The Diabetes-associated 3243 Mutation in the Mitochondrial tRNALeu(UUR) Gene Causes Severe Mitochondrial Dysfunction without a Strong Decrease in Protein Synthesis Rate*

Cells harboring patient-derived mitochondria with an A-to-G transition at nucleotide position 3243 of their mitochondrial DNA display severe loss of respiration when compared with cells containing the wild-type adenine but otherwise identical mitochondrial DNA sequence. The amount and degree of leucylation of tRNALeu(UUR) were both found to be highly reduced in mutant cells. Despite the low level of leucyl-tRNALeu(UUR), the rate of mitochondrial translation was not seriously affected by this mutation. Therefore, decrease of mitochondrial protein synthesis as such does not appear to be a necessary prerequisite for loss of respiration. Rather, the mitochondrially encoded proteins seem subject to elevated degradation, leading to a severe reduction in their steady state levels. Our results favor a scheme in which the 3243 mutation causes loss of respiration through accelerated protein degradation, leading to a disequilibrium between the levels of mitochondrial and nuclear encoded respiratory chain subunits and thereby a reduction of functional respiratory chain complexes. The possible mechanisms underlying the pathogenesis of mitochondrial diabetes is discussed.

Nucleotide substitutions in the mitochondrial tRNA genes are important factors in the pathogenesis of several multisystem disorders including mitochondrial encephalomyopathies and maternally inherited diabetes and deafness (MIDD) 1 (1)(2)(3)(4)(5)(6). A point mutation in the structural gene for a tRNA may be expected to influence protein synthesis either quantitatively or qualitatively. A strong reduction in mitochondrial protein synthesis, in particular in the proteins with a high lysine content, has been reported in the case of the 8344 mutation in the tRNA Lys gene of myoclonus epilepsy with ragged red fibers (MERRF). This reduction is thought to be caused by the premature termination of mitochondrial translation on lysine codons and the subsequent formation of truncated polypeptides (7). Mitochondrial protein synthesis has also been found de-creased in the case of the 3243 mutation associated with mitochondrial myopathy, encephalopathy, lactic acidosis, and stroke-like episodes (MELAS) (8,9). Therefore, the general view has emerged that mutations in tRNA genes cause mitochondrial dysfunction by lowering mitochondrial protein synthesis and, as a consequence, the amount of mitochondrially encoded respiratory chain enzymes. Indeed, mitochondrial respiration is severely reduced in these cases (7)(8)(9). However, the large variation in clinical phenotypes that associate with tRNA mutations is difficult to explain solely by the lowered respiration rate.
MIDD accounts for 1-2% of the diabetic population in Europe and Japan and is clinically characterized by an early middle age onset of diabetes and sensorineural hearing loss, progressive insulin secretory defect, absence of islet cell antibodies, and absence of obesity (10,11). As a first step in understanding the molecular mechanisms caused by the 3243 mutation, we have established mitochondrial transformants (cybrids) carrying either the wild-type adenine or the mutant-type guanine at position 3243 but having otherwise identical mitochondrial DNA sequences (12). Care was taken to ensure that all cell lines contained high and comparable mitochondrial DNA copy numbers. The mutant-type cells displayed a severe reduction of cellular respiration and respiratory chain activity, which has provided proof for the pathogenicity of the 3243 mutation in MIDD (12). In the present work, the analysis of these cybrids has been extended toward mitochondrial tRNA and protein metabolism in order to understand better the molecular mechanisms leading from the 3243 mutation to impaired respiration and MIDD.

EXPERIMENTAL PROCEDURES
Cell Culture-The mitochondrial transformants (cybrids) used in this study had been generated earlier by the transfer of mitochondria from two genetically unrelated diabetes patients (V and A) into mitochondrial DNA-less B°-3 cells (12). Patient V displays the phenotype of maternally inherited diabetes and deafness. Patient A suffers from Alport-like syndrome, which is characterized by progressive kidney disease, sensorineural hearing loss, and maternal inheritance (13). Diabetes also develops after transplantation and immunosuppressive therapy with steroids but is overshadowed by the kidney disease. Therefore, MIDD and Alport-like syndrome have in common maternal inheritance, sensorineural deafness, and diabetes, with renal failure as a prominent characteristic for Alport-like syndrome. From both donors we obtained fully wild-type cybrids (W6 and W7 from patient V; W20 from patient A) as well as mutant-type cybrids (M48 and M50 from patient V; M12, M26, and M30 from patient A) (12). Cybrids, 143B osteosarcoma cells and B°-3 cells were all cultured in Dulbecco's modified Eagle's medium containing 4.5 mg/ml glucose and 110 g/ml pyruvate (DMEM) supplemented with 50 g/ml uridine and 10% fetal bovine serum (complete DMEM). For long-term culturing (maximum ϳ3 months), the cybrids were cultured in DMEM supplemented with 10% dialysed fetal calf serum to preserve a high level of mitochondrial DNA. The mitochondrial DNA copy numbers of the cybrids used were all comparable with that of 143B cells. Cybrid cells were never cultured for longer than about 3 months. After this period a fresh vial of cells from the stock, stored in liquid nitrogen, was used.
Oxygen Consumption-The rate of oxygen consumption by ϳ5 ϫ 10 6 exponentially growing cells in 2 ml of complete DMEM but without glucose was measured at 37°C using a Clark-type electrode as described (14).
Northern Blot Analysis-Total RNA (20 g), isolated from exponentially growing cells by the guanidinium thiocyanate method (15), was subjected to electrophoresis through a 1% agarose-formaldehyde gel, transferred onto Hybond N ϩ membranes and hybridized overnight at 42°C with 32 P-labeled probes specific for the tRNA Leu(UUR) gene (nucleotides 3245-3861) or the tRNA Phe gene (nucleotides 245-706), and washed with several changes of 1ϫ SSC, 0.5% SDS at 42°C (16). tRNA bands were visualized by autoradiography and quantified using Phos-phorImager detection and ImageQuant analysis.
The degree of in vivo aminoacylation of tRNAs was determined by high-resolution electrophoresis and subsequent Northern analysis as described (17). To prevent deacylation of tRNA during isolation, RNA was isolated under acidic conditions by the guanidinium thiocyanate method (15) but using 0.15 M sodium acetate, pH 4.0, instead of pH 7 buffers throughout. The final RNA preparation was dissolved in 10 mM sodium acetate, pH 4.0.
Mitochondrial Protein Synthesis and Degradation-[ 35 S]Methionine incorporation into mitochondrially encoded proteins was analyzed on SDS-polyacrylamide gels essentially as described (18). In brief, ϳ5 ϫ 10 5 exponentially growing cells in a 3.4-cm dish were washed twice with methionine-free DMEM and incubated for 30 min at 37°C with Promix (a 2:1 mixture of [ 35 S]methionine and [ 35 S]cysteine from Amersham Pharmacia Biotech; 10 Ci/l, 1,000 Ci/mmol) in 0.75 ml of methioninefree DMEM supplemented with 10% dialysed fetal bovine serum and emetine (100 g/ml), which was added 15 min before the addition of label. The cells were washed three times with phosphate-buffered saline and finally dissolved in 100 l of SDS sample buffer. Samples (40 g) were analyzed on 15% SDS-polyacrylamide gels. Mitochondrial translation products were assigned according to Chomyn et al. (18). Protein concentration was determined by the bicinchoninic acid assay (Pierce).
In the case of [ 3 H]leucine incorporation, [ 35 S]methionine was replaced by L- [4, H]leucine (10 Ci/0.75 ml on a 3.4-cm dish; 68 Ci/ mmol) in leucine-free complete DMEM. After the specified incubation period, the ϳ5 ϫ 10 5 cells were washed three times with phosphatebuffered saline and dissolved in 0.5 ml of 0.1 M NaOH. 100-l aliquots were immediately withdrawn in duplicate, added to 100 l of cold 20% trichloroacetic acid, and heated for 10 min at 90°C. The precipitate was collected on a glass fiber filter, washed, and counted as described (19). The significance of the differences between wild-type and mutant-type incorporation was estimated by linear regression analysis using SPSS 7.0 software (Chicago).
The protein degradation rate of individual mitochondrial proteins was estimated from pulse-chase experiments (20). For pulse values, ϳ5 ϫ 10 5 cells (pretreated with chloramphenicol) were incubated for 2 h at 37°C with Promix and the reversible cytoplasmic protein synthesis inhibitor cycloheximide (200 g/ml) instead of emetine and washed three times with phosphate-buffered saline. For chase values, the pulsed cells were washed with complete DMEM instead of phosphatebuffered saline and subsequently cultured for 48 h in complete DMEM with two daily changes of medium. The decay rate constant, , was calculated from ϭ Ϫ(2.3 log(n/n 0 ))/t, where t ϭ 2880 min and n 0 and n denote the amount of 35 S in an individual protein band following pulse and chase, respectively. Autoradiograms were digitalized and processed using a GDS8000 gel analysis system (Ultra Violet Products, Lambridge) for that purpose. Statistical analysis of the decay data was performed using Student's t test (SPSS version 7.0).
Western Blotting and Immunoprecipitation-Western blotting was performed essentially as described (21). In brief, total cellular protein was separated on a 15% SDS gel and transferred to a polyvinylidene difluoride membrane (Immobilon P, Millipore). Subsequently, the proteins were specifically visualized using monoclonal antibodies against cytochrome c oxidase subunits I, II, and IV (Molecular Probes) and enhanced chemiluminescence (Amersham Pharmacia Bitotech).
Immunoprecipitates (22) were prepared by a 16-h incubation at 0°C of mitochondria-enriched fractions (35 g) (20) with monoclonal antibody (2 g) and protein G-Sepharose (10 l) (Amersham Pharmacia Biotech) in 0.5 ml of 10 mM Tris-HCl, pH 7.5, 150 mM NaCl, and 1% Triton X-100. The beads were washed twice with the same buffer and twice with 5 mM Tris-HCl, pH 7.5, and were then suspended in SDS sample buffer. The composition of the eluted proteins was analyzed by SDS-polyacrylamide gel electrophoresis followed by autoradiography.

RESULTS
Oxygen Consumption-Stable mitochondrial transformants containing wild-type or mutant-type mitochondrial DNA have been generated and characterized earlier (12). Cybrids were cultured on selective media to obtain and preserve high and comparable mitochondrial DNA copy numbers. Mutant-type cybrids displayed an ϳ5-fold reduced rate of respiration (Fig. 1) and a 79 -98% reduction of individual respiratory chain enzymes (12).
Steady State Level of tRNA-Northern blot analysis revealed that the level of tRNA Leu(UUR) in the mutant-type cells was reduced to ϳ25% of the level found in wild-type cybrids or the parental 143B cells, whereas the amount of tRNA Phe remained constant within this set of cybrids ( Fig. 2A). The relative ratio of tRNA Leu(UUR) to tRNA Phe was estimated to be 0.19 Ϯ 0.05 for the mutant-type and 0.80 Ϯ 0.12 for the wild-type cybrids. Furthermore, there was no change observed in the level of mature ND1, 12 S, and 16 S rRNA transcripts. "RNA19" however, which represents an unspliced transcript encompassing 16 S rRNA, tRNA Leu(UUR) , and the ND1 mRNA, was increased in the mutant-type cells (not shown) comparable with the results previously described for the MELAS 3243 cybrids (9).
Aminoacylation of tRNA-To determine whether the 3243 mutation also affects the degree of aminoacylation of tRNA Leu(UUR) , we examined the ratio of mitochondrial leucyl-tRNA Leu(UUR) and non-acylated tRNA Leu(UUR) in wild-type versus mutant-type cybrids. RNA was isolated under acidic conditions to prevent deacylation and analyzed on an acidic urea gel followed by Northern blotting (17). Because aminoacylation results in a lowered mobility of most tRNAs, this method allows the resolution and quantitation of both forms. The in vivo degree of aminoacylation of tRNA Leu(UUR) was found to be reduced from ϳ75% leucylation in wild-type cells to ϳ40% in all of the mutant-type cybrids (Fig. 2B). In contrast, tRNA Lys showed a more constant level of ϳ80% aminoacylation in both types of cells. Therefore, when taking into account the reduced steady state level of tRNA Leu(UUR) , the amount of aminoacylated leucyl-tRNA Leu(UUR) available for mitochondrial protein synthesis is very low in mutant-type cells and is estimated to be ϳ15% of the wild-type level.
Mitochondrial Protein Synthesis-Because all mitochondrial mRNAs contain a number of UUR codons, it may be expected that a low level of leucyl-tRNA Leu(UUR) would reduce the rate of mitochondrial translation, at least when leucyl-tRNA Leu(UUR) is a limiting factor for mitochondrial protein synthesis. Therefore, we estimated the rate of mitochondrial protein synthesis by measuring the incorporation of [ 35 S]methionine into protein in the presence of the cytoplasmic protein synthesis inhibitor, emetine. Unexpectedly, we did not observe a severe decrease in the rate of [ 35 S]methionine incorporation into mitochondrially encoded proteins (Fig. 3, left panel). Rather, the overall rate of mitochondrial protein synthesis of most of the mutant-type cybrids (4 of 5) was very comparable with that of the wild-type cybrids. The MERRF cybrid, which was included as a control, did show a sharp decrease in mitochondrial translation (Fig. 3, right panel) (7,18). We did not find in our cybrids a correlation between labeling intensity of a particular protein and the number of UUR codons in its respective mRNA (cf. Ref. 7). Remarkably, the synthesis of ATPase subunits, especially ATPase 8, was strongly enhanced in mutant-type cybrids (Fig. 3, left  panel). Furthermore, mutant cells displayed four additional, but weakly 35 S-labeled protein bands (indicated by arrowheads in Fig. 3, left panel), which may represent degradation products of the mutant mitochondrial proteins. The cybrids originating from the MIDD patient were indistinguishable from that of the patient with Alport-like syndrome, with the exception of M12, which showed a ϳ50% reduction of protein synthesis rate.
Because the rate of translation does not seem to be seriously diminished in most cybrids, the 3243 mutation may possibly affect the accuracy of mitochondrial protein synthesis by inducing the incorporation of amino acids other than leucine at UUR codons. Therefore, we measured the rate of [ 3 H]leucine incor-poration into mitochondrial protein as a first estimate of the degree of misincorporation at UUR codons of the mitochondrial mRNAs. [ 3 H]Leucine incorporation was measured at various cell densities because we have noticed that the rate of mitochondrial protein synthesis depends on cell density, probably reflecting a gradual shutdown of protein synthesis as soon as the cells become more confluent. Because the mitochondrial genome contains two leucyl-tRNA genes with UUR as the minor codon for leucine (89 UUR codons versus 553 for CUN), the maximal difference between wild-type and mutant-type cybrids is expected to be in the order 15-20%, at least when complete misincorporation occurs at UUR codons. [ 3 H]Leucine incorporation was indeed slightly (ϳ15%), but significantly (p Ͻ 0.01) lower in mutant-type versus wild-type cells at all of the cell densities analyzed (Fig. 4). Therefore, the misincorporation of amino acids at UUR codons seems indicated.
Mitochondrial Protein Degradation-The rate of mitochondrial protein degradation was estimated from pulse-chase experiments (20). After pretreatment with chloramphenicol to stabilize the mitochondrial proteins after removal of the drug (20), the cybrids were labeled for 2 h with [ 35 S]methionine in the presence of the reversible cytoplasmic protein synthesis inhibitor cycloheximide. The decay of 35 S during the 48-h period following labeling was subsequently determined for individual protein bands. The stability of mitochondrially encoded proteins was found to be slightly but significantly (p Ͻ 0.01) lower in the mutant versus wild-type cybrids (Fig. 5). Furthermore, the decay rates of individual mitochondrial proteins are in a remarkably close range, suggesting that mitochondrial protein turnover be coordinately regulated.
Western blot analysis revealed a decreased steady state level of the mitochondrially encoded proteins cytochrome c oxidase subunits I and II, whereas that of the nuclear-encoded mitochondrial proteins cytochrome c oxidase subunit IV and elongation factor EF-Tu were both found to be comparable in mutant versus wild-type cybrids (Fig. 6). As may be expected, the mitochondrially encoded proteins were not present in the mitochondrial DNA-less B°-3 cells. Thus, the specific reduction of mitochondrially encoded proteins seems to reflect the increased rate of protein degradation.

FIG. 2. Highly reduced leucyl-tRNA Leu(UUR) levels in mutanttype cybrids.
A, steady state levels of tRNA Leu(UUR) and tRNA Phe in wild-type and mutant-type cybrids were determined by Northern blot analysis. Each of the two labeled cDNA probes for tRNA Leu(UUR) and tRNA Phe detected a single band corresponding to the size expected for tRNA Leu(UUR) (78 nucleotides) or tRNA Phe (74 nucleotides) in both the wild-type as well as the mutant-type cybrids. B, the in vivo aminoacylation levels of tRNA Leu(UUR) and tRNA Lys in wild-type and mutant-type cybrids were estimated by acidic high resolution Northern blot analysis using specific 32 P-labeled cDNA probes for tRNA Leu(UUR) and tRNA Lys , respectively. The position of the aminoacylated and non-acylated tRNAs are indicated on the right. In contrast to tRNA Leu(UUR) , the two tRNA Lys species were found to be well resolved, in agreement with Enriquez and Attardi (24). Incubation at 100°C was performed to enhance deacylation of aminoacyl-tRNA and to identify the relative positions of the aminoacylated and non-acylated tRNA species. . After a prolonged 270-min labeling, the additional 35 S-labeled protein bands as found previously in mutant-type MERRF cybrids (7) also became visible.

DISCUSSION
The total gene content of the mitochondrion, 2 rRNA, 22 tRNA, and 13 protein-coding genes, is completely involved in the synthesis of components for oxidative phosphorylation. Therefore, any mutation in mitochondrial DNA may be expected to affect mitochondrial respiration in the first instance. Indeed, in the case of most pathogenic mitochondrial DNA mutations, respiration is reduced in near-homoplasmic mutant cells, independent of the nature of mitochondrial DNA mutation. The clinical phenotype related to these mutations, however, varies strongly. For instance, MERRF, which is caused by a mutation in the tRNA Lys gene, is a severe neuromuscular disease with a low life expectancy, whereas MIDD with the tRNA Leu mutation presents as a milder disorder. Furthermore, the same mitochondrial DNA mutation can even produce distinct clinical phenotypes in different individuals (6). The latter is especially evident for the 3243 mutation in the tRNA Leu gene, which may lead to MIDD, Alport-like syndrome, MELAS, or chronic progressive external ophthalmoplegia (12). This difference in pathogenesis may best be explained by assuming the involvement of additional nuclear or mitochondrial factors that determine the spatial and temporal distribution of the 3243 mutation over the different tissues and thereby determine the organ most affected. Because tRNA also seems to be involved in processes other than translation (23), it cannot be ruled out that the mutated tRNA gene product exerts its effect instead via other processes.
We have generated distinct clonal cell lines carrying the 3243 mutation from patients with MIDD and Alport-like syndrome in a near-homoplasmic form. The mitochondrial genome from the patient with MIDD has been extensively sequenced with the only heteroplasmic nucleotide found at position 3243 (3). Therefore, it may be assumed that upon mitochondriamediated transformation the only difference between wild-type and mutant-type cybrids is the A-to-G transition at position 3243. Contrary to reports on cybrids from MERRF and MELAS patients (7,8,18,24), our mutant-type cybrids show a serious deficiency in cellular respiration without a strong decrease in the rate of protein synthesis. The same applies for our mutanttype cybrids derived from Alport-like syndrome and also for some of the 3243 cybrids isolated in Dr. H. Jacobs' laboratory (University of Tampere). 2 In this context it should be mentioned that another panel of cybrids with low copy numbers of mitochondrial DNA often displayed a serious decrease of mitochondrial protein synthesis (25). Therefore, we took great care to ensure that all cell lines contained a high and comparable amount of mitochondrial DNA (12). Because we have generated cybrids with severe mitochondrial dysfunction without a strong decrease in protein synthesis rate, it is evident that a decrease in mitochondrial protein synthesis is not a necessary prerequisite for mitochondrial dysfunction. Therefore, our findings call into doubt the belief that a decrease in mitochondrial protein synthesis is a necessary prerequisite for loss of respiration and induction of pathogenesis. Our data suggest rather that the 3243 mutation reduces the stability of the mitochondrially 2 H. Jacobs, personal communication.

FIG. 4. [ 3 H]Leucine incorporation into wild-type versus mutant-type cybrids. [ 3 H]Leucine incorporation into trichloroacetic acid
precipitable material appears to be highly dependent on cell density. Density is expressed as g of whole-cell protein/cm 2 of tissue culture area available and incorporation as cpm/g of whole-cell protein; 30 g/cm 2 corresponds to ϳ50% confluency. The data obtained for wildtype cybrids (q) and mutant-type cybrids (E), respectively, were fit to a straight line by using a least-squares fitting procedure. Linear regression analysis of these data revealed that mutant-type cybrids incorporated significantly (p Ͻ 0.01) less [ 3 H]Leu than wild-type cybrids.
FIG. 5. Enhanced degradation in mutant-type cybrids. The decay rate constant, , of individual mitochondrially encoded proteins was estimated from pulse-chase experiments as detailed under "Experimental Procedures." Because measurements are based on equal protein load, the degradation rate is somewhat underestimated in the slower growing mutant cybrids. Well resolved and strongly labeled protein bands were selected, except for cytochrome c oxidase subunit II (COII), which represents rather the sum of COII and ATPase subunit 6 (A6) degradation. Note that the stability of the selected proteins is remarkably comparable, with half-lives estimated to be ϳ50 h. Open bars, wild-type; filled bars, mutant-type cybrids. Results are expressed as means Ϯ S.D. When treated as one group, the mutant-type proteins were found to degrade significantly (p Ͻ 0.01) faster than wild-type proteins. When treated separately, only the differences of COII and A8 reached statistical significance (p Ͻ 0.05). A similar loss of stability of mutant-type cytochrome c oxidase subunit I was also observable after its immunoprecipitation from M50 versus W7 extracts, respectively. NDII, NADH dehydrogenase subunit II.
FIG. 6. Lowered levels of mitochondrially encoded proteins in mutant-type cybrids. Right, Western blots of mitochondrially encoded proteins cytochrome c oxidase subunit I (CO I) and II (CO II) versus the nuclear encoded protein CO IV and protein synthesis elongation factor EF-Tu. Comparable results were obtained with a polyclonal antiserum against CO II. Left, for comparison, the rate of synthesis of CO I and CO II, as reflected by immunoprecipitation from [ 35 S]methionine-labeled W7 and M50 cybrids, is shown. This result indicates that the two monoclonal antibodies recognize the wild-type and mutant-type CO I and CO II proteins equally well. encoded proteins, which may be due to misincorporated amino acids at UUR codons of the mitochondrial mRNAs (Fig. 4). Interestingly, a similar mechanism has been proposed to occur in SV40 transformed B-cells containing ϳ70% 3243 mutant mitochondrial DNA (26). Finally, it is important to mention that our cybrids did not contain a suppressor mutation in the anti-codon of the tRNA Leu(CUN) gene as found recently by Jacobs and co-workers (27).
The low level of tRNA Leu(UUR) itself may be explained by enhanced susceptibility for degradation as a result of decreased affinity toward its natural partners, leucyl-tRNA synthetase, EF-Tu, or the ribosome. In this context, it has been postulated that a tRNA is never free but is always bound to one of these three partners (28). On the other hand, because RNA19 is elevated, it cannot be ruled out that the synthesis of the mutated tRNA Leu(UUR) is impaired at the processing level (29). It has been suggested in MERRF disease that a relative level of ϳ40% Lys-tRNA Lys is sufficiently low to reduce the synthesis of mitochondrial proteins in MERRF cybrids (7). However, with a level of ϳ15% Leu-tRNA Leu(UUR) , we did not find such dramatic behavior, even though the absolute levels of tRNA Leu(UUR) and tRNA Lys are quite comparable in wild-type cells (30). But how should the low level of leucyl-tRNA Leu(UUR) be reconciled with the high rate of protein synthesis? With the assumption of an elongation rate of 20 amino acids/s/ribosome (31), which corresponds to an average transit time of 50 ms at each codon, the synthesis of a complete protein of 500 amino acids will require 25 s (500 ϫ 50ms). The ϳ6-fold decrease in leucyl-tRNA Leu(UUR) may diminish the rate of elongation at UUR at most by a factor of 6 but probably by less (32). Thus, the transit time at UUR will rise to 300 ms at the most. Theoretically, the time to synthesize the complete protein, containing on average 10 UUR codons, would be elevated to (490 ϫ 50 ms) ϩ (10 ϫ 300 ms) ϭ 27.5 s, which corresponds to a 10% reduction of the rate of protein synthesis at the most.
The 3243 mutation induces enhanced degradation of mitochondrial DNA-encoded proteins (Figs. 5 and 6). Although we have not directly shown that the primary sequence of these proteins is different at the UUR-encoded leucine residues in mutant-type cells, two independently obtained results both demonstrate the decreased stability of these proteins, which at least strongly suggests an alteration of their primary structure. The pulse-chase experiments directly revealed an elevated degradation rate (Fig. 5), whereas the same may also be inferred from the low steady state levels ( Fig. 6) in the absence of a strong decrease of de novo protein synthesis (Figs. 3, 4, and 6). Direct sequencing of isolated mitochondrial proteins will provide evidence as to whether alterations in the primary structure are indeed involved in the pathogenesis of MIDD.