The genetic and epigenetic factors underlying the variable penetrance of homoplasmic mitochondrial DNA mutations are poorly understood. We investigated a 16-year-old patient with hypertrophic cardiomyopathy harboring a homoplasmic m.4277T>C mutation in the mt-tRNAIle (MTTI) gene. Skeletal muscle showed multiple respiratory chain enzyme abnormalities and a decreased steady-state level of the mutated mt-tRNAIle. Transmitochondrial cybrids grown on galactose medium demonstrated a functional effect of this mutation on cell viability, confirming pathogenicity. These findings were reproduced in transmitochondrial cybrids, harboring a previously described homoplasmic m.4300A>G MTTI mutation. The pathogenic role of the m.4277T>C mutation may be ascribed to misfolding of the mt-tRNA molecule, as demonstrated by the altered electrophoretic migration of the mutated mt-tRNA. Indeed, structure and sequence analyses suggest that thymidine at position 4277 of mt-tRNAIle is involved in a conserved tertiary interaction with thymidine at position 4306. Interestingly, the mutation showed variable penetrance within family members, with skeletal muscle from the patient's clinically unaffected mother demonstrating normal muscle respiratory chain activities and steady-state levels of mt-tRNAIle, while homoplasmic for the m.4277T>C mutation. Analysis of mitochondrial isoleucyl-tRNA synthetase revealed significantly higher expression levels in skeletal muscle and fibroblasts of the unaffected mother when compared with the proband, while the transient over-expression of the IARS2 gene in patient transmitochondrial cybrids improved cell viability. This is the first observation that constitutively high levels of aminoacyl-tRNA synthetases (aaRSs) in human tissues prevent the phenotypic expression of a homoplasmic mt-tRNA point mutation. These findings extend previous observations on aaRSs therapeutic effects in yeast and human.
Mutations in the human mitochondrial genome [mitochondrial DNA (mtDNA)] cause a variety of clinical disorders, whose main unifying feature is an alteration in energy homeostasis due to decreased ATP production. Highly energy-dependent tissues such as central nervous system and skeletal and cardiac muscles are commonly involved either as multisystem or as isolated organ disease (1). Pathogenic mtDNA mutations can be heteroplasmic (e.g. mutated and non-mutated mtDNA molecules coexist within the same cell) or homoplasmic (all mtDNA molecules within an organism's cells are mutated). Homoplasmic mtDNA mutations are generally associated with tissue-specific disorders, including Leber's hereditary optic neuropathy (2), mitochondrial non-syndromic sensorineural hearing loss (3) and maternally inherited hypertrophic cardiomyopathy (MIC) (4). The latter is typically associated with mutations in mt-tRNA genes, with mt-tRNAIle being considered a ‘hot spot’ for isolated cardiac disease (MITOMAP: a Human Mitochondrial Genome Database, http://www.mitomap.org; 2011). In addition, homoplasmic mt-tRNA mutations have been described in families associated with a range of devastating multisystem disorders (5,6). An intriguing feature of all homoplasmic mtDNA mutations is their extremely variable penetrance, even within members of the same family. This phenomenon may be explained by differences in the efficiency of compensatory mechanisms such as mitochondrial biogenesis, which has been linked to the generation of reactive oxygen species (ROS) due to impaired oxidative phosphorylation (OXPHOS) (7). In addition, mtDNA haplotype (8–11), nuclear modifiers and epigenetic or environmental factors have been hypothesized to play a role in the severity and tissue specificity of homoplasmic mt-tRNA mutations (12–15).
To establish the pathogenicity of homoplasmic mutations, stringent criteria including unequivocal evidence linking the genetic defect with molecular dysfunction are required (16,17). Although some criteria have been proposed, pathogenicity may be difficult to prove, due to the absence of segregation of the mutation with the biochemical or clinical defect. In addition, fibroblasts and transmitochondrial cybrids from patients with tissue-specific disease may fail to show a clear biochemical phenotype (4,14), and the clinically affected tissue [e.g. cardiac muscle in patients with hypertrophic cardiomyopathy (HCM)] is often unavailable for functional analyses. In the present study, we report a further case of a homoplasmic mutation in the mitochondrial tRNAIle (MTTI) gene associated with HCM. By introducing mitochondria bearing the m.4277T>C mutation into transmitochondrial cybrids and forcing cells to utilize OXPHOS, we were able to demonstrate a functional effect on cell viability and energetic competence, reproducing these observations in transmitochondrial cybrids carrying the homoplasmic m.4300A>G MTTI mutation (4). We then investigated the effects of the mutation on the structure and aminoacylation of the mutated mt-tRNAIle molecule. Finally, we provide evidence that the expression levels of the cognate isoleucyl-tRNA synthetase (IleRS) can modulate the phenotypic expression of the homoplasmic m.4277T>C mt-tRNAIle mutation.
The index case is a 16-year-old male, born to non-consanguineous healthy parents. Pregnancy, birth and early psychomotor development were normal. He showed progressive hearing impairment at the age of 2 years and was later diagnosed with HCM at the age of 12 years. There was no family history of HCM or premature sudden death. On examination, he was mildly symptomatic for fatigue and exertional dyspnea (New York Heart Association functional class II), but had never experienced syncope. His blood pressure levels were normal. Electrocardiogram showed sinus rhythm, with short P-R interval, left ventricular (LV) hypertrophy and deep negative T waves in V3–V6. Two-dimensional echocardiography revealed a mildly dilated LV with symmetric hypertrophy (maximal thickness 27 mm at the inferolateral wall, Fig. 1A) and a slightly reduced ejection fraction (about 45–50%). Cardiac magnetic resonance confirmed the anatomical and functional features of the echocardiogram. The contrast study with gadolinium showed no areas of late enhancement, suggestive of fibrosis (data not shown). Based on cardiac imaging, a non-sarcomeric form of HCM was suspected. Repeated blood analyses revealed mildly increased serum creatine kinase levels (324 IU/l, normal range 22–269 IU/l). General neurological examination was normal. He had no evidence of ptosis or external ophthalmoplegia. A skeletal muscle biopsy revealed numerous cytochrome c oxidase (COX)-deficient fibers on sequential COX–succinate dehydrogenase (SDH) histochemistry (Fig. 1B), with numerous fibers showing subsarcolemmal accumulation of mitochondria typical of ‘ragged-red’ changes; skeletal muscle biopsy of his clinically unaffected mother was normal. A follow-up evaluation at the age of 19 years revealed that the patient had developed symptoms of advanced congestive heart failure, later requiring two hospital admissions. Echocardiogram showed a further decrease in LV ejection fraction (about 25%) and mild right ventricular involvement. A cardioverter defibrillator was implanted in primary prevention, and at present, the patient is on a waiting list for cardiac transplant.
Identification of the homoplasmic m.4277T>C mutation in mt-tRNAIle
Based on the histochemical demonstration of a mosaic COX deficiency in the proband's skeletal muscle, respiratory chain biochemical studies of muscle homogenate were undertaken, revealing a combined respiratory chain defect involving complexes I, III and IV, all of which contain mtDNA-encoded subunits. In contrast, the activity of complex II, which consists entirely of nuclear-encoded subunits, was normal, prompting us to search for possible mtDNA abnormalities. Respiratory chain enzyme activities in skeletal muscle from the mother were unremarkable (Table 1). A screen for mtDNA rearrangements in muscle DNA proved to be negative, whereas a significant increase in the mtDNA amount was found (Fig. 1C). This was paralleled by the up-regulation of the peroxisome proliferator-activated receptor gamma, coactivator 1 alpha (PPARGC1A) gene, consistent with the induction of mitochondrial biogenesis (Fig. 1D). These features were not observed in the skeletal muscle from the unaffected mother (Fig. 1C and D). Sequence analysis of the entire mitochondrial genome in the patient's muscle revealed an m.4277T>C variant within the dihydrouridine (DHU) loop of the mt-tRNAIle (MTTI) gene (Fig. 2A and B), a region that is not strongly conserved phylogenetically. In addition to the m.4277T>C base change that best fitted the observed combined respiratory chain enzyme defect, we noted 13 homoplasmic reported polymorphisms (MITOMAP: a Human Mitochondrial Genome Database, http://www.mitomap.org; May 2011), which allowed us to assign the mtDNA to haplogroup H (GenBank accession no. NC012920) (http://www.phylotree.org; mtDNA tree Build 11; February 7, 2011) (18), and four unreported homoplasmic variants (Supplementary Material, Table S1). Identical homoplasmic changes were found in mtDNA from peripheral lymphocytes of the patient and in mtDNA from the skeletal muscle and peripheral lymphocytes of his clinically unaffected mother. Last hot cycle-restriction fragment length polymorphism (RFLP) analysis confirmed that the m.4277T>C mutation was homoplasmic in skeletal muscle from both the patient and his mother (Fig. 2C). Homoplasmic levels of the m.4277T>C mutation were also demonstrated in peripheral lymphocytes obtained from the patient and in skin-derived fibroblasts from the patient and his mother (Supplementary Material, Fig. S1).
|Complex I/CS||Complex II/CS||Complex III/CS||Complex IV/CS|
|Controls (n = 25)||0.104 ± 0.036||0.145 ± 0.047||0.554 ± 0.345||1.124 ± 0.511|
|Complex I/CS||Complex II/CS||Complex III/CS||Complex IV/CS|
|Controls (n = 25)||0.104 ± 0.036||0.145 ± 0.047||0.554 ± 0.345||1.124 ± 0.511|
Enzyme activities are expressed as nmol NADH oxidized per minute per unit citrate synthase (CS) for complex I, nmol DCPIP reduced per minute per unit CS for complex II (succinate: ubiquinone-1 reductase) and the apparent first-order rate constant per second per unit CS for complexes III and IV (×103).
Control values are shown as mean ± SD.
DCPIP, 2,6-dichlorophenol-indophenol; SD, standard deviation.
Decreased steady-state level of mutated mt-tRNAIle in skeletal muscle from the proband
To evaluate the effect of the m.4277T>C mutation on mt-tRNAIle stability, we determined the steady-state level of mt-tRNAIle by high-resolution northern blot analysis, relative to the amount of mt-tRNALeu(UUR) as control. A selective 60–70% reduction in mt-tRNAIle levels was observed in the skeletal muscle of the proband when compared with normal, age-matched controls. Interestingly, the steady-state levels of mutated mt-tRNAIle in the skeletal muscle of the proband's mother were unchanged (Fig. 2D), in accordance with the normal respiratory chain enzyme activities (histochemically and biochemically).
Reduced viability and energetic competence of mutant transmitochondrial cybrids in galactose medium
To shed light on the role of the m.4277T>C mutation in determining the observed biochemical and molecular phenotypes, we undertook further experiments using transmitochondrial cybrids obtained by fusing cytoplasts derived from the patient and his unaffected mother with mtDNA-less (ρ°) 143B.206 human osteosarcoma cells. We performed all the experiments on two to three selected clones from each transmitochondrial cybrid cell line (Supplementary Material, Table S2) both under basal conditions (glucose medium) and in glucose-free galactose medium. The latter is a well-established method to reveal an OXPHOS defect by forcing cells to rely on the mitochondrial respiratory chain for ATP synthesis (19). Under this condition, cells increase ROS production with consequent induction of mitochondrial biogenesis and, possibly, reduced cell viability due to an increased rate of apoptotic cell death (8,20,21).
Under basal conditions, no significant differences in ROS levels, rate of oxygen consumption and growth capacity were observed in the m.4277T>C mutant cells when compared with control transmitochondrial cybrids (data not shown). Interestingly, a slight—albeit not statistically significant—increase in the mtDNA amount was observed in mutant cells when compared with controls after 24 h incubation in glucose medium (Fig. 3A). After 1 h incubation in galactose, all transmitochondrial cybrids showed an increase in the steady-state levels of ROS (Fig. 3B). This increase was significantly higher in 4277-mutant clones (n= 5) when compared with controls (n= 4) and was associated with the induction of mitochondrial biogenesis, as demonstrated by the increase in the mtDNA content per cell, both in m.4277T>C and in controls (Fig. 3A). These findings were reproduced in transmitochondrial cybrids (n= 4) obtained from two patients (IV-09 and V-14 in 4) harboring the homoplasmic m.4300A>G MTTI cardiomyopathy mutation.
Next, we evaluated transmitochondrial cybrid viability in galactose medium, showing a significant decrease in the m.4277T>C and m.4300A>G mutant transmitochondrial cybrid cell number after 48 h incubation when compared with controls (Fig. 3C). To examine whether decreased viability was related to apoptosis, we labeled cells with annexin V, which binds phosphatidylserine exposed on the cytoplasmic surface of the cell membrane of apoptotic cells (22). After incubation in galactose medium for 24 h, a significant number of mutant transmitochondrial cybrids shifted from being viable to an apoptotic state (Fig. 3D and Supplementary Material, Fig. S2). Interestingly, a slight but significant percentage of m.4300A>G cells shifted toward apoptosis after 24 h incubation in glucose medium (Fig. 3D). In keeping with these findings, both m.4277T>C and m.4300A>G transmitochondrial cybrids showed a marked decrease in the rate of oxygen consumption after 12 h incubation in galactose medium (Fig. 3E).
Decreased steady-state levels and abnormal electrophoretic mobility of mutated mt-tRNAIle in cultured cells
When compared with the m.4277T>C proband skeletal muscle, a weaker decrease of mutated mt-tRNAIle steady-state levels [up to 40% when normalized to mt-tRNALeu(UUR)] was observed in cultured cells. This is in keeping with previous reports, showing that the very low levels of mutated tRNAs in tissues carrying homoplasmic, pathogenic mt-tRNA mutations are not recapitulated in cell lines (23). Interestingly, a further mild (up to 15%) decrease was observed in transmitochondrial cybrids after 24 h incubation in galactose medium (Fig. 4A). In addition, by using a partially denaturing polyacrylamide sequencing gel, we were able to demonstrate that both the acylated and non-acylated forms of mutated mt-tRNAIle had a slower electrophoretic migration when compared with controls (Fig. 4B). This may be ascribed to an alteration in the conformation of the mutated mt-tRNAIle with respect to the wild-type molecule under partially denaturing conditions. A similar alteration has previously been observed in several other mutated tRNAs and found to disappear under completely denaturing conditions (24). No isoleucylation impairment was observed in transmitochondrial cybrids grown either in glucose (Fig. 4B) or in galactose medium (data not shown).
The m.4277T>C mutation affects the conserved 15–48 mt-tRNAIle tertiary interaction
High sequence conservation at a site affected by a mutation is among the commonly accepted criterion to establish pathogenicity; however, some genuinely pathological mutations escape this criterion (16). To investigate the potential effect of the m.4277T>C mutation on the three-dimensional structure of human mt-tRNAIle, we analyzed several representative three-dimensional structures of tRNA molecules determined by X-ray crystallography and available from the Protein Data Bank (PDB) (http://www.rcsb.org) (25). In addition, we studied the tRNA sequences from: (i) all species and cellular compartments (26) and (ii) mammalian mitochondria [(27), see Materials and Methods for details]. Both structure and sequence analyses support the hypothesis that position m.4277T>C in human mt-tRNAIle is equivalent to position 15 of conventional tRNA numbering. Indeed, the nucleotide (nt) A4276 is equivalent to the almost universally conserved A14, which is involved in the conserved tertiary interaction U8-A14-A21. The adjacent T4277 is, therefore, likely to assume a position structurally correspondent to that of nt 15 in three-dimensional tRNA structures.
In all the analyzed tRNA structures (n= 7), the nt at position 15 of the D-loop interacts with the nt at position 48 of the V-loop (Fig. 4C) and, in one structure, with G59 in the T-loop. Although the D-arm of human mt-tRNAIle is shorter (12 nt) than that observed in known mt-tRNA structures (comprising 15, 16 or 17 nt), the conservation of the 15–48 interaction observed in known structures (Fig. 4C) supports the hypothesis that the same interaction occurs between T4277 and T4306 (equivalent to position 48) of mt-tRNAIle and plays a role in the maintenance of the native structure.
Analysis of mt-tRNA gene sequences available from the two databases (26,27) showed that T occurs at least twice more frequently than C at position 15, although pyrimidines are about 4-fold less frequent than purines. While in both databases the T–T and C–T pairs represent collectively only ∼10% of the combinations occurring at positions 15–48, the T–T nt pair occurs 3- and 1.5-fold more frequently than C–T, respectively.
Results observed in transmitochondrial cybrids are reproduced in skin-derived fibroblasts from the proband
Next, we compared the viability of skin-derived fibroblasts from the m.4277T>C proband and his mother. As controls, we used skin-derived fibroblasts from age-matched individuals (n = 3). In addition, we analyzed two affected patients bearing the previously described homoplasmic m.4300A>G mutation in mt-tRNAIle (IV-09 and V-14 in 4). There were no differences in growth capability of controls and mutant fibroblasts maintained in glucose medium (data not shown). However, after 96 h in galactose medium, only fibroblasts from the m.4277T>C proband and the m.4300A>G patients showed a significant decrease in viability (Fig. 5A). Consistent with this finding, after 72 h in galactose medium, mutant fibroblasts showed a significant increase in the rate of apoptosis (Fig. 5B and Supplementary Material, Fig. S3). Interestingly, most of the fibroblasts bearing the m.4300A>G mutation showed a combined stain for annexin V and propidium iodine, consistent with a late apoptotic phase. Fibroblasts from the unaffected mother of the m.4277T>C mutation proband showed growth capacity and apoptotic rate similar to controls.
Expression levels of IleRS correlate with the biochemical phenotype of m.4277T>C in skeletal muscle and fibroblasts
Together, our data strongly indicate a pathogenic role for the m.4277T>C mt-tRNAIle mutation. However, the variable penetrance between the patient and his mother suggests that additional factors, either detrimental or protective, are involved in the modulation of the clinical and biochemical phenotype. Previous studies have shown that over-expression of mitochondrial aminoacyl-tRNA synthetases (aaRSs) can rescue the pathological phenotype both in transmitochondrial cybrids and in yeast models (23,28–30). Thus, we speculated that the penetrance of the m.4277T>C mt-tRNAIle mutation in this small kindred could be modulated by the expression levels of the mitochondrial IleRS. According to our hypothesis, both the IleRS mRNA and protein were increased in fibroblasts and skeletal muscles, but not in transmitochondrial cybrids, from the clinically unaffected mother when compared with the patient (Fig. 6). A small although significant increase in the expression levels of the non-cognate leucyl-tRNA synthetase gene (LARS2) was observed in the same tissues, whereas the expression levels of the glutamyl-tRNA synthetase gene (EARS2) were similar to controls (Supplementary Material, Fig. S4).
Direct sequencing of the whole coding region ruled out the presence of pathogenic mutations and/or major rearrangements in the IleRS gene (IARS2) and showed the presence of an identical, previously reported synonymous coding polymorphism in the proband and his mother (variation ID rs116823205, www.ensemble.org; April 2011) (31). IARS2 expression levels in skeletal muscle and fibroblasts from patients with the m.4300A>G mt-tRNAIle mutation were similar to normal controls. However, skeletal muscle and fibroblasts from unaffected carriers bearing homoplasmic levels of the same mutation were not available for comparison.
Over-expression of exogenous IARS2 ameliorates viability of mutant transmitochondrial cybrids in galactose medium
We then over-expressed IARS2 in the m.4277T>C mutant transmitochondrial cybrids by transient transfection. As shown in Figure 7A, the IleRS-green fluorescent protein (GFP) fusion colocalized with Mitotracker Red, indicating mitochondrial localization. This result was paralleled by an approximate 5-fold increase in IARS2 expression and by an increase in the intracellular protein levels, as demonstrated by immunofluorescence (Fig. 7B and C). Transfected cells maintained in galactose medium showed a ∼50% increase in viability after 48 h and a ∼30% decrease in apoptotic rate after 24 h when compared with non-transfected ones (Fig. 7D and Supplementary Material, Fig. S6).
Homoplasmic mtDNA mutations show a remarkably variable penetrance, and little is known about the genetic and epigenetic factors underlying this phenomenon. We report a patient with HCM and sensorineural deafness whose skeletal muscle biopsy revealed numerous COX-deficient fibers and a combined defect of respiratory chain complexes I, III and IV. Sequencing of the entire mitochondrial genome identified a homoplasmic m.4277T>C mutation in mt-tRNAIle, which has been previously reported among a group of mt-tRNA variants associated with essential hypertension in a Chinese Han population (32). Our patient showed a marked decrease in the steady-state levels of the mutated mt-tRNAIle in skeletal muscle (∼70% reduction when compared with controls).
To analyze the pathogenic role of this mutation, we transferred the mutated patient mitochondria into a neutral nuclear background (143B.TK-) by generating transmitochondrial cybrids. Although we could not observe significant changes in basal conditions (cells grown in glucose medium), we were able to demonstrate the presence of a pathological phenotype by growing several cybrid clones in glucose-free medium supplemented with galactose. The enforced use of OXPHOS in galactose medium led to an increase in ROS production in mutant transmitochondrial cybrid cells, which was significantly higher than controls, and reduced cell viability due to apoptotic cell death. These findings were reproduced in transmitochondrial cybrids homoplasmic for the m.4300A>G MTTI gene mutation which is associated with cardiomyopathy (4), showing that homoplasmic mt-tRNA mutations with a cardiac-restricted clinical phenotype are able per se to induce mitochondrial dysfunction. In keeping with increased ROS levels, transmitochondrial cybrids showed an activation of mitochondrial biogenesis, in agreement with previous reports (7,8,21).
The pathogenic role of the m.4277T>C mutation may be ascribed to changes in the conformation of the mt-tRNAIle molecule, as suggested by the slower electrophoretic migration of the mutated mt-tRNA with respect to the wild-type molecule. According to our results, the aminoacylation of mt-tRNAIle is not impaired, at least in transmitochondrial cybrids; we obtained similar results for the m.4300A>G mutation, in agreement with previous data (33).
Analysis of representative three-dimensional structures of tRNA molecules determined by X-ray crystallography shows that T4277 of mt-tRNAIle is likely to be involved in a tertiary interaction with T4306. Indeed, these positions are structurally equivalent to positions 15 and 48, according to the standard tRNA numbering, and the interaction between nucleotides occurring at these sites is strongly conserved. This observation highlights the importance of exploiting structural as well as sequence information in the assessment of the pathogenicity of mt-tRNA mutations.
In the analyzed structures of tRNA molecules in complex with aaRSs, position 15 is not involved in interactions with the enzyme, in agreement with previous reports, showing that this position is not among those involved in aaRS recognition in most tRNAs (34). Together with our finding of normal aminoacylation of the m.4277T>C mutant mt-tRNAIle from transmitochondrial cybrids, these observations strongly suggest that T4277 is not an identity element for human mt-tRNAIle.
Structural changes can impair the processing and stability of mutated tRNAs (33), providing an explanation for the reduced steady-state level of mutated mt-tRNAIle observed in our patient. Low levels of a mature tRNA available for aminoacylation may in turn affect mitochondrial protein synthesis, thus resulting in a biochemical phenotype. Interestingly, the steady-state level of mutated m.4277T>C was lower in the affected skeletal muscle than in fibroblasts and transmitochondrial cybrids carrying the mutation, in agreement with the more severe phenotype of the former.
An unanswered question regarding homoplasmic mtDNA mutations is their variable clinical penetrance. Accordingly, it was interesting to note that skeletal muscle biopsy from the patient's unaffected mother showed normal respiratory chain enzyme activities and mt-tRNAIle steady-state levels, while being homoplasmic for the m.4277T>C mutation. In addition, there was no evidence of a compensatory increase in mitochondrial biogenesis. Her cultured skin fibroblasts incubated in galactose medium showed growth capability similar to controls, at variance with the patient's fibroblasts, which showed a severe growth impairment and apoptotic cell death. Altogether, these findings point to the existence of a nuclear factor capable of modulating or preventing the phenotypic expression of this homoplasmic mt-tRNAIle mutation.
Previous work by our group has shown that the over-expression of cognate aaRSs can rescue the defects of yeast mt-tRNA mutants (35,36). Different groups confirmed these findings on human cell lines (23,28,30). Based on these results, we speculated that the expression level of mitochondrial IleRS could modulate the penetrance of m.4277T>C mt-tRNAIle mutation in this family. Concordant with this hypothesis, we observed significantly higher expression of the cognate IleRS in skeletal muscle and fibroblasts from the unaffected mother when compared with the proband. This observation was confirmed by results showing an improvement of the phenotype of mutant transmitochondrial cybrids following over-expression of exogenous IARS2.
This is the first observation that constitutively high levels of aaRSs in human tissues are able to suppress the phenotypic expression of a homoplasmic mt-tRNA point mutation.
Since isoleucylation does not seem to be affected by the m.4277T>C transition, the suppressive effect of IleRS may be mediated by a stabilizing, chaperone-like activity on the structurally altered tRNA (29,37). In agreement with this hypothesis, increased expression of IleRS was paralleled by higher steady-state levels of mutated mt-tRNAIle, in skeletal muscle and, to a lesser extent, in fibroblasts from the proband's unaffected mother. However, further investigations are required to clarify the mechanisms of the suppressive effect of IleRS.
In addition, our observation of increased expression levels of the non-cognate LARS2 is in agreement with a recent report, showing the cross-suppressing effects of non-cognate aaRS on mitochondrial defects in yeasts (29), and deserves a much deeper investigation.
We could not extend our speculation to the homoplasmic m.4300A>G mt-tRNAIle mutation, which has also been associated with HCM. In fact, no unaffected mutation carriers were available for comparison with affected patients.
Since no disease-associated mutations/major rearrangements were identified in the coding region of the IARS2 gene in our patient and his mother, we may speculate that a regulatory mutation and/or epigenetic factors are the most likely explanation for the observed differences in expression levels.
The recent observation that mutations in mitochondrial alanyl-tRNA synthetase cause severe infantile mitochondrial cardiomyopathy confirms a causative role for aaRSs in mitochondrial diseases (38). We have extended this observation by showing that constitutively high expression levels of non-mutated aaRSs can modulate the penetrance of homoplasmic mt-tRNA mutations. Based on the results of this work and previously published data, we believe that a more extensive analysis of the baseline levels of aaRSs in normal and affected tissues is required, in light of their potential therapeutic effect.
MATERIALS AND METHODS
All studies were undertaken with informed patient consent and conformed to local protocols, as established by the Ethics Committee of Sapienza University of Rome.
Biochemical analysis of skeletal muscle biopsy
Biochemical analysis of individual mitochondrial respiratory chain complexes was carried out on frozen skeletal muscle homogenates, as described previously (39). Specific enzyme activities were normalized to that of citrate synthase, a marker of mitochondrial mass.
Skin fibroblasts were obtained from the index case and his mother bearing the m.4277T>C mutation, from two patients with MIC bearing the homoplasmic m.4300A>G mutation in mt-tRNAIle that belong to a previously described pedigree (patients IV-9 and V-14, 4) and from three unrelated age-matched controls. Fibroblasts were cultured in Dulbecco's modified Eagle's medium (DMEM), supplemented with 4.5 g/l d-glucose, 10% fetal bovine serum (FBS), 2 mml-glutamine, 100 U/ml penicillin and 100 µg/ml streptomycin (referred to as glucose medium) in a humidified atmosphere of 95% air and 5% CO2 at 37°C. A subset of experiments was performed in parallel either in glucose medium or in glucose-free DMEM, supplemented with 10% dialyzed FBS (dFBS), 5 mm galactose and 110 mg/ml sodium pyruvate (referred to as galactose medium).
Transmitochondrial cybrids were generated by polyethylene glycol fusion of enucleated m.4277T>C, m.4300A>G and control fibroblasts with the mtDNA-less (ρ°) osteosarcoma (143B.TK-) cell line (a kind gift of Andrea Martinuzzi), as described previously (40). The cells were replated 24 h after fusion in uridine-free DMEM, supplemented with 100 µg/ml bromodeoxyuridine and 10% dFBS to allow selection, and individual cell clones (about 16–25 for each transmitochondrial cybrid cell line) were isolated 10–20 days later using glass cylinders. For each transmitochondrial cybrid cell line, we selected two to three clones for the present study. For each clone, the relative amount of mtDNA was determined by real-time quantitative PCR (Supplementary Material, Table S2). In addition, the m.4277T>C and m.4300A>G mutations were shown to be homoplasmic by RFLP analysis using the Agilent 2100 bioanalyzer instrument with the 1000 LabChip kit (discussed subsequently) (data not shown). In all cases, the complete sequence of the transmitochondrial cybrid-repopulating mtDNA was determined and the mtDNA haplogroup assigned (Supplementary Material, Tables S1–S3). Three additional control transmitochondrial cybrids used in this study (a kind gift of Valerio Carelli) have been previously characterized (41,42).
Mitochondrial DNA analysis
Total DNA was isolated using Wizard Genomic DNA Purification Kit (Promega, Madison, WI, USA). Possible mtDNA deletions were screened in muscle, as described previously (43). The total amount of mtDNA was calculated by quantitative real-time PCR (AB 7500 Fast, Applied Biosystems, Warrington, UK), using a previously described method (44) with slight modifications. Briefly, an mtDNA fragment (nt 4625–4714) and a nuclear DNA fragment (FasL gene) were amplified using a TaqMan-MGB probe system (PCR conditions, primers and probes are detailed in Supplementary Material, Table S4). The amount of mtDNA relative to nuclear genomic DNA was evaluated with the comparative Ct method (2−ΔCt), where ΔCt = Ct mtDNA−Ct nuclear DNA. The amplification efficiencies were determined by generating a standard curve from serial dilutions (10–10−6) of a vector (a kind gift of Andrea Cossarizza), in which the templates for the two amplifications were cloned tail to tail in a 1:1 ratio. Standard curves were also generated with genomic DNA, showing that the amplification efficiencies of the mitochondrial and nuclear DNA were similar.
The entire mitochondrial genome was sequenced in a series of overlapping fragments using M13-tagged oligodeoxynucleotide primers to facilitate direct sequencing of the PCR-amplified products with BigDye® terminator chemistries on an Applied Biosystem 3100 automated sequencer (Applied Biosystems) (45). All sequences were directly compared with the revised Cambridge reference sequence for human mtDNA (GenBank accession no. NC012920).
The levels of mutated mtDNA in skeletal muscle from the patient and his mother were determined by last hot cycle PCR/RFLP analysis. A 161 bp fragment encompassing the mutation site was amplified with the forward mismatch primer (positions 4251–4275) 5′-CCCTCAAACCTAAGAAATATGTGTG-3′ and the reverse M13-tagged mismatch primer (positions 4393–4374) 5′-caggaaacagctatgaccGGGTGTGATAGGTGTCACGG-3′ (mismatch bases are shown in bold). Samples were subjected to 35 cycles of amplification with an annealing temperature of 60°C; the final extension proceeded for 8 min. After the addition of 0.4 μm of each primer, 5 μCi (alpha-32P)-deoxycytidine triphosphate (3000 Ci/mmol) and 1 U AmpliTaq Gold® polymerase (Applied Biosystems), the PCR products were subjected to an additional cycle of amplification. Labeled amplicons were precipitated and equal amounts (1000 counts) digested with 8 U Tsp45I (New England Biolabs, Hitchin, UK). Restriction fragments were separated by 12% non-denaturing polyacrylamide gel electrophoresis (PAGE), dried onto a support and analyzed with ImageQuant software (Molecular Dynamics, Eugene, OR, USA), following PhosphorImager exposure (Molecular Dynamics). On digestion, a single Tsp45I site in the wild-type product generates fragments of 125 and 36 bp. The m.4277 T>C mutation introduces an additional Tsp45I recognition site that cleaves the 125 bp fragment into two smaller products of 103 and 22 bp.
To confirm homoplasmic levels of the m.4277T>C mutation in blood, fibroblasts and transmitochondrial cybrids, restriction fragments (143 bp) amplified using non-M13-tailed mismatch primers as detailed earlier and digested with 8 U Tsp45I were separated using the Agilent 2100 bioanalyzer instrument, which utilizes LabChip® technology, with the 1000 LabChip kit (Agilent Technology, Palo Alto, CA, USA).
Steady-state levels and aminoacylation status of mutated mt-tRNAs
To determine mt-tRNAIle steady-state levels, total cytosolic RNA from cultured cells and tissue homogenate was obtained using Trizol reagent (Life Technologies, Invitrogen, Carlsbad, CA, USA) and analyzed (1 µg) on a high-resolution northern blot (13% urea/polyacrylamide), as described previously (4).
To investigate the aminoacylation status of mt-tRNAIle, mitochondria were isolated from 28 × 106 transmitochondrial cybrid cells by standard differential centrifugation under acidic conditions (pH 4.7). RNA obtained from isolated mitochondria was analyzed on 6% polyacrylamide/8 m urea sequencing gels (40 cm length, 2 days run). These gels are partially denaturing when electrophoresis is conducted at 4°C and completely denaturing at room temperature. They allow analysis of the characteristics of mutated tRNAs, including size, amount, conformation and acylation (24). Complementary oligonucleotides specific for mt-tRNAIle (5′-GGTTTAAGCTCCTATTATTTA-3′) and for mt-tRNALeu(UUR) (5′-GTTTTATGCGATTACCGGGCTCTGC-3′) were 5′ end-labeled as described previously (46) and used as probes for northern blot experiments.
Reactive oxygen species
ROS measurement was carried out on living cells using 2′,7′-dichlorofluorescin-diacetate (H2-DCF-DA; Molecular Probes and Invitrogen) (47). Cells (5 × 103) were seeded into 96-well plates and allowed to adhere overnight. After 1 h incubation in galactose medium, cells were supplemented with 10 µm H2-DCF-DA and maintained for 30 min at 37°C. The probe accumulates in cells and is hydrolyzed by cytoplasmic esterases to 2′,7′-dichlorofluorescin, which then reacts with H2O2 to give the fluorescent 2′,7′-dichlorofluorescein (DCF). Cells were washed with phosphate-buffered saline (PBS), and DCF fluorescence intensity (excitation = 485 nm and emission = 535 nm) was measured using a Multilabel Plate Reader (VICTOR™ X3 PerkinElmer). Fold increases in ROS production were calculated using the following equation: (Fgal–Fblank)/(Fglu–Fblank), where F is the fluorescence reading.
Cell growth was measured by the Trypan blue dye exclusion assay. Multiple series of 30 mm/60 mm dishes were seeded with a constant number of cells (3 × 105) for 24–96 h. Cells were harvested with 0.25% trypsin and 0.2% EDTA, washed, suspended in PBS in the presence of Trypan blue solution (Sigma, St Louis, MO, USA) at 1:1 ratio and counted using a hemocytometer. Experiments were performed both in glucose and in galactose medium. For each time point, the number of viable cells in galactose medium was expressed as a percentage of the number of cells in glucose.
Annexin V/propidium iodide staining for apoptotic cells
Cells were seeded at 1.5 × 105 cells/well, incubated overnight and then switched to glucose or galactose medium for 24/96 h. Cells were harvested by quick trypsinization to minimize potentially high annexin V background levels in adherent cells, washed and stained with Alexa 488/annexin V/propidium iodide (PI) (Molecular Probes and Invitrogen) and analyzed on an Epics XL-flow cytometer as described (21). Transfected cells were stained with APC-conjugated annexin V/7-amino actinomycin (BD Biosciences, Franklin Lakes, NJ, USA) and analyzed on a FACS-Calibur flow cytometer (BD Biosciences).
Oxygen consumption was measured in intact cells (5 × 106) using a Clark-type oxygen electrode (Hansatech, Norfolk, UK) in 1.85 ml DMEM lacking glucose and supplemented with 10% dFBS at 37°C, as described previously (48).
Gene expression analysis by quantitative real-time PCR
Gene expression experiments were performed on fibroblasts and skeletal muscle from patients bearing the m.4277T>C (n = 2) and the m.4300A>G (n = 2) mutations. As controls, four fibroblast primary cell lines from age-matched controls and nine skeletal muscle biopsies, obtained for diagnostic purposes from patients in whom metabolic disorders were excluded based on clinical investigation and histological analysis of biopsy specimens, were used. Total RNA was isolated using SV Total RNA isolation kit (Promega) and measured with a NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies, Wilmington, DE, USA). Total RNA (0.1–1 µg) was reverse-transcribed to cDNA using random hexamer primers. The relative expression levels of PPARGC1A and mitochondrial isoleucyl-, leucyl- and glutamyl-tRNA synthetases (IARS2, LARS2 and EARS2) genes were evaluated. We used TaqMan probe chemistry by means of inventoried and custom FAM-labeled TaqMan MGB probes (Applied Biosystems; Supplementary Material, Table S5), according to the manufacturer's instructions. In all samples, the relative expression of each target gene was evaluated with respect to one control (reference sample) using the comparative threshold cycle (ΔCt) method. All values were normalized for both the housekeeper hypoxanthine phosphoribosyltransferase 1 (HPRT1) and the mitochondrial voltage-dependent anion channel 1 (VDAC1) gene as a marker of mitochondrial mass. Each experiment was performed in triplicate.
Immunostaining and western blot analysis
The following primary antibodies were used: rabbit polyclonal antibody anti-IleRS (HPA024212, ATLAS Antibodies AB, AlbaNova University Center, Stockholm, Sweden); mouse monoclonal antibody anti-mitochondria extract clone MTC (UCS Diagnostic, Morlupo, Italy); mouse monoclonal antibody anti-porin (Invitrogen, Paisley, UK) and mouse monoclonal anti-β-actin (Sigma).
For immunocytochemistry, cells grown on cover slips were fixed with 4% formaldehyde freshly prepared from paraformaldehyde in PBS (pH 7.4) with 0.1% Triton X-100. Primary antibodies were visualized using secondary FITC- and Cy3-conjugated antibodies (Jackson Laboratories, Bar Harbor, ME, USA).
For western blot analysis, cells were rinsed twice with ice-cold PBS, lysed in ice-cold RIPA buffer [50 mm Tris–HCl pH 8, 150 mm NaCl, 1% NP-40, 0.5% sodium deoxycholate, 1% sodium dodecyl sulfate (SDS), 1 mm phenylmethylsulfonyl fluoride, 10 mg/ml aprotinin, 10 mg/ml leupeptin and 10 mg/ml pepstatin] and centrifuged at 10 000g for 10 min at 4°C. Protein concentration was measured by bicinchoninic acid (Beyotime Biotechnology, Haimen, China). Equal amounts of protein (80 μg) were separated by 12% SDS–PAGE and transferred onto a nitrocellulose membrane (Millipore, Bedford, MO, USA). Primary antibodies were visualized using horseradish peroxidase-conjugated secondary antibodies (Dako, Glostrup, Denmark). Signals were detected by enhanced chemiluminescence (Amersham Biosciences, UK).
Analysis of IARS2 gene
Direct sequencing of the entire coding region and intron–exon boundaries of the IARS2 gene (GenBank accession no. AK022665) was performed on an ABI Prism 3100 Genetic analyzer following standard procedures. Primer sequences and PCR conditions are reported in Supplementary Material, Table S6. The possible presence of major rearrangements was evaluated by quantitative real-time PCR with a specific inventoried TaqMan Copy Number Assay (Applied Biosystems; Supplementary Material, Table S4), following the manufacturer's instructions. As endogenous control, the nuclear FasL gene was amplified.
A pTune inducible vector containing the full-length cDNA fragment of IARS2 with a C-terminal fusion tGFP (True ORF.IARS2-GFP, 1 µg, Origene Technologies, Inc., Rockville, MD, USA) was used to transiently transfect transmitochondrial cybrids. The empty vector was used as control for mock transfection. Transfection was performed using Lipofectamine XLS (Invitrogen), according to the manufacturer's protocol. The transfected cells (pIARS2 cells) were maintained in DMEM-glucose for 6 h, and then 150 μm isopropyl-β-d-1-thiogalactopyranoside (IPTG, Sigma) was added to the medium for plasmid induction. After 24 h in IPTG-supplemented medium, subcellular localization of the fusion protein was evaluated by fluorescence microscopy with 0.3 μm Mitotracker Red FM (Molecular Probes and Invitrogen). Immunofluorescent staining of transfected cells with a specific antibody against IleRS and quantitative real-time PCR for IARS2 expression were performed in parallel. To test growth capability and rate of apoptosis, transfected cells were maintained in DMEM-glucose with 150 μm IPTG for 24 h and then harvested and plated (30 × 104) in either glucose or galactose medium plus 150 μm IPTG.
All data are expressed as mean ± SEM. Standard analysis of variance procedures followed by multiple pairwise comparisons adjusted with the Bonferroni correction were performed. Results were considered to be significant at P< 0.05.
Analysis of three-dimensional structures and gene sequences of tRNA molecules
To investigate the potential effect of the m.4277T>C mutation on the three-dimensional structure of human mt-tRNAIle, we analyzed several representative three-dimensional structures of tRNA molecules determined by X-ray crystallography and available from the PDB (http://www.rcsb.org/) (25). The analyzed structures comprise: (i) all available tRNA molecules in the free state, all of which have 8 nts in the D-loop, namely yeast tRNAPhe (PDB ID: 1EHZ) and tRNAAsp (PDB ID: 2TRA, 3TRA), HIV-1 tRNALys3 (PDB ID: 1FIR) and Escherichiacoli tRNAPhe (PDB ID: 3L0U) and (ii) tRNA molecules with D-loops of different lengths with respect to those in the previous set, namely E. coli tRNACys in complex with elongation factor EF-Tu (PDB ID: 1B23) and tRNAGln2 in complex with glutaminyl-tRNA synthetase (PDB ID: 1QTQ), both of which have 7 nts in the D-loop, and Staphylococcusaureus tRNAIle in complex with IleRS (PDB ID: 1FFY), which has 9 nts in the D-loop. Structure visualization and analyses were performed using the software InsightII (Accelrys Inc.) and PyMOL (Schrödinger LLC).
In addition, we analyzed the sequences of tRNA genes available from two databases: (i) ‘Compilation of tRNA sequences and sequences of tRNA genes’ (http://www.staff.uni-bayreuth.de/~btc914/search/index.html) (26), comprising approximately 3700 tRNA sequences from organisms belonging to different kingdoms (e.g. viruses, archaea, eubacteria, single cells and fungi, plants and animals) and cellular compartments (e.g. mitochondria, cytoplasm and chloroplasts) and (ii) ‘Compilation of mammalian mitochondrial tRNA genes’ (http://mamit-trna.u-strasbg.fr) (27), comprising nearly 3000 tRNA sequences from mammalian mitochondria, a large fraction of which is not included in the previous database.
Conflict of Interest statement. None declared.
This work has been supported by Associazione Serena Talarico per i Giovani nel Mondo and Fondazione Giuseppe Tomasello Onlus, Sapienza University of Rome Grant (C26A1089SW), a Wellcome Trust Programme Grant (074454/Z/04/Z) and the UK NHS Specialised Services ‘Rare Mitochondrial Disorders of Adults and Children’ Diagnostic Service (http://www.mitochondrialncg.nhs.uk).