ClinVar

The 3243A-G MTTL1 mutation is the most common heteroplasmic mtDNA mutation associated with disease. The percentage of mutated mtDNA decreases in blood as patients get older (Pyle et al., 2007).

PHENOTYPES

Goto et al. (1990) and Kobayashi et al. (1990) independently reported an mtDNA point mutation associated with the MELAS syndrome (540000). Kobayashi et al. (1991) showed that the A-to-G transition at nucleotide 3243 in the tRNA-leu (UUR) gene was indeed the cause of MELAS. (In UUR, R = A or G.) They isolated, from the same muscle tissue of a patient with MELAS, cell lines with distinctly different phenotypes: one was respiration-deficient and the other was apparently normal. The respiration-deficient cells were found to carry the abnormality in mtDNA. The mutation was found in 8 patients from unrelated families who had the mutation in heteroplasmic form, but was not found in control persons. Enter et al. (1991) likewise found this mutation in heteroplasmic form in Caucasian patients with the MELAS syndrome. This point mutation lies within a DNA segment responsible for transcription termination of the rRNA genes.

By histochemical, immunohistochemical, and single-fiber polymerase chain reaction (PCR) analysis, Moraes et al. (1992) demonstrated that ragged-red fibers in MELAS syndrome were associated with high levels of mutant mitochondrial genomes and with partial cytochrome c oxidase deficiency.

Ciafaloni et al. (1992) found the nucleotide 3243 mutation in 21 of 23 patients with MELAS and in all 11 oligosymptomatic and 12 of 14 asymptomatic relatives, but in only 5 of 50 patients with mitochondrial disease without features of MELAS. The proportion of mutant genomes in muscle ranged from 56 to 95% and was significantly higher in the patients with MELAS than in their oligosymptomatic or asymptomatic relatives. In those in whom both muscle and blood were studied, the percentage of mutations was significantly lower in blood and was not detected in 3 of 12 asymptomatic relatives.

In 2 patients with classic features of MELAS, Lertrit et al. (1992) found the A-to-G base substitution at nucleotide 3243 of tRNA(leu) in one and an 11084A-G mutation of ND4 in the other (516003.0001). The patient with the 3243A-G mutation was a 29-year-old woman with a 2-year history of classic migraine, recurrent stroke-like episodes with bilateral occipital infarction, recurrent occipital seizures, electroencephalogram consistent with encephalopathy, and high lactate and pyruvate levels in blood and cerebrospinal fluid. Muscle biopsy demonstrated extensive ragged-red fibers. Mosewich et al. (1993) studied a large family with the 3243 mutation as the cause of the MELAS syndrome. Family members had various combinations of sensorineural hearing loss, retinal pigmentary degeneration, migraine, hypothalamic hypogonadism, and mild myopathy. Only one member had a stroke-like episode at the age of 46 years. This patient had the highest percentage of mitochondrial chromosomes carrying the point mutation.

Matthews et al. (1994) described a woman without neurologic symptoms who died suddenly at the age of 42 of cardiomyopathy and lactic acidosis. Heteroplasmy for the 3243 mutation was found with a higher proportion of mutant mtDNA in heart (0.49), skeletal muscle (0.56), and liver (0.55) than in other tissues studied, for example, kidney (0.03). Vilarinho et al. (1997) reported a 6-year-old Portuguese boy with dilated cardiomyopathy, lactic acidosis, and no evidence of neurologic abnormality. Molecular genetic analysis showed the 3243 mutation in 88% of total muscle mtDNA and in 68% of the blood mitochondrial genome. Lower percentages were detected in blood from the mother (43%) and brother (49%). The brother had asymptomatic mild hyperlactic acidemia.

In a family of Indonesian descent, de Vries et al. (1994) found that clinical severity of MELAS varied directly in proportion to the quantity of mutated mitochondrial DNA in different tissues. The eldest of 5 children suffered from increasing sensorineural hearing loss after 15 years of age. The third-born sib had severe convulsions at the age of 18 years with stroke-like episodes and lactic acidosis. The fifth-born sib had unusually severe manifestations of MELAS with first manifestations at 7 years of age and death at the age of 14 years from severe cardiomyopathy. The clinically unaffected mother had 35% mutated mtDNA in her muscle but no mutated mtDNA in blood or fibroblasts.

As discussed in 520000, the maternally inherited diabetes-deafness syndrome (MIDD) without the manifestations of MELAS syndrome has been observed in association with an A-to-G transition at nucleotide 3243 in the MTTL1 gene (van den Ouweland et al., 1992). Reardon et al. (1992) and Schulz et al. (1993) described kindreds in which members with the maternally transmitted diabetes-deafness syndrome had an A-to-G transition at nucleotide 3243 of MTTL1.

Manouvrier et al. (1995) observed the 3243A-G mutation in the MTTL1 gene segregating with maternally inherited diabetes mellitus, sensorineural deafness, hypertrophic cardiomyopathy, or renal failure in a large pedigree with 35 affected members in 4 generations. Presenting symptoms almost consistently involved deafness and recurrent attacks of migraine-like headaches, but the clinical course of the disease varied within and across generations. Hypertrophic cardiomyopathy had not previously been a common finding in persons with the 3243A-G mutation and renal failure had not been reported.

Odawara et al. (1995) found the 3243 mutation in 3 of 300 patients with noninsulin-dependent diabetes mellitus or impaired glucose tolerance, and in none of 94 insulin-dependent diabetes mellitus or 115 nondiabetic controls. None of the patients with the mutation had significant sensorineural hearing loss.

Yang et al. (1995) described a 32-year-old Taiwanese woman in whom MELAS syndrome associated with diabetes mellitus and hyperthyroidism was caused by an A-to-G transition at nucleotide 3243 in the MTTL1 gene. In blood cells, approximately 60% of mtDNA was of the mutant type. In Taiwan, Chuang et al. (1995) found the A-to-G mutation at position 3243 in the MTTL1 gene in 1 of 23 pedigrees with multiple sibs affected with NIDDM. The pedigree was consistent with mitochondrial disease in terms of maternal transmission, relatively early onset, nonobesity, insulin-requirement, and association with hearing impairment. There was no correlation between the degree of heteroplasmy of the mitochondrial gene mutations in leukocyte DNA and clinical severity. Thus, the A-to-G transition at nucleotide 3243 is associated with 2 clinically distinct maternally transmitted syndromes: MELAS and noninsulin-dependent diabetes mellitus (NIDDM) with sensorineural hearing loss. In some populations the latter syndrome of maternally-inherited diabetes and deafness may represent 1 to 3% of all cases of NIDDM (Velho et al., 1996).

Yorifuji et al. (1996) reported this mutation in a mother with diabetes mellitus and sensorineural hearing loss and in her son. The clinical picture in the boy included short stature (as a result of deficient growth hormone secretion), moderate mental retardation, progressive nephropathy, and diabetes mellitus. Estimated percentages of mutated mtDNA were 11.8% in the child and 5.1% in the mother. In the renal biopsy specimen from the child, the percentage of mutated mtDNA was 65.6%. Yorifuji et al. (1996) proposed that nephropathy in the child was a result of this mutation. They suggested that mitochondrial disease should be taken into account when a patient has nephropathy of unknown cause.

Morten et al. (1995) sequenced part of the mtDNA control region in 11 unrelated patients heteroplasmic for the 3243 mutation associated with the MELAS phenotype. Only 2 patients shared the same sequence haplotype, implying that the 3243 mutation occurs independently in the maternal lineages of most MELAS patients.

Damian et al. (1996) reported a family in which a female infant with VACTERL (the association of vertebral, anal, cardiovascular, tracheoesophageal, renal, and limb defects; 192350) died at age 1 month due to renal failure. Her mother and sister later developed classic mitochondrial cytopathy associated with the A-to-G point mutation at nucleotide 3243 of mtDNA. Molecular analysis of mtDNA in preserved kidney tissue from the infant with VACTERL demonstrated 100% mutant mtDNA in multicystic and 32% mutant mtDNA in normal kidney tissue. Mild deficiency of complex I respiratory chain enzyme activity was found in the mother's muscle biopsy. Other maternal relatives were healthy but had low levels of mutant mtDNA in blood. Damian et al. (1996) stated that this was the first report to provide a precise molecular basis for a case of VACTERL. Stone and Biesecker (1997) were studying a cohort of 62 patients with VACTERL association. All of the children had normal chromosomes; only 1 had symptoms suggestive of mitochondriopathy, i.e., deafness and muscle weakness. All children had at least 3 of the 6 categories of anomalies associated with VACTERL. None of the affected children had levels of the 3243 mutation that were detectable by the methods used. Stone and Biesecker (1997) recognized the limitations of their study such as heteroplasmy of different tissues (only lymphocytes were studied). Stone and Biesecker (1997) suggested that the proposita reported by Damian et al. (1996) may have had oculoauriculovertebral dysplasia (164210) rather than VACTERL association.

Feigenbaum et al. (1996) also described a family that expanded the extreme clinical variability known to be associated with the A-to-G transition at nucleotide position 3243 of MTTL1. The propositus, a fraternal twin, presented at birth with clinical manifestations consistent with diabetic embryopathy, including anal atresia, caudal dysgenesis, and multicystic dysplastic kidneys. His cotwin, also male, was normal at birth, but at 3 months of age presented with intractable seizures later associated with developmental delay. The twins' mother developed diabetes mellitus type I at the age of 20 years and gastrointestinal problems at 22 years. Since age 19 years, the maternal aunt had had recurrent strokes, seizures, mental deterioration, and deafness, later diagnosed as MELAS syndrome due to the A-to-G mutation. A maternal uncle had diabetes mellitus type I, deafness, and normal intellect, and died at 35 years of age after recurrent strokes. This pedigree raised the possibility that, in some cases, diabetic embryopathy may be due to a mitochondrial cytopathy that affects both the mother's pancreas (and results in diabetes mellitus and the metabolic dysfunction associated with it) and the embryonic/fetal and placental tissues which make the embryo more vulnerable to this insult.

Velho et al. (1996) detected the 3243 mutation in 25 of 50 tested members of 5 white French pedigrees. Mutation-positive family members presented variable clinical features, ranging from normal glucose tolerance to insulin-requiring diabetes. They described the clinical phenotypes of affected members and detailed evaluations of insulin secretion and insulin sensitivity in 7 mutation-positive individuals who had a range of glucose tolerance from normal to impaired to NIDDM. All subjects, including those with normal glucose tolerance, demonstrated abnormal insulin secretion on at least 1 test. The data suggested to Velho et al. (1996) that a defect of glucose-regulated insulin secretion is an early and possible primary abnormality in carriers of the mutation. They speculated that this defect may result from the progressive reduction of oxidative phosphorylation and may implicate the glucose-sensing mechanism of beta cells.

Tamagawa et al. (1997) described the audiologic features in patients with hearing loss associated with the 3243A-G mutation. Four patients without and 5 patients with MELAS were studied. Most of the patients had bilateral progressive sensorineural hearing loss. The most common shape of the audiogram was sloping, while cases in the advanced stages had flat audiograms. Speech discrimination scores were generally poor and did not parallel the degree of hearing loss. The studies suggested that the lesion for hearing loss could include both cochlear and retrocochlear involvement, but did not demonstrate a significant difference in the audiologic findings between patients with and those without MELAS.

Lam et al. (1997) reported the case of a boy in whom MELAS appeared to be precipitated by valproate therapy. He had mild mental retardation and a right-sided convulsion at the age of 12 years. About 1 year later he experienced a similar seizure and was then treated with sodium valproate (200 mg 3 times daily). Eight days after initiation of valproate, he developed right hemiparesis and hypotonia and had 2 seizures. The serum levels of valproate were not excessive, but when no improvement occurred an idiosyncratic drug reaction was suspected and valproate was discontinued. Blood lactate and pyruvate were found to be elevated. Brain CT scan showed a left parietooccipital infarct and bilateral basal ganglia calcification. Muscle biopsy showed ragged-red fibers. Electron microscopy of muscle showed increased numbers of mitochondria and atypical mitochondria in the subsarcolemmal region, some with inclusion bodies. A 3243A-G mutation was detected in mitochondrial DNA from this patient. This mutation predisposed the patient to the detrimental effects of valproate on oxidative phosphorylation. The findings of Lam et al. (1997) supported the suggestion that valproate should not be given to patients suspected of having mitochondrial diseases. In addition, for patients whose seizures worsen with valproate therapy, an inborn error of mitochondrial metabolism should be suspected.

In a 21-year-old male patient with overlapping MELAS and Kearns-Sayre syndrome, Wilichowski et al. (1998) demonstrated the 3243A-G mutation in the MTTL1 gene. Progressive external ophthalmoplegia, pigmentary retinopathy, and right bundle branch block were present when he experienced the first stroke-like episode at 18 years of age. The 3243A-G mutation was found in 79% of mitochondrial DNA and was present at low levels in fibroblasts (49%) and blood cells (37%). Biochemical analysis showed diminished activities of pyruvate dehydrogenase (23%) and respiratory chain complexes I and IV (57%) in muscle, but normal activities in fibroblasts. Immunochemical studies showed normal content of E1-alpha, E1-beta, and E2 proteins. No nuclear mutation of the E1-alpha gene (PDHA1; 300502) was found. These observations suggested that mitochondrial DNA defects may be associated with altered nuclear encoded enzymes that are actively imported into mitochondria and constitute components of the mitochondrial matrix.

Dashe and Boyer (1998) discussed the case of a 13-year-old girl with a relapsing-remitting neurologic disorder that proved to be MELAS. The 3243A-G mitochondrial mutation was identified. The patient had had a normal childhood, although she was awkward at running, bicycle riding, and gymnastics. Twenty-six months before admission, she had an acute febrile episode with otitis, pharyngitis, headache, drowsiness, imbalance, and confusion, and abnormalities of the cerebral cortex were found by MRI and computed tomography (CT). The following month the girl had a grand mal seizure, followed by cortical blindness for 18 hours, with slow resolution. Three sibs of the mother had died in childhood. One, an older brother, never walked and died at 2 years of age of 'Schilder's disease;' another brother died in infancy of a 'high fever;' and a sister died of 'malnutrition' at 4 years of age. Both the proband's mother and a maternal uncle had hearing loss. Disorders causing strokes or stroke-like episodes in children were reviewed, including disorders of blood vessels or blood and metabolic disorders. In this case, analysis of mitochondrial DNA in skeletal muscle and cerebellum showed that 88% of the mitochondrial DNA was mutant. Over a period of 7 or 8 years the patient continued to have seizures that were difficult to control, accompanied by decline in cognition, hearing, vision, and balance. By the age of 20 years, she could no longer walk. Endocrinopathies, including hypothyroidism, inappropriate release of antidiuretic hormone, and diabetes mellitus developed in her late teens. She died with sepsis and multiorgan failure at the age of 23 years.

Chinnery et al. (1998) demonstrated that for both the MELAS 3243A-G mutation and the MERRF 8344A-G mutation (590060.0001) higher levels of mutant mtDNA in the mothers' blood were associated with an increased frequency of affected offspring. They also found that at any one level of maternal mutation load there was a greater frequency of affected offspring for the MELAS 3243A-G mutation than for the MERRF 8344A-G mutation.

Sue et al. (1999) reported 3 unrelated children with the 3243A-G mutation who presented with severe psychomotor delay in early infancy. One patient's clinical picture was more consistent with Leigh syndrome, with apneic episodes, ataxia, and bilateral striatal lesions on brain MRI. A second patient had generalized seizures refractory to treatment and bilateral occipital lesions on brain MRI. The third child had atypical retinal pigmentary changes, seizures, areflexia, and cerebral atrophy on brain MRI. All patients had several atypical features in addition to early onset: absence of an acute or focal neurologic deficit, variable serum and cerebrospinal fluid lactate levels, and lack of ragged-red fibers in muscle biopsy specimens. The proportion of mutant mtDNA in available tissues was relatively low (range, 5 to 51% in muscle and 4 to 39% in blood). The observations extended the phenotypic expression of the 3243A-G 'MELAS' mutation, and confirmed previous observations that there is 'poor correlation between abundance of mutant mtDNA in peripheral tissues and neurologic phenotype.'

Deschauer et al. (1999) detected the 3243A-G mutation heteroplasmatically in DNA from muscle and blood of a 61-year-old patient who at the age of 54 developed a myopathy with painful muscle stiffness as the predominant symptom. Additionally, a hearing impairment requiring a hearing aid for the left ear, numbness of the left arm and leg, and impaired glucose tolerance were present. Nocturnal sleep was severely disturbed by the pain. Neurologic examination showed generalized induration of muscles of the trunk and the upper and lower limbs at rest. Palpation of muscles was painful. On passive motion, there was a contracture-like resistance. A stiff-legged gait was observed, caused by pain and contracture. Muscle histopathology showed a few ragged-red fibers.

Smith et al. (1999) studied 13 subjects with the 3243A-G mutation from 7 different pedigrees with maternally inherited diabetes and deafness to evaluate the association of retinal disease. Visual symptoms, particularly loss of visual acuity, appeared to be infrequent. Test findings suggested that the retinal dystrophy involved defective functioning of retinal pigment epithelial cells and of both rod and cone photoreceptors. The pigmentary retinopathy did not prevent diabetic retinopathy; one subject had evidence of both disorders. The authors stated that the 3243A-G mutation accounts for 0.5 to 2.8% of diabetes.

Latkany et al. (1999) reported the ocular findings in 4 family members with MELAS syndrome caused by the 3243A-G transition in the MTTL1 gene. Findings included ophthalmoplegia, neurosensory deafness, reduced photopic and scotopic electroretinogram (ERG) b-wave amplitudes, myopathy, and slowly progressive geographic macular retinal pigment epithelium atrophy.

Aggarwal et al. (2001) found the same mutation in a 29-year-old woman with gestational diabetes, deafness, Wolff-Parkinson-White (WPW) syndrome (194200), placenta accreta, and premature graying. Premature graying of the hair had started at age 15 years. Placenta accreta is a rare disorder, occurring especially in primigravidae. The features are 'retained placenta' and severe postpartum hemorrhage. The authors claimed this to be the first report of an association of the 3243A-G mtDNA mutation with WPW syndrome; a study of 27 other patients with WPW syndrome failed to reveal this mutation. The patient's sister also suffered from WPW syndrome, premature graying, and sensorineural deafness.

De Kremer et al. (2001) described a child with a 3243A-G mutation of mtDNA in all tissues. The patient had severe failure to thrive, severely delayed gross motor milestones, marked muscle weakness, and dilated cardiomyopathy. He also developed neutropenia at age 4 years. Laboratory studies showed persistently elevated urinary levels of 3-methylglutaconic and 2-ethylhydracrylic acids and low blood levels of cholesterol. The child died at age 4.5 years. De Kremer et al. (2001) suggested that Barth-like syndrome should be added to the list of phenotypes observed with the MELAS mutation.

In Finland, Uimonen et al. (2001) estimated the rate of progression of deafness in patients with the 3243A-G mutation. They examined 14 men and 24 women. The impairment of hearing was worse in men than in women, and women outnumbered men among patients with normal hearing or mild hearing impairment. The rate of progression was calculated to be 2.9 dB per year in men and 1.5 dB per year in women. A high degree of mutant heteroplasmy, male gender, and age were found to increase the severity of hearing impairment.

Deschauer et al. (2001) found the 3243A-G mutation in 16 patients with mitochondrial encephalomyopathies (10 index patients and 6 symptomatic relatives). Only 6 of these patients presented with stroke-like episodes and met the classic criteria of MELAS syndrome. One had MELAS/MERRF overlap syndrome. Two patients presented with stroke-like episodes but did not meet the classic criteria of MELAS. Of the 8 other patients, 1 had myopathy with hearing loss and diabetes mellitus, 1 had chronic progressive external ophthalmoplegia, 1 had diabetes mellitus with hearing loss, 1 had painful muscle stiffness with hearing loss, 1 had cardiomyopathy, 1 had diabetes mellitus, and 2 had hearing loss as predominant features. In 11 of the 16 patients, hearing impairment was obvious on clinical examination. Furthermore, all 5 patients with normal hearing on clinical examination showed subclinical hearing loss.

Chinnery et al. (2001) studied 9 patients from 4 families with the 3243A-G mutation; only 1 of the patients demonstrated severe neurologic disease with a proximal myopathy. Detailed study of resting and active muscle phosphate, creatine, and ATP synthesis failed to show any relationship between mutation load (percentage of mutated mtDNA in the muscle) and mitochondrial function in vivo. The authors suggested that nuclear genetic factors may play a modifying role in mitochondrial dysfunction.

Petruzzella et al. (2004) described cerebellar ataxia as an atypical manifestation of the 3243A-G MELAS mutation. The patient was a 55-year-old man who denied a family history of neurologic diseases and presented a 10-year history of progressive speech and balance disturbances. IQ was not affected. There was mild proximal muscle weakness and hypotrophy in the upper limb girdle muscles. The concurrence of muscle weakness and atrophy raised the suspicion of mitochondrial involvement, prompting testing. The 3243A-G mutation was found in relatively low abundance.

Jones et al. (2004) tested 570 patients with early or late age-related maculopathy (see 603075) for the 3243A-G mutation in mtDNA. Only 1 study participant with early ARM, hypertension, ischemic heart disease, and asthma was found to carry the 3243A-G mutation. Jones et al. (2004) concluded that the 3243A-G mutation is a very rare cause of typical ARM in the general population.

Salpietro et al. (2003) identified the 3243A-G mutation in 4 affected members of an Italian family with cyclic vomiting syndrome (500007). The youngest affected member, a 5-year-old boy, had 70% mutant mtDNA in peripheral blood. The boy's mother, maternal aunt, and maternal grandmother, who were all affected, had 35%, 30%, and 25% mutant mtDNA, respectively. There was a positive correlation between amount of mutant mtDNA and clinical severity. The 3 adults were affected by the syndrome during childhood and developed migraine headaches as adults.

Bohm et al. (2006) identified the 3243A-G mutation in 6 unrelated patients with mitochondrial complex IV deficiency (220110).

Donovan and Severin (2006) reported a kindred in which 4 of 7 sibs had adult-onset diabetes mellitus and sensorineural hearing loss with a confirmed mutation at position 3243 in the tRNA. Two other sibs in this kindred demonstrated different phenotypes of mitochondrial disease. After 1 year of treatment with coenzyme Q10, repeat stress thallium testing demonstrated improvement in the exercise tolerance of the proband from 7 to 12 minutes. Audiometry testing did not demonstrate a change in the rate of hearing decline.

In a study of 51 Danish individuals with the 3243A-G mutation, Jeppesen et al. (2006) found that skeletal muscle mutation load correlated inversely with maximal oxygen update and maximal workload during cycling exercise. Resting venous lactate directly corresponded to muscle mutation load. Those with COX-negative and/or ragged red fibers had over 50% mutation load in muscle, and all those with hearing impairment and diabetes mellitus or hearing impairment alone had more than 65% mutation load. In contrast, mutation load in blood was not correlated, or only weakly correlated, with these parameters. Jeppesen et al. (2006) concluded that the threshold at which oxidative impairment and muscle symptoms occur is as low as 50 to 65% in patients with the 3243A-G mutation, a level that is lower than previously reported based on cell culture studies.

Janssen et al. (2008) defined the 'mitochondrial energy-generating system' (MEGS) capacity as a measurement encompassing mitochondrial enzymatic reactions from oxidation of pyruvate to the export of ATP, which can be used as an indicator for overall mitochondrial function. In an analysis of muscle tissue from 24 MELAS patients with the 3243A-G mutation, MEGS capacity correlated better with mutation load than did analysis of individual respiratory chain enzyme activities, including complex I, III, and IV. The sensitivity and specificity of measurement using MEGS reached 78% and 100%, respectively, for a mitochondriopathy, which was significantly more accurate than measuring individual enzyme activities alone.

In a review of 45 probands with the 3243A-G mutation, Kaufmann et al. (2009) found variable involvement of multiple organ systems. In addition to classic MELAS features, patients had psychiatric problems, hearing loss, diabetes, exercise intolerance, gastrointestinal disorders, short stature, and learning disabilities.

In a patient with MELAS due to the 3243A-G mutation, Costello and Sims (2009) reported safe and effective treatment of symptomatic myoclonus with lamotrigine.

Nakamura et al. (2010) reported a family in which 4 members carried both the 3243A-G mutation and a 8356T-C transition in the MTTK gene (590060.0002), which is usually associated with MERFF syndrome. The female proband and her cousin had MERFF, a deceased aunt had a MERFF/MELAS overlap syndrome, and the mother of the proband was asymptomatic. Genetic analysis showed that the double mutations were heteroplasmic in blood of the proband and her cousin but at low levels in her asymptomatic mother. In muscle tissue of the proband and her aunt, the proportion of the 3243A-G mutation was higher than in blood, and the 8356T-C mutation was homoplasmic. Nakamura et al. (2010) hypothesized that the phenotype in affected individuals began with MERFF and evolved into MELAS later in life.

In a questionnaire-based survey, Parsons et al. (2010) found that 28 (80%) of 35 patients with MELAS due to the 3243A-G mutation and 33 (62%) of 53 carrier relatives reported autonomic symptoms compared to 2 (12%) of 16 controls. Gastrointestinal symptoms, orthostatic dizziness, and cold or discolored hands and feet were the most common complaints among mutation carriers.

In a retrospective study, Malfatti et al. (2013) found that 38 of 41 individuals with the MTTL1 c.3243A-G mutation had symptoms consistent with MELAS, whereas 3 were asymptomatic. Cardiac investigations identified left ventricular hypertrophy and/or left ventricular dysfunction in 18 patients, along with Wolff-Parkinson-White syndrome in 7, conduction system disease in 4, and atrial fibrillation in 1. Over a median 5-year follow-up period, 11 patients died, including 3 due to heart failure. Malfatti et al. (2013) concluded that, after central neurologic disease, cardiac disease has the greatest impact on prognosis in patients with the c.3243A-G mutation. Left ventricular hypertrophy was the only independent prognostic risk factor for adverse cardiac events, suggesting that these patients should be closely monitored. The severity of cardiac disease and adverse events did not correlate with mutation load in blood or urine or with ragged-red fibers on muscle biopsy.

STUDIES OF THE 3243A-G MUTATION

Yoneda et al. (1992) studied the growth characteristics of cells carrying the MELAS mitochondrial mutation by introducing patient mitochondria into human mtDNA-less cells. Five of 13 clonal cell lines containing mixtures of wildtype and mutant mtDNAs were found to undergo a rapid shift of their genotype toward the pure mutant type. On the other hand, the other 8 cell lines, which included 6 exhibiting nearly homoplasmic mutant mtDNA, maintained a stable genotype. Subcloning experiments and growth rate measurements clearly indicated that an intracellular replicative advantage of mutant mtDNA was mainly responsible for the dramatic shift toward the mutant genotype observed in the unstable cell lines.

Matthews et al. (1995) showed that fibroblast clones from subjects heteroplasmic for the MELAS 3243A-G mutation show wide variability in the degree of heteroplasmy. The distribution of mutant mtDNA between different cells was not random about the mean, suggesting that selection against cells with high proportions of mutant mtDNA had occurred. To explore the way in which heteroplasmic mtDNA segregates in mitosis, they followed the distribution of heteroplasmy between clones over approximately 15 generations. There was either no change or a decrease in the variance of intercellular heteroplasmy for the MELAS 3243A-G mutation, which was considered most consistent with segregation of heteroplasmic units of multiple mtDNA molecules in mitosis. After mitochondria from one of the MELAS 3243A-G fibroblast cultures were transferred to a mitochondrial DNA-free cell line derived from osteosarcoma cells by cytoplast fusion, the mean level and intercellular distribution of heteroplasmy was unchanged. Matthews et al. (1995) interpreted the findings as evidence that somatic segregation (rather than nuclear background or cell differentiation state) is the primary determinant of the level of heteroplasmy.

Janssen et al. (1999) showed that cells harboring patient-derived mitochondria with an A-to-G transition at nucleotide position 3243 display severe loss of respiration. Despite the low level of leucyl-tRNA-Leu(UUR), the rate of mitochondrial translation was not seriously affected by this mutation. Therefore, a decrease of mitochondrial protein synthesis as such did not appear to be a necessary prerequisite for loss of respiration. Rather, the mitochondrially encoded proteins seemed subject to elevated degradation, leading to a severe reduction in their steady state levels. The results were interpreted as favoring 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.

Borner et al. (2000) used an assay that combined tRNA oxidation and circularization to determine the relative amounts and states of aminoacylation of mutant and wildtype tRNAs in tissue samples from MELAS and MERRF (545000) patients. In most, but not all, biopsies from MELAS patients carrying the 3243A-G substitution, the mutant tRNA was underrepresented among processed and/or aminoacylated tRNAs. In contrast, in biopsies from MERRF patients harboring the 8344A-G substitution in the tRNA-lys gene (590060) neither the relative abundance nor the aminoacylation of the mutated tRNA was affected. The authors concluded that whereas the 3243A-G mutation may contribute to the pathogenesis of MELAS by reducing the amount of aminoacylated tRNA-leu, the 8344A-G mutation does not affect tRNA-lys function in MERRF patients in the same way.

Chomyn et al. (2000) presented several lines of evidence indicating that the protein synthesis defect in 3243A-G MELAS mutation-carrying cells is mainly due to a reduced association of mRNA with ribosomes, possibly as a consequence of the tRNA-leu(UUR) aminoacylation defect.

In a longitudinal study, Olsson et al. (2001) determined the proportion of the mitochondrial 3243A-G mutation in DNA obtained from cervical smear samples collected from 3 patients with maternally inherited diabetes and deafness. The proportion of mtDNA with the 3243A-G mutation decreased over time in the samples from all 3 patients: in patient 1 the mutant mtDNA decreased from 32 to 10% in 18 years, in patient 2 from 11 to 5% in 8 years, and in patient 3 from 26 to 19% in 4 years. This corresponded to a relative annual decline of 5.8% of the initial heteroplasmy level. The authors suggested that the observed decrease in mutational load with time was probably the consequence of negative selection acting against high levels of mutation load either at the level of cells or mitochondria due to impairment of the oxidative phosphorylation. The results may explain the observation of molecular genetic anticipation seen in some pedigrees with mitochondrial disorders. The percentage level of mutant mtDNA may be higher in the recent generation compared to the preceding one simply because the samplings were made on individuals of different ages in the pedigree.

The amount of 3243A-G heteroplasmy in blood tends to slowly decrease over time. Rahman et al. (2001) compared the levels of 3243A-G mutant mtDNA in blood at birth from Guthrie cards and at the time of diagnosis in a blood DNA sample from patients with MELAS syndrome. Paired blood DNA samples separated by 9 to 19 years were obtained from 6 patients. Quantification of mutant load demonstrated a decline (range 12 to 29%) in the proportion of mutant mtDNA in all cases. These results suggested that mutant mtDNA is slowly selected from rapidly dividing blood cells in MELAS. The authors showed that in elderly patients false-negative results may be obtained in testing for mitochondrial diseases when DNA from leukocytes is used. They proposed muscle and hair follicles as a better source of DNA.

Nam and Kang (2005) found that the 3243A-G mutation reduced, but did not eliminate, binding of recombinant human MTERF to the tRNA-leu(UUR) termination sequence.

Pyle et al. (2007) established an accurate fluorescent assay for 3243A-G heteroplasmy and the amount of mtDNA in blood with real-time PCR. The amount of mutated and wildtype mtDNA was measured at 2 time points in 11 subjects. The percentage of mutated mtDNA decreased exponentially during life, and peripheral blood leukocytes in patients harboring 3243A-G were profoundly depleted of mtDNA. A similar decrease in mtDNA had been seen in other mitochondrial disorders, and in 3243A-G cell lines in culture, indicating that depletion of mtDNA may be a common secondary phenomenon in several mitochondrial diseases. Thus, depletion of mtDNA is not always due to mutation of a nuclear gene involved in mtDNA maintenance.

Durham et al. (2007) showed that segments of human skeletal muscle fibers harboring 2 pathogenic mtDNA mutations retain normal cytochrome c oxidase (COX) activity by maintaining a minimum amount of wildtype mtDNA. For these mutations, direct measurements of mutated and wildtype mtDNA molecules within the same skeletal muscle fiber are consistent with the 'maintenance of wildtype' hypothesis, which predicts that there is nonselective proliferation of mutated and wildtype mtDNA in response to the molecular defect. However, for the 3243A-G mutation, a superabundance of wildtype mtDNA was found in many muscle fiber sections with negligible COX activity, indicating that the pathogenic mechanism for this particular mutation involves interference with the function of the wildtype mtDNA or wildtype gene products.

Using an anti-complex IV immunocapture technique and mass spectrometry, Janssen et al. (2007) found that COX I and COX II existed exclusively with the correct amino acid sequences in 3243A-G cells. The findings excluded a dominant-negative effect of the 3243A-G mutation, leaving tissue-specific accumulation by mtDNA segregation as the most likely cause of variable mitochondrial disease expression in patients with this mutation.

Rajasimha et al. (2008) observed an exponential decrease of mutant 3243A-G mtDNA in blood over time by simulating the segregation of the mutation in hematopoietic stem cells and leukocyte precursors. The findings were consistent with a selective process acting at the stem cell level and could explain why the level of mutant mtDNA in blood is almost always lower than that observed in nondividing tissues such as skeletal muscle. From human data (Rahman et al., 2001; Pyle et al., 2007), Rajasimha et al. (2008) derived a formula to correct for the change in heteroplasmy over time. The findings indicated that there is selection against pathogenic mtDNA in stem cells but not in committed blood cell precursors. The mechanism of loss of stem cells with high mutation levels was sufficient to explain the data, but other mechanisms may also be involved.

Sasarman et al. (2008) analyzed myoblasts isolated from a MELAS patient who was homoplasmic for the 3243A-G mutation. MELAS myoblasts showed only moderately affected translation of mitochondrial proteins, but an almost complete lack of assembly of respiratory chain complexes I, IV, and V. Pulse-chase labeling showed reduced stability of all mitochondrial translation products consistent with an assembly defect. There was also evidence of amino acid misincorporation in 3 polypeptides: MTCO3 (516050), MTCO2 (516040), and ATP6 (516060). The assembly defect could be partially rescued by overexpression of mitochondrial translation elongation factors EFTu (TUFM; 602389) or EFG2 (GFM2; 606544). The data indicated that the 3243A-G mutation produces both loss- and gain-of-function phenotypes, explaining the apparent discrepancy between the severity of the translation and respiratory chain assembly defects.

Yakubovskaya et al. (2010) identified 3243A as 1 of 3 nucleotides in leu-tRNA(UUR) that was everted during MTERF1 binding. The 3243A-G mutation reduced the affinity of MTERF1 binding to leu-tRNA(UUR).

Dvorakova et al. (2016) reported the comprehensive clinical phenotype of 50 Czech patients with the m.3243A-G mutation. Symptoms developed in 33 patients (66%) and 17 carriers remained unaffected (34%). The age of onset varied from 1 month to 47 years of age, with juvenile presentation occurring in 53% of patients. Myopathy was the most common presenting symptom (18%), followed by CPEO/ptosis and hearing loss, with the latter also being the most common second symptom. Stroke-like episodes occurred in 14 patients, although never as a first symptom, and were frequently preceded by migraines (58%). Rhabdomyolysis developed in 2 patients. The second symptom appeared 5.0 +/- 8.3 years (range 0-28 years) after the first, and the interval between the second and third symptom was 2.0 +/- 6.0 years (range 0-21 years). Four patients remained monosymptomatic up to 12 years of follow-up. The sequence of symptoms according to their time of manifestation was migraines, myopathy, seizures, CPEO/ptosis, stroke-like episodes, hearing loss, and diabetes mellitus. The average age at death was 32.4 +/- 17.7 years (range 9-60 years) in the juvenile form and 44.0 +/- 12.7 years (range 35-53 years) in the adult form. Some patients with stroke-like episodes harbored very low heteroplasmy levels in various tissues. No threshold for any organ dysfunction could be determined based on these levels.

POPULATION GENETICS

In an adult population of 245,201 individuals, Majamaa et al. (1998) studied the frequency of the 3243A-G mutation. In addition, they ascertained 615 patients with diabetes mellitus, sensorineural hearing impairment, epilepsy, occipital brain infarct, ophthalmoplegia, cerebral white matter disease, basal ganglia calcifications, hypertrophic cardiomyopathy, or ataxia, and examined 480 samples for the mutation. The mutation was found in 11 pedigrees, and its frequency was calculated to be 16.3/100,000 in the adult population. As determined by mtDNA haplotyping, the mutation had arisen in the population at least 9 times. Clinical evaluation of the probands revealed a syndrome that most frequently consisted of hearing impairment, cognitive decline, and short stature. The high prevalence of the common MELAS mutation in the adult population suggested that mitochondrial disorders constitute one of the largest diagnostic categories of neurogenetic diseases.

The low frequency of deleterious point mutations in mtDNA in human populations has suggested that they are under strong negative selection, and a reduced genetic fitness of mutation carriers had been assumed (Zeviani et al., 1998). In a population-based study of the 3243A-G mutation in the province of Oulu, Finland (Majamaa et al., 1998), Moilanen and Majamaa (2001) calculated the net reproduction rate for 31 mutation carriers and found it to be similar to that of the general population. The average fertility of mutation carriers was not reduced, and the generation time was not different between the mutation carriers and the general population. Moilanen and Majamaa (2001) stated that morbidity and mortality from the mutation may have changed compared to that of historical populations (e.g., diabetes mellitus may have proved fatal in earlier times but is now treatable). Moilanen and Majamaa (2001) concluded that the low frequency of this mutation in populations may still be explained by a mild uniform selection or a selection that depends on the degree of the mutant mtDNA heteroplasmy; alternatively, there may be no host-level selection at all, implying that other factors are responsible for the low population frequency.

Among 230 patients with sensorineural hearing loss in otorhinolaryngology clinics in Japan, Nagata et al. (2001) found 4 instances of the 3243A-G mutation (1.74%). Association of maternally inherited diabetes mellitus, cardiomyopathy, a family history of possible maternal inheritance of sensorineural hearing loss, and an onset of sensorineural hearing loss between the teenage years and the forties were signs suggesting the mutation.

Using high-resolution restriction fragment length polymorphism analysis and control-region sequencing, Torroni et al. (2003) studied 35 mtDNAs from Spain that harbored the 3243A-G mutation in association with either MELAS or a wide array of disease phenotypes. A total of 34 different haplotypes were found, indicating that all instances of the 3243A-G mutation are probably due to independent mutational events. Haplotypes were distributed into 13 haplogroups whose frequencies were close to those of the general Spanish population. Moreover, there was no statistically significant difference in haplogroup distribution between patients with MELAS and those with disease phenotypes other than MELAS. Torroni et al. (2003) concluded that the 3243A-G mutation may harbor all the evolutionary features expected from a severely deleterious mtDNA mutation under strong negative selection, and they reveal that European mtDNA backgrounds do not play a substantial role in modulating the mutation's phenotypic expression.

Uusimaa et al. (2007) estimated the prevalence of the 3243A-G mutation in northern Finland to be 18.4 in 100,000 children. The most common clinical manifestations appearing in childhood were sensorineural hearing loss, short stature or delayed maturation, migraine, learning difficulties, and exercise intolerance. Segregation analyses did not show a correlation between mother and child heteroplasmy, arguing against the random drift hypothesis.

Elliott et al. (2008) determined the frequency of 10 mitochondrial point mutations in 3,168 neonatal cord blood samples from sequential live births in North Cumbria in England, analyzing matched maternal blood samples to estimate the de novo mutation rate. Mitochondrial DNA mutations were detected in 15 offspring (0.54%, 95% confidence interval = 0.30-0.89%). Of these live births, 0.00107% (95% confidence interval = 0.00087-0.0127) harbored a mutation not detected in the mother's blood, providing an estimate of the de novo mutation rate. The most common mutation was mitochondrial 3243A-G. Elliott et al. (2008) concluded that at least 1 in 200 healthy humans harbors a pathogenic mitochondrial DNA mutation that potentially causes disease in the offspring of female carriers.