP), the average levels of ATP production in mutant cells carrying both m.1555A→G and homozygous or heterozygous TRMU A10S mutations and only m.1555A→G mutation were 93, 95, and 85%, relative to the mean value measured in the control cell lines, respectively. In the presence of pyruvate and 2-deoxy-d-glucose to inhibit the glycolysis (mitochondrial levels of ATP), as shown in Fig. 8B, the levels of ATP production in mutant cell lines carrying both m.1555A→G and homozygous or heterozygous TRMU A10S mutations and only m.1555A→G mutation were 48, 61, and 83% of the mean value measured in the control cell lines, respectively.

FIGURE 8.

Measurement of cellular and mitochondrial ATP levels using bioluminescence assay. The cells were incubated with 10 mm glucose or 5 mm 2-deoxy-d-glucose plus 5 mm pyruvate to determine ATP generation under mitochondrial ATP synthesis. The average rates of ATP level per cell line are shown. A, ATP level in total cells. B, ATP level in mitochondria. Six to seven determinations were made for each cell line. Graph details and symbols are explained in the legend to Fig. 4.

The Increase of ROS Production

It was anticipated that respiration defects increase the production of ROS. The levels of the ROS generation in the vital cells derived from six mutant cell carrying the m.1555A→G mutation with or without TRMU A10S mutation and two control cell lines lacking both mutations were measured with flow cytometry under normal and H₂O₂ stimulation (51, 52). Geometric mean intensity was recorded to measure the rate of ROS of each sample. The ratio of geometric mean intensity between unstimulated and stimulated with H₂O₂ in each cell line was calculated to delineate the reaction upon increasing level of ROS under oxidative stress. As shown in Fig. 9, the levels of ROS generation in the mutant cell lines cells carrying both m.1555A→G and homozygous TRMU A10S mutations ranged from 155 to 167%, with an average 161% of the mean value measured in the control cell lines. Moreover, ROS generation levels of cell lines carrying both m.1555A→G and heterozygous TRMU A10S mutations and mutant cells carrying only m.1555A→G mutation were 132 and 120% of controls, respectively.

FIGURE 9.

Ratio of geometric mean intensity between levels of the ROS generation in the vital cells with or without H₂O₂ stimulation. The rates of production in ROS from mutant cell lines and control cell lines were analyzed by BD-LSR II flow cytometer system with or without H₂O₂ stimulation. The relative ratio of intensity (stimulated versus unstimulated with H₂O₂) was calculated. The average of four determinations for each cell line is shown. Graph details and symbols are explained in the legend to Fig. 4.

Discussion

The nuclear modifier genes were proposed to modulate the phenotypic manifestation of deafness-associated 12S rRNA mutations (3, 4, 53). In the present study, we further characterized the nuclear modifier allele (A10S) in the TRMU, which interacts with m.1555A→G mutation to cause deafness. Human TRMU encodes a highly conserved 5-methylaminomethyl-2-thiouridylate-methyltransferase responsible for the biosynthesis of 5-taurinomethyl-2-thiouridine (τm⁵s²U) nucleotides at the wobble position of mitochondrial tRNA^Gln, tRNA^Glu, and tRNA^Lys (30, 31). The highly conserved Ala¹⁰ residue locates at the N-terminal region of this polypeptide. In all available TRMU/MnmA protein sequences, residue 10 is highly conserved, either Ala or Gly. Based on our MD simulation results, either Ala¹⁰ or Gly¹⁰ does not interact with helix 4 of the protein (43). The change of alanine 10 residue with serine introduces the Ser¹⁰ dynamic electrostatic interaction with the Lys¹⁰⁶ residue of helix 4 within the catalytic domain of TRMU. Thus, it was hypothesized that the A10S mutation altered the stability and catalytic activity of TRMU. A Western blotting analysis showed markedly reduced levels of TRMU in cell lines carrying the A10S mutation. Furthermore, the thermal shift assay revealed that the T_m value of mutant TRMU protein was lower than those of wild-type TRMU. These data are strong evidences that A10S mutation caused the instability of TRMU.

The primary defect in the A10S mutation was the deficient 2-thiouridine modification of U₃₄ of tRNA^Lys, tRNA^Glu, and tRNA^Gln. In the present study, 48, 38, and 50% decreases in 2-thiouridine modification of U₃₄ of tRNA^Lys, tRNA^Glu, and tRNA^Gln were observed in mutant cell lines carrying both m.1555A→G and homozygous A10S mutations, as compared with controls. These results were consistent with the fact that the small interfering RNA down-regulation of or other mutations of TRMU led to the defects in 2-thiouridylation in mitochondrial tRNA^Lys, tRNA^Glu, and tRNA^Gln (34, 54, 55). Furthermore, in vitro assays showed that the deficient synthesis of s²U₃₄ altered the tRNA aminoacylation, because s²U₃₄ serves as a determinant for tRNA recognition by cognate aminoacyl-tRNA synthetases in bacteria (39, 56). However, in vivo assays revealed that the lack of 5-methyl-aminomethyl group did not affect the charging levels for tRNA^Lys, tRNA^Glu, and tRNA^Gln in bacteria (56). In the present study, the efficiencies of aminoacylated tRNA^Lys, tRNA^Leu(CUN), and tRNA^Ser(AGY) in cell lines carrying both m.1555A→G and homozygous TRMU A10S mutations were 137, 135, and 129% of those in control cell lines. These data suggested that the TRMU A10S mutation may increase the charging levels for tRNA^Lys, tRNA^Leu(CUN), and tRNA^Ser(AGY). An increase in aminoacylation of tRNAs in mutant cell lines may be due to the instability of the mutant tRNA, where aminoacylation may provide some levels of stabilization by compensatory effect (15, 57, 58). Alternatively, these tRNAs may be mischarged with a noncognate amino acids, because anticodon modifications act as antideterminants (59). Therefore, an inefficient modification of tRNAs caused by the TRMU mutations may then make these tRNAs to be metabolically less stable and more subject to degradation, thereby lowering the level of the tRNAs (25, 35, 54, 55).

A failure in the tRNA metabolism caused by the TRMU A10S mutation should be responsible for the impairment of mitochondrial translation. In particular, the mischarged tRNAs may cause the global protein misfolding (32, 59). In fact, the mtDNA encoded 13 polypeptides in the complexes of the oxidative phosphorylation system (ND1–6; ND4L of complex I; CYTB of complex III; CO1, CO2, and CO3 of complex IV; and ATP6 and ATP8 of complex V) (6, 60). In the present study, 58, 45, and 27% reductions in the levels of mitochondrial proteins were observed in mutant cell lines carrying both m.1555A→G and homozygous or heterozygous TRMU A10S mutations and only m.1555A→G mutation, respectively. These results were comparable with the in vivo pulse-labeling mitochondrial protein synthesis assay (21, 25). Notably, variable decreases in the levels of seven mtDNA-encoded polypeptides were observed in mutant cell lines. In particular, cell lines carrying both m.1555A→G and homozygous TRMU A10S mutations exhibited marked reductions (from 46 to 76%) in the levels of seven polypeptides. However, the levels of synthesis of polypeptides in mutants relative to that in controls did not correlate with either the number of codons or proportion of glutamic acid, glutamine, and lysine residues. These data were not fully comparable with the case of MERRF-associated m.8344A→G mutation in tRNA^Lys gene (61). The impairment of mitochondrial protein synthesis was apparently responsible for the reduced rates in the basal OCR or ATP-linked OCR reserve capacity and maximal OCR among the control and mutant cell lines. In particular, the 12, 34, and 46% decreases in basal OCR were observed in cell lines harboring only m.1555A→G mutation, both m.1555A→G and heterozygous or homozygous TRMU A10S mutations, respectively. This correlation is clearly consistent with the importance that a failure in tRNA metabolism plays a critical role in producing their respiration defects in deafness patients carrying the m.1555A→G mutation.

The respiratory deficiency then affects the efficiency of mitochondrial ATP synthesis. In this investigation, the m.1555A→G mutation caused 17% reduction of mitochondrial ATP production in lymphoblastoid cell lines, as in the cases of cell lines bearing the LHON-associated m.11778G→A mutations (62, 63). By contrast, the ∼52% drop in mitochondrial ATP production in lymphoblastoid cell lines bearing both m.1555A→G and homozygous TRMU A10S mutations may result from the defective activities of respirations caused by both m.1555A→G mutation and altered tRNA metabolism associated with TRMU A10S mutation. These data are consistent with the fact that 20 and 51% reductions in mitochondrial ATP production were observed in lymphoblastoid cell lines carrying the only m.11778G→A and both m.11778G→A and homozygous YARS2 p.191Gly→Val mutations (62). Alternatively, the reduction in mitochondrial ATP production in mutant cells was likely a consequence of the decrease in the proton electrochemical potential gradient of mutant mitochondria (64). As a result, the hair cells carrying the mtDNA mutation may be particularly sensitive to increased ATP demand (3, 4, 65). The impairment of oxidative phosphorylation can lead to more electron leakage from electron transport chain and, in turn, elevate the production of ROS in mutant cells (66), thereby damaging mitochondrial and cellular proteins, lipids, and nuclear acids (67). However, a 61% increase of ROS production in cells carrying both m.1555A→G and homozygous TRMU A10S mutations was the consequence of the altered activities of respiration. The hair cells and cochlear neurons may be preferentially involved because they are somehow exquisitely sensitive to subtle imbalance in cellular redox state or increased level of free radicals (68–70). This would lead to dysfunction or apoptosis of hair cells and cochlear neurons carrying both m.1555A→G and TRMU A10S mutations, thereby producing a phenotype of deafness.

In summary, our study demonstrated the role of the first nuclear modifier allele (A10S) in the TRMU gene in the phenotypic manifestation of deafness-associated m.1555A→G mutation. The A10S mutation altered the structure and function of TRMU. The mutated TRMU caused the deficient thiolation of tRNA^Gln, tRNA^Glu, and tRNA^Lys but increased the aminoacylation of tRNAs. The failures in tRNA metabolism led to impairment of mitochondrial translation, respiratory phenotype, defects in mitochondrial ATP production, and increasing ROS production. The resultant biochemical defects aggravate the mitochondrial dysfunction associated with m.1555A→G mutation, below the threshold for normal cell function, thereby expressing the deafness phenotype. Therefore, the mutated TRMU, acting as a nuclear modifier, triggers the deafness in individuals harboring the m.1555A→G mutation.

Experimental Procedures

MD Simulations

Simulation systems. The starting coordinates of wild-type TRMU were taken from the crystal structure of TRMU-tRNA^Glu complex (Protein Data Bank entry 2DER) (44). The coordinates of A10S mutation mutated protein were generated from the wild-type TRMU coordinates through PyMOL (Schrödinger). All terminal residues adopted the neutral state. Each system was solvated in a cubic box of TIP3P water with an extension of at least 10 Å from each side. Approximately 50 mm NaCl were added to the solvent in addition to the neutralizing Na⁺ or Cl⁻. This leads to a wild-type system of 109422 atoms and a mutant system of 109426 atoms.

Simulation protocol. MD simulations were carried out with the GROMACS 4.5.5 package 2 (42). The CHARMM363 force field with CMAP modification 4 was applied for protein (43). Energy minimizations were performed to relieve unfavorable contacts, followed with equilibration steps to fully equilibrate the solvent. Each system was equilibrated in the NPT ensemble at 310 K and 1 bar in periodic boundary condition, and the time step was set to 1 fs. Positional restraints were first applied on all the heavy atoms of protein for 50 ps, then main chain atoms for 50 ps, and then α-atoms for 500 ps. After equilibrations, the production simulation was carried out with a time step of 2 fs, and each system was run up to 300 ns. Electrostatic interactions were calculated with the particle mesh Ewald algorithm (71). SETTLE constraint was applied on hydrogen-involved covalent bonds in water, and LINCS constraint was applied on the hydrogen-involved covalent bonds in those molecules other than waters in the system (72).

Cell Lines and Culture Conditions

Eight human immortalized lymphoblastoid cell lines derived from members of an Arab-Israeli family (two subjects carrying only m.1555A→G mutation (F12H and F6D), two individuals (F20C and F8A) harboring both m.1555A→G and heterozygous TRMU A10S mutations, two individuals (F20A and F20D) carrying both m.1555A→G and homozygous TRMU A10S mutations, and two control individuals (F7A and F7E) lacking both mutations) were cultured in RPMI 1640 medium supplemented with 1 mm sodium pyruvate and 10% FBS (15, 21).

Differential Scanning Fluorimetry

The wild-type and mutant human TRMU cDNAs were amplified by PCR and cloned in-frame with a C-terminal His tag into pET-28a vector (25). Recombinant wild-type and mutant TRMU were produced as His tag fusion proteins in Escherichia coli Rosetta (DE3) as detailed elsewhere (73). The proteins were purified using a nickel-nitrilotriacetic acid column (Qiagen). The stability of proteins was assessed using a protein thermal shift dye kit (Life Technologies) unfolding temperature (T_m) test performed on a 7900 HT Fast real time polymerase chain reaction detection system, according to the modified manufacturer's instructions. The data were analyzed as detailed elsewhere (44). The T_m value was estimated from the transition midpoint of the fluorescence curve, which corresponds to temperature at which half of the protein population is unfolded.

Mitochondrial tRNA Thiolation Analysis

Total mitochondrial RNAs were obtained using a Totally RNA^TM kit (Ambion) from mitochondria isolated from mutant and wild-type cell lines, as described previously (74). The presence of the thiouridine modification in the tRNAs was verified by the retardation of electrophoretic mobility in a polyacrylamide gel that contains 0.05 mg/ml APM (45–47). 2 μg of total mitochondrial RNA was separated by polyacrylamide gel electrophoresis and blotted onto positively charged membrane (Roche Applied Science). Each tRNA fraction was detected with a specific non-radioactive DIG oligodeoxynucleoside probe at the 3′ termini according to the method as described elsewhere (15, 25, 47). Oligonucleotide probes for tRNA^Lys, tRNA^Glu, and tRNA^Gln were detailed previously (15, 25). DIG-labeled oligodeoxynucleosides were generated by using the DIG oligonucleoside tailing kit (Roche). APM gel electrophoresis and quantification of 2-thiouridine modification in tRNAs were conducted as detailed (25).

Mitochondrial tRNA Aminoacylation Analysis

Total mitochondrial RNAs were isolated under acidic condition. 2 μg of total mitochondrial RNA was electrophoresed at 4 °C through an acid (pH 5.2) 10% polyacrylamide with 7 m urea gel to separate the charged and uncharged tRNA as detailed elsewhere (48). The gels were then electroblotted onto a positively charged nylon membrane (Roche) for the hybridization analysis with oligodeoxynucleotide probes of tRNA^Lys, tRNA^Tyr, tRNA^Leu(CUN), and tRNA^Ser(AGY), as described elsewhere (25, 48–49). DIG-labeled oligodeoxynucleotides were generated by using a DIG oligonucleotide tailing kit (Roche). The hybridization and quantification of density in each band were carried out as detailed elsewhere (48, 49).

Western Blotting Analysis

Western blotting analysis was performed as detailed previously (15, 49). The antibodies used for this investigation were from Abcam (TRMU (ab50895), VDAC (ab14734), NDUFB8 (ab110411), ND1 (ab74257), ND5 (ab92624) and A6 (ab101908), and CO2 (ab110258)), Santa Cruz Biotechnology (ND4 (sc-20499-R), ND6 (sc-20667)), MTO1 (sc-398760), and Proteintech (CYTB (55090-1-AP)). Peroxidase Affini Pure goat anti-mouse IgG and goat anti-rabbit IgG (Jackson) were used as secondary antibodies, and protein signals were detected using the ECL system (CWBIO). Quantification of density in each band was performed as detailed previously (15, 49).

Measurements of Oxygen Consumption

The rates of oxygen consumption in cybrid cell lines were measured with a Seahorse Bioscience XF-96 extracellular flux analyzer (Seahorse Bioscience), as detailed previously (15, 50).

ATP Measurements

The Cell Titer-Glo® luminescent cell viability assay kit (Promega) was used for the measurement of cellular and mitochondrial ATP levels, according to the modified manufacturer's instructions (15, 49, 62).

Measurement of ROS Production

ROS measurements were performed following the procedures detailed previously (51–52).

Computer Analysis

Statistical analysis was carried out using the unpaired, two-tailed Student's t test contained in the Microsoft Excel program for Macintosh (version 2007). Differences were considered significant at a p < 0.05.

Author Contributions

F. M. performed the experiments and contributed to data analysis in Figs. 2, 3, and 6–9. X. C. and Z. Z. carried out the MD simulations. Y. P. performed the aminoacylation experiment. R. L. did the thiolation analysis. F. L. and Q. F. contributed to the Western blotting analysis. A. S. G. performed the statistical analysis. N. F.-G. provided the cell lines. M.-X. G. designed the experiments, and M.-X. G. and X. Z. monitored the project progression, data analysis, and interpretation. F. M. prepared the initial draft of the manuscript. M.-X. G. made the final version of the manuscript.