- Email: [email protected]
Leigh encephalopathy: Histologic and biochemical analyses of muscle biopsies
Leigh Encephalopathy: Histologic and Biochemical Analyses of Muscle Biopsies T o s h i r o N a g a i , M D * t , Yu-ichi G o t o , M D * , Taro M a t s u o k a , M D * , R y o u i c h i S a k u t a , M D * , E t s u o Naito, MD*, Y a s u h i r o K u r o d a , MD*, a n d I k u y a N o n a k a , M D *
To elucidate the pathogenesis of Leigh encephalopathy, histologic, biochemical, and mitochondrial DNA analyses were performed on biopsied muscles from 33 patients with the clinical characteristics of this disorder. On muscle histochemistry, cytochrome c oxidase activity was decreased or absent in 7 patients (21%), although none had ragged-red fibers. In 2 patients with cytochrome c oxidase deficiency, staining for this enzyme was poor in the muscle fibers and fibroblasts but was normal in the arterial wail, indicating tissue-specific involvement. Ten patients (30%) had biochemical defects, including 2 with pyruvate dehydrogenase complex, 4 with cytochrome c oxidase, 1 with NADH-cytochrome c reductase (complex I), and 3 with multiple complex deficiencies. None of the 28 patients in whom muscle mitochondrial (mt)DNA was analyzed had DNA deletions or point mutation at nucleotide positions 3,243 or 8,344. These results indicate that the underlying defect in Leigh encephalopathy is heterogeneous because only 30% of patients had enzyme defects demonstrable in muscle biopsy material. Nagai T, Goto Y, Matsuoka T, Sakuta R, Naito E, Kuroda Y, Nonaka I. Leigh encephalopathy: Histologic and biochemical analyses of muscle biopsies. Pediatr Neurol 1992;8:328-32.
Because lactate and pyruvate levels in blood and cerebrospinal fluid (CSF) are increased in most patients [4], mitochondrial enzyme defects probably are responsible for the pathogenesis of Leigh encephalopathy. B iochemically, pyruvate carboxylase (PC) [5,6], pyruvate dehydrogenase complex (PDHC) [7-9], cytochrome c oxidase (COX) [1012], and complex I [13,14] deficiencies all have been reported, suggesting heterogeneous enzyme defects. No characteristic or specific histologic abnormalities on muscle biopsy have been reported in this disease, except in patients with COX and complex I deficiencies who had abnormal mitochondria by electron microscopy [10,14]. Among the distinct disorders dominated by central nervous system (CNS) involvement with lactic acidosis are mitochondrial myopathy, encephalopathy, lactic acidosis, and stroke-like episodes (MELAS); myoclonus, epilepsy associated with ragged-red fibers (MERRF); and chronic, progressive external ophthalmoplegia (CPEO). All of these diseases are known to have ragged-red fibers and (mt)DNA aberrations on muscle biopsies, in Leigh encephalopathy, however, no reports have described raggedred fibers or (mt)DNA abnormalities. In this report, we examined skeletal muscle samples by histologic, biochemical, and molecular biologic methods, to characterize the enzyme defect(s) in Leigh encephalopathy in more detail and to elucidate the pathogenetic mechanism.
Methods Introduction Leigh encephalopathy initially manifests most commonly in infancy or early childhood and is characterized by progressive neurologic deterioration consisting of hypotonia, mental impairment, nystagmus, and respiratory failure [1,2]. The final diagnosis is confirmed neuropathologically by the presence of symmetric multifocal necrotic lesions extending from the thalamus to the pons. Computed tomographic (CT) findings often suggest this diagnosis by the demonstration of symmetric, low-density lesions in the basal ganglia, thalamus, and pons which correlate with the neuropathologic findings [3].
From the *Division of Ultmstructural Research; National Institute of Neuroscience; National Center of Neurology and Psychiatry; tDepartment of Neurology; Tokyo Metropolitan Kiyose Children's Hospital; Tokyo, Japan; ~Department of Pediatrics; Tokushima University School of Medicine; Tokushima, Japan.
328
PEDIATRIC NEUROLOGY
Vol. 8 No. 5
From 1985 to 1990, we obtained muscle biopsies from 33 children (19 males, 14 females), ranging in age from 4 months to 10 years with the clinical diagnosis of Leigh encephalopathy. All had the following criteria: (1) Subacute, progressive nearodegenerative symptoms before the age of 2 years; (2) High lactate level in blood and/or CSF; and, (3) Symmetric low-density lesions in the basal ganglia on CT. The frequency of the main neurodegenerative symptoms was 33 of 33 with hypotonia (100%), 30 of 33 with regression (91%), 18 of 33 with apnea or irregular breathing (55%), and 13 of 33 with seizures (39%). The age at symptom onset ranged from birth to 20 months of age. Lactate levels ranged in blood from 22.2 to 67.0 mg/dl and in CSF from 20.1 to 87.0 mg/dl; lactatelpyrovate ratio in blood was 16.5-34.0 and in CSF was 17.3-23.8. CT findings consisted of focal, symmetric
Communications should be addressed to: Dr. Nagai; Division of Ultrastruetural Research; National Institute of Neuroscience; NCNP; Kodaira, Tokyo 187, Japan. Received March 23, 1992; accepted June 24, 1992.
Table 1. Clinical summary of patients with PDHC deficiency Lactate (mg/dl) Blood CSF
Patient No./ Age/Sex
Onset
Main Symptoms
1/7 mos/M
1 mo
Muscle weakness, irregular breathing, seizure
27.0
Gait disturbance, seizure, apnea
28.0
2/9 yrs/F
2 yrs
NP
PDHC Activity* Native Total
1.52
1.42
Outcome
Died at age 13 mos
47.0
Control (n = 13; mean + S.D.)
0.48
1.16
1.46 _+
4.46 + 1.18
0.52
On respirator when last examined at age 9 yrs
* mmol/min/mg protein. Abbreviation: NP = Not performed hypodensity involving the basal ganglia. No definite clinical difference existed between those with or without biochemical defects. The patients were selected retrospectively. Muscle samples were obtained from the biceps brachii (10 patients) or rectus femoris (23 patients) muscles. The samples were divided into 3 pieces for biochemistry, histochemistry, and electron microscopy. Biochemistry. Mitochondria were isolated from fresh muscles and stored at -80°C until analyzed. Spectrophotometric assays were applied to measure NADH-ubiquinone oxidoreductase [15], rotenone-sensitive NADH-cytochrome c reductase [16], succinate-ubiquinoneoxidoreductase (complex II) [17], succinate cytochrome c reductase, ubiquinolcytochrome c oxidoreductase (complex III) [18], and COX [19]. PDHC activity also was measured in all muscles [20]. Muscle Histochemistry. Muscle specimens were frozen in isopentane-cooled liquid nitrogen. Serial frozen sections were stained with hematoxylin and eosin, modified Gomori trichrome, and various histochemical methods [21], including succinate dehydrogenase (SDH) and COX. On ATPase stain, each fiber was classified into type 1, 2A, 2B, or 2C. On COX stain, muscle fiber enzyme activity in intrafusal fibers, fibroblasts, and blood vessel walls was evaluated. Analysis of (mt)DNA. Total DNA was prepared from a small portion of biopsied muscle. Southern blot analysis was performed, as described previously to detect deletions [22]. In addition, (mt)DNA was examined for point mutations at nucleotide positions 3,243 for MELAS [23] and 8,344 for MERRF.
Results
Biochemistry (Tables 1,2). In 2 patients, total PDHC muscle activity was decreased to 26% and 32% of the mean control value. Eight patients had significant decreases in mitochondrial respiratory chain enzyme activities; complex I deficiency in 1, C O X in 4, and multiple complexes in 3. No definite clinical differences were observed among these groups. Muscle Histochemistry (Table 2). Most of the biopsies d i s c l o s e d subtle abnormalities, including type 2 fiber atrophy in 29 patients (88%; particularly type 2B fibers in 13), increased number of type 2C fibers in 11 (33%), type 1 fiber predominance in 4 (12%), increased amount of lipid droplets in 2 (6%), and fiber type grouping in 3 (9%). Seven patients with biochemical defect o f C O X activity, including 3 patients with multiple respiratory chain enz y m e defects, revealed diffusely decreased density on
COX stain. All tissue components of muscle biopsy in Patients 3, 5, and 9 had no enzyme activity while tissuespecific involvement was observed in 4 patients; Patients 4 and 7 had C O X activity only in arterial vessels with almost complete absence o f the enzyme in the extra- and intra-fusal fibers and fibroblasts (Table 2; Fig 1). No ragged-red fibers were found in any of the muscle samples. Most of the arterioles in Patients 4 and 7 were stained dark with SDH and disclosed the morphologic characteristics of the strongly SDH-reactive blood vessels (SSV). Analyses of Mitochondrial DNA. Neither deletion of (mt)DNA detectable by Southern blot analysis nor (mt)DNA mutations at nucleotide positions 8,344 or 3,243 were demonstrated. Discussion
The biochemical defects in Leigh encephalopathy are probably heterogeneous because PC, PDHC, COX, and complex I deficiencies all have been reported. PC deficiency is probably a secondarily induced abnormality [12, 25]. DeVivo et al. analyzed mitochondrial enzyme activities in fibroblasts from 23 patients with Leigh encephalopathy and found COX deficiency in 2 patients, PDHC deficiency in 1, and intermediate PC activities in 2 [26]. Similarly, Miyabayashi et al. reported 2 patients with C O X and 4 with PDHC deficiency in fibroblasts from 28 patients [27]; Robinson et al. reported 6 patients with C O X and 4 with complex I deficiency among 95 patients [28]. Enzyme defects in mitochondrial encephalomyopathies often reveal tissue-specific involvement sparing fibroblasts. For instance, in the fatal infantile form of C O X deficiency, the enzyme activity is defective in skeletal muscle fibers but not in fibroblasts [29]. No enzymatic or (mt)DNA defect is detectable in fibroblasts and blood from most patients with CPEO who have focal COX deficiency and (mt)DNA mutation in the muscle specimens [30]. In Leigh encephalopathy caused by C O X deficiency, decreased COX activity sometimes has been found in extracts of various organs but not in cultured skin fibroblasts
Nagai et al: Leigh Encephalopathy 329
Z
z
Z
z
Z
z
z
,q
=e.+
o.
, ~...., +1 ~.o
t..q t-q
r
I
I
+
i
I
I
I
I
I
1
+1
I
+1
I
¢
i
-t-
+
I
+
+~
I
-b
8 °~
..= b °~
o.
_:im
~
CD
Z
~ d
..=
.o eh
d~
Z: =
O d::
<=
=. r~
+
b.~
+
'= 9
i II
8 C~
•g z ~R z u o ~ . e,i
330
II
PEDIATRIC
NEUROLOGY
Vol. 8 No. 5
II
Ii
II
II
N z II
II
II
Figure I. Muscle biopsy, Patient 7, COX stain is positive only in the media of the arterial wall (arrow): original magnification x330.
[12]. Even though only 1 of 7 patients (Patient 3) with respiratory chain enzyme defects also had decreased COX activity in fibroblasts (Table 2), above cited reports [29,30] suggested that the enzyme analysis performed on fibroblasts may be falsely negative. We, therefore, stress that it is also necessary to evaluate the biochemical defect on biopsied muscle besides skin fibroblasts in patients with Leigh encephalopathy. Histochemical study of muscles from patients with Leigh encephalopathy has, so far, not been fully evaluated. Only 2 studies reported ragged-red fibers in muscle biopsies [31,32]. The patient reported by Crosby was later identified as having Kearns-Sayre syndrome [33]. In our study, no specific morphologic changes were observed in this disease and the principal change was type 2, particularly type 2B, fiber atrophy which is known to be a nonspecific change induced by a variety of conditions, including CNS disorders, disuse, malnutrition, and aging. An increased number of type 2C fibers was found in 11 patients, suggesting a delay in muscle fiber type differentiation. In 7 patients, COX activity was diffusely absent in skeletal muscle fibers using histochemical methods. Complete concordance existed between histochemical and biochemical assessment in our materials. None had raggedred fibers. We believe that muscle histochemistry, except COX deficiency, is not helpful in the diagnosis of Leigh encephalopathy and that high blood and CSF lactate without ragged-red fibers may support the diagnosis of this disease. On COX stain, tissue-specific involvement was observed in 4 of 7 patients with COX and multiple respiratory enzyme complex deficiencies. Two patients had COX activity in the arterial walls and 2 in fibroblasts, revealing tissue-specific involvement. Two patients had abnormal blood vessels with the histochemical characteristics of the
SSV which are commonly observed in muscle biopsies from MELAS patients and suggesting systemic vascular involvement [34]. These findings are compatible with various clinical features of this disease, including strokelike episodes in some patients with Leigh encephalopathy. Although most cases of Leigh encephalopathy are sporadic, there is increasing evidence that Leigh encephalopathy with COX deficiency is inherited as an autosomalrecessive trait [35]. In our study, 2 patients with COX deficiency and 2 with unknown enzyme defects had affected siblings; 1 with COX deficiency was the product of consanguineous parents. No sexual preponderance was observed in either familial or nonfamilial patients. When respiratory enzyme defects cause MELAS, MERRF, or CPEO, patients usually have ragged-red fibers and (mt)DNA aberrations; however, when respiratory enzyme defects clinically cause Leigh encephalopathy, patients have neither ragged-red fibers nor definite (mt)DNA abnormalities. This finding may indicate that the pathogenesis of Leigh encephalopathy is different from that of MELAS, MERRF, and CPEO. Because the biogenesis of respiratory enzymes are under dual control from mitochondrial and nuclear genomes, nuclear DNA mutation may cause Leigh encephalopathy, in contrast with (mt)DNA mutation causing MELAS, MERRF, and CPEO.
The authors would like to thank Dr. Nobutake Matsuo, Professor of Pediatrics, Keio University School of Medicine, for his critical reading of the manuscript.
References
[1] Leigh D. Subacute necrotizing encephalomyelopathy in an infant. J Neurol Neurosurg Psychiatry 1951; 14:216-21.
Nagai et al: Leigh Encephalopathy 331
[2] DeVivo DC, DiMauro S, Rapin 1. Mitochondrial disorders. In: Rudolph AM, ed. Pediatrics, 18th ed. Norwalk: Appleton & Lange, 1987;1736-8. [3] Schwartz W J, Hutchinson HT, Berg BO. Computerized tomography in subacute necrotizing encephalomyelopathy. Ann Neurol 1981; 10:268-71. [4] Worsely HE, Brookfield RW, Elwood JS, Noble RL, Taylor WH. Lactic acidosis with necrotizing eneephalopathy in two sibs. Arch Dis Child 1965;40:492-501. [5] Hommes FA, Polman HA, Reering JD. Leigh encephalomyelopathy: An inborn error of gluconeogenesis. Arch Dis Child 1968; 43:423-6. [6] Tang TT, Good TA, Dyken PR, et al. Pathogenesis of Leigh's encephalomyelopathy. J Pediatr 1972;81:189-90. [7] Farmer TW, Veath L, Miller AL, O'Brien J, Rosenberg RN. Pyruvate decarboxylase deficiency in a patient with subacute necrotizing encephalomyelopathy. Neurology 1973;23:429. [8] Kretzsehmar HA, DeArmond S J, Koch TK, et al. Pyruvate dehydrogenase complex deficiency as a cause of subacute necrotizing encephalopathy (Leigh disease). Pediatrics 1987;79:370-3. [9] DeVivo DC, Uziel G. Disturbance of pyruvate metabolism in neuromuscular disease. In: Scalato G, Cerri C, eds. Mitochondrial pathology in muscle disease. Padova: Piccin Medical Books, 1983;58-70. [10] Willems JL, Monnens LAH, Trijhels JMF, et al. Leigh encephalomyelopathy in a patient with cytochrome c oxidase deficiency in muscle tissue. Pediatrics 1977;60:850-7. [11] Miyabayashi S, Narisawa K, Tada K. Two siblings with cytochrome c oxidase deficiency. J Inherited Metab Dis 1983;6:121-2. [12] DiMauro S, Servidei S, Zeviani M, et al. Cytochrome c oxidase deficiency in Leigh syndrome. Ann Neurol 1987;22:498-506. [13] Van Erven PMM, Cahreels FJM, Ruitenbeek W, et al. Subacute necrotizing encephalomyelopathy (Leigh syndrome) associated with disturbed oxidation of pyruvate, malate and 2-oxoglutarate in muscle and liver. Acta Neurol Scand 1985;72:36-42. [14] Fujii T, Ito M, Okuno T, Mutoh K, Nishikomori R, Mikawa H. Complex I (reduced nicotinamide-adenine dinucleotide-coenzyme Q reductase) deficiency in two patients with probable Leigh syndrome. J Pediatr 1990; 116:84-7. [15] Hatefi Y, Rieske JS. Preparation and properties of DPNHcoenzyme Q reductase (complex I of the respiratory chain). Methods Enzymol 1967;10:235-9. [16] Macider B. Microsomal DPNH-cytochrome e reductase. Methods Enzymol 1967; 10:551-3. [17] Hatefi Y, Stiggal DL. Preparation and properties of succinate: Ubiquinone oxidoreductase (complex II). Methods Enzymol 1978;53: 21-7. [18] Shimomura Y, Nishikimi M, Ozawa T. Isolation and reconstitution of the iron-sulfur protein in ubiquinol-cytochrome c oxidoreductase complex. J Biol Chem 1984;259:14059-63.
332
PEDIATRIC NEUROLOGY
Vol. 8 No. 5
[19] Orii Y, Okunuki K. Studies on cytochrome ft. ~'yt
To elucidate the pathogenesis of Leigh encephalopathy, histologic, biochemical, and mitochondrial DNA analyses were performed on biopsied muscles from 33 patients with the clinical characteristics of this disorder. On muscle histochemistry, cytochrome c oxidase activity was decreased or absent in 7 patients (21%), although none had ragged-red fibers. In 2 patients with cytochrome c oxidase deficiency, staining for this enzyme was poor in the muscle fibers and fibroblasts but was normal in the arterial wail, indicating tissue-specific involvement. Ten patients (30%) had biochemical defects, including 2 with pyruvate dehydrogenase complex, 4 with cytochrome c oxidase, 1 with NADH-cytochrome c reductase (complex I), and 3 with multiple complex deficiencies. None of the 28 patients in whom muscle mitochondrial (mt)DNA was analyzed had DNA deletions or point mutation at nucleotide positions 3,243 or 8,344. These results indicate that the underlying defect in Leigh encephalopathy is heterogeneous because only 30% of patients had enzyme defects demonstrable in muscle biopsy material. Nagai T, Goto Y, Matsuoka T, Sakuta R, Naito E, Kuroda Y, Nonaka I. Leigh encephalopathy: Histologic and biochemical analyses of muscle biopsies. Pediatr Neurol 1992;8:328-32.
Because lactate and pyruvate levels in blood and cerebrospinal fluid (CSF) are increased in most patients [4], mitochondrial enzyme defects probably are responsible for the pathogenesis of Leigh encephalopathy. B iochemically, pyruvate carboxylase (PC) [5,6], pyruvate dehydrogenase complex (PDHC) [7-9], cytochrome c oxidase (COX) [1012], and complex I [13,14] deficiencies all have been reported, suggesting heterogeneous enzyme defects. No characteristic or specific histologic abnormalities on muscle biopsy have been reported in this disease, except in patients with COX and complex I deficiencies who had abnormal mitochondria by electron microscopy [10,14]. Among the distinct disorders dominated by central nervous system (CNS) involvement with lactic acidosis are mitochondrial myopathy, encephalopathy, lactic acidosis, and stroke-like episodes (MELAS); myoclonus, epilepsy associated with ragged-red fibers (MERRF); and chronic, progressive external ophthalmoplegia (CPEO). All of these diseases are known to have ragged-red fibers and (mt)DNA aberrations on muscle biopsies, in Leigh encephalopathy, however, no reports have described raggedred fibers or (mt)DNA abnormalities. In this report, we examined skeletal muscle samples by histologic, biochemical, and molecular biologic methods, to characterize the enzyme defect(s) in Leigh encephalopathy in more detail and to elucidate the pathogenetic mechanism.
Methods Introduction Leigh encephalopathy initially manifests most commonly in infancy or early childhood and is characterized by progressive neurologic deterioration consisting of hypotonia, mental impairment, nystagmus, and respiratory failure [1,2]. The final diagnosis is confirmed neuropathologically by the presence of symmetric multifocal necrotic lesions extending from the thalamus to the pons. Computed tomographic (CT) findings often suggest this diagnosis by the demonstration of symmetric, low-density lesions in the basal ganglia, thalamus, and pons which correlate with the neuropathologic findings [3].
From the *Division of Ultmstructural Research; National Institute of Neuroscience; National Center of Neurology and Psychiatry; tDepartment of Neurology; Tokyo Metropolitan Kiyose Children's Hospital; Tokyo, Japan; ~Department of Pediatrics; Tokushima University School of Medicine; Tokushima, Japan.
328
PEDIATRIC NEUROLOGY
Vol. 8 No. 5
From 1985 to 1990, we obtained muscle biopsies from 33 children (19 males, 14 females), ranging in age from 4 months to 10 years with the clinical diagnosis of Leigh encephalopathy. All had the following criteria: (1) Subacute, progressive nearodegenerative symptoms before the age of 2 years; (2) High lactate level in blood and/or CSF; and, (3) Symmetric low-density lesions in the basal ganglia on CT. The frequency of the main neurodegenerative symptoms was 33 of 33 with hypotonia (100%), 30 of 33 with regression (91%), 18 of 33 with apnea or irregular breathing (55%), and 13 of 33 with seizures (39%). The age at symptom onset ranged from birth to 20 months of age. Lactate levels ranged in blood from 22.2 to 67.0 mg/dl and in CSF from 20.1 to 87.0 mg/dl; lactatelpyrovate ratio in blood was 16.5-34.0 and in CSF was 17.3-23.8. CT findings consisted of focal, symmetric
Communications should be addressed to: Dr. Nagai; Division of Ultrastruetural Research; National Institute of Neuroscience; NCNP; Kodaira, Tokyo 187, Japan. Received March 23, 1992; accepted June 24, 1992.
Table 1. Clinical summary of patients with PDHC deficiency Lactate (mg/dl) Blood CSF
Patient No./ Age/Sex
Onset
Main Symptoms
1/7 mos/M
1 mo
Muscle weakness, irregular breathing, seizure
27.0
Gait disturbance, seizure, apnea
28.0
2/9 yrs/F
2 yrs
NP
PDHC Activity* Native Total
1.52
1.42
Outcome
Died at age 13 mos
47.0
Control (n = 13; mean + S.D.)
0.48
1.16
1.46 _+
4.46 + 1.18
0.52
On respirator when last examined at age 9 yrs
* mmol/min/mg protein. Abbreviation: NP = Not performed hypodensity involving the basal ganglia. No definite clinical difference existed between those with or without biochemical defects. The patients were selected retrospectively. Muscle samples were obtained from the biceps brachii (10 patients) or rectus femoris (23 patients) muscles. The samples were divided into 3 pieces for biochemistry, histochemistry, and electron microscopy. Biochemistry. Mitochondria were isolated from fresh muscles and stored at -80°C until analyzed. Spectrophotometric assays were applied to measure NADH-ubiquinone oxidoreductase [15], rotenone-sensitive NADH-cytochrome c reductase [16], succinate-ubiquinoneoxidoreductase (complex II) [17], succinate cytochrome c reductase, ubiquinolcytochrome c oxidoreductase (complex III) [18], and COX [19]. PDHC activity also was measured in all muscles [20]. Muscle Histochemistry. Muscle specimens were frozen in isopentane-cooled liquid nitrogen. Serial frozen sections were stained with hematoxylin and eosin, modified Gomori trichrome, and various histochemical methods [21], including succinate dehydrogenase (SDH) and COX. On ATPase stain, each fiber was classified into type 1, 2A, 2B, or 2C. On COX stain, muscle fiber enzyme activity in intrafusal fibers, fibroblasts, and blood vessel walls was evaluated. Analysis of (mt)DNA. Total DNA was prepared from a small portion of biopsied muscle. Southern blot analysis was performed, as described previously to detect deletions [22]. In addition, (mt)DNA was examined for point mutations at nucleotide positions 3,243 for MELAS [23] and 8,344 for MERRF.
Results
Biochemistry (Tables 1,2). In 2 patients, total PDHC muscle activity was decreased to 26% and 32% of the mean control value. Eight patients had significant decreases in mitochondrial respiratory chain enzyme activities; complex I deficiency in 1, C O X in 4, and multiple complexes in 3. No definite clinical differences were observed among these groups. Muscle Histochemistry (Table 2). Most of the biopsies d i s c l o s e d subtle abnormalities, including type 2 fiber atrophy in 29 patients (88%; particularly type 2B fibers in 13), increased number of type 2C fibers in 11 (33%), type 1 fiber predominance in 4 (12%), increased amount of lipid droplets in 2 (6%), and fiber type grouping in 3 (9%). Seven patients with biochemical defect o f C O X activity, including 3 patients with multiple respiratory chain enz y m e defects, revealed diffusely decreased density on
COX stain. All tissue components of muscle biopsy in Patients 3, 5, and 9 had no enzyme activity while tissuespecific involvement was observed in 4 patients; Patients 4 and 7 had C O X activity only in arterial vessels with almost complete absence o f the enzyme in the extra- and intra-fusal fibers and fibroblasts (Table 2; Fig 1). No ragged-red fibers were found in any of the muscle samples. Most of the arterioles in Patients 4 and 7 were stained dark with SDH and disclosed the morphologic characteristics of the strongly SDH-reactive blood vessels (SSV). Analyses of Mitochondrial DNA. Neither deletion of (mt)DNA detectable by Southern blot analysis nor (mt)DNA mutations at nucleotide positions 8,344 or 3,243 were demonstrated. Discussion
The biochemical defects in Leigh encephalopathy are probably heterogeneous because PC, PDHC, COX, and complex I deficiencies all have been reported. PC deficiency is probably a secondarily induced abnormality [12, 25]. DeVivo et al. analyzed mitochondrial enzyme activities in fibroblasts from 23 patients with Leigh encephalopathy and found COX deficiency in 2 patients, PDHC deficiency in 1, and intermediate PC activities in 2 [26]. Similarly, Miyabayashi et al. reported 2 patients with C O X and 4 with PDHC deficiency in fibroblasts from 28 patients [27]; Robinson et al. reported 6 patients with C O X and 4 with complex I deficiency among 95 patients [28]. Enzyme defects in mitochondrial encephalomyopathies often reveal tissue-specific involvement sparing fibroblasts. For instance, in the fatal infantile form of C O X deficiency, the enzyme activity is defective in skeletal muscle fibers but not in fibroblasts [29]. No enzymatic or (mt)DNA defect is detectable in fibroblasts and blood from most patients with CPEO who have focal COX deficiency and (mt)DNA mutation in the muscle specimens [30]. In Leigh encephalopathy caused by C O X deficiency, decreased COX activity sometimes has been found in extracts of various organs but not in cultured skin fibroblasts
Nagai et al: Leigh Encephalopathy 329
Z
z
Z
z
Z
z
z
,q
=e.+
o.
, ~...., +1 ~.o
t..q t-q
r
I
I
+
i
I
I
I
I
I
1
+1
I
+1
I
¢
i
-t-
+
I
+
+~
I
-b
8 °~
..= b °~
o.
_:im
~
CD
Z
~ d
..=
.o eh
d~
Z: =
O d::
<=
=. r~
+
b.~
+
'= 9
i II
8 C~
•g z ~R z u o ~ . e,i
330
II
PEDIATRIC
NEUROLOGY
Vol. 8 No. 5
II
Ii
II
II
N z II
II
II
Figure I. Muscle biopsy, Patient 7, COX stain is positive only in the media of the arterial wall (arrow): original magnification x330.
[12]. Even though only 1 of 7 patients (Patient 3) with respiratory chain enzyme defects also had decreased COX activity in fibroblasts (Table 2), above cited reports [29,30] suggested that the enzyme analysis performed on fibroblasts may be falsely negative. We, therefore, stress that it is also necessary to evaluate the biochemical defect on biopsied muscle besides skin fibroblasts in patients with Leigh encephalopathy. Histochemical study of muscles from patients with Leigh encephalopathy has, so far, not been fully evaluated. Only 2 studies reported ragged-red fibers in muscle biopsies [31,32]. The patient reported by Crosby was later identified as having Kearns-Sayre syndrome [33]. In our study, no specific morphologic changes were observed in this disease and the principal change was type 2, particularly type 2B, fiber atrophy which is known to be a nonspecific change induced by a variety of conditions, including CNS disorders, disuse, malnutrition, and aging. An increased number of type 2C fibers was found in 11 patients, suggesting a delay in muscle fiber type differentiation. In 7 patients, COX activity was diffusely absent in skeletal muscle fibers using histochemical methods. Complete concordance existed between histochemical and biochemical assessment in our materials. None had raggedred fibers. We believe that muscle histochemistry, except COX deficiency, is not helpful in the diagnosis of Leigh encephalopathy and that high blood and CSF lactate without ragged-red fibers may support the diagnosis of this disease. On COX stain, tissue-specific involvement was observed in 4 of 7 patients with COX and multiple respiratory enzyme complex deficiencies. Two patients had COX activity in the arterial walls and 2 in fibroblasts, revealing tissue-specific involvement. Two patients had abnormal blood vessels with the histochemical characteristics of the
SSV which are commonly observed in muscle biopsies from MELAS patients and suggesting systemic vascular involvement [34]. These findings are compatible with various clinical features of this disease, including strokelike episodes in some patients with Leigh encephalopathy. Although most cases of Leigh encephalopathy are sporadic, there is increasing evidence that Leigh encephalopathy with COX deficiency is inherited as an autosomalrecessive trait [35]. In our study, 2 patients with COX deficiency and 2 with unknown enzyme defects had affected siblings; 1 with COX deficiency was the product of consanguineous parents. No sexual preponderance was observed in either familial or nonfamilial patients. When respiratory enzyme defects cause MELAS, MERRF, or CPEO, patients usually have ragged-red fibers and (mt)DNA aberrations; however, when respiratory enzyme defects clinically cause Leigh encephalopathy, patients have neither ragged-red fibers nor definite (mt)DNA abnormalities. This finding may indicate that the pathogenesis of Leigh encephalopathy is different from that of MELAS, MERRF, and CPEO. Because the biogenesis of respiratory enzymes are under dual control from mitochondrial and nuclear genomes, nuclear DNA mutation may cause Leigh encephalopathy, in contrast with (mt)DNA mutation causing MELAS, MERRF, and CPEO.
The authors would like to thank Dr. Nobutake Matsuo, Professor of Pediatrics, Keio University School of Medicine, for his critical reading of the manuscript.
References
[1] Leigh D. Subacute necrotizing encephalomyelopathy in an infant. J Neurol Neurosurg Psychiatry 1951; 14:216-21.
Nagai et al: Leigh Encephalopathy 331
[2] DeVivo DC, DiMauro S, Rapin 1. Mitochondrial disorders. In: Rudolph AM, ed. Pediatrics, 18th ed. Norwalk: Appleton & Lange, 1987;1736-8. [3] Schwartz W J, Hutchinson HT, Berg BO. Computerized tomography in subacute necrotizing encephalomyelopathy. Ann Neurol 1981; 10:268-71. [4] Worsely HE, Brookfield RW, Elwood JS, Noble RL, Taylor WH. Lactic acidosis with necrotizing eneephalopathy in two sibs. Arch Dis Child 1965;40:492-501. [5] Hommes FA, Polman HA, Reering JD. Leigh encephalomyelopathy: An inborn error of gluconeogenesis. Arch Dis Child 1968; 43:423-6. [6] Tang TT, Good TA, Dyken PR, et al. Pathogenesis of Leigh's encephalomyelopathy. J Pediatr 1972;81:189-90. [7] Farmer TW, Veath L, Miller AL, O'Brien J, Rosenberg RN. Pyruvate decarboxylase deficiency in a patient with subacute necrotizing encephalomyelopathy. Neurology 1973;23:429. [8] Kretzsehmar HA, DeArmond S J, Koch TK, et al. Pyruvate dehydrogenase complex deficiency as a cause of subacute necrotizing encephalopathy (Leigh disease). Pediatrics 1987;79:370-3. [9] DeVivo DC, Uziel G. Disturbance of pyruvate metabolism in neuromuscular disease. In: Scalato G, Cerri C, eds. Mitochondrial pathology in muscle disease. Padova: Piccin Medical Books, 1983;58-70. [10] Willems JL, Monnens LAH, Trijhels JMF, et al. Leigh encephalomyelopathy in a patient with cytochrome c oxidase deficiency in muscle tissue. Pediatrics 1977;60:850-7. [11] Miyabayashi S, Narisawa K, Tada K. Two siblings with cytochrome c oxidase deficiency. J Inherited Metab Dis 1983;6:121-2. [12] DiMauro S, Servidei S, Zeviani M, et al. Cytochrome c oxidase deficiency in Leigh syndrome. Ann Neurol 1987;22:498-506. [13] Van Erven PMM, Cahreels FJM, Ruitenbeek W, et al. Subacute necrotizing encephalomyelopathy (Leigh syndrome) associated with disturbed oxidation of pyruvate, malate and 2-oxoglutarate in muscle and liver. Acta Neurol Scand 1985;72:36-42. [14] Fujii T, Ito M, Okuno T, Mutoh K, Nishikomori R, Mikawa H. Complex I (reduced nicotinamide-adenine dinucleotide-coenzyme Q reductase) deficiency in two patients with probable Leigh syndrome. J Pediatr 1990; 116:84-7. [15] Hatefi Y, Rieske JS. Preparation and properties of DPNHcoenzyme Q reductase (complex I of the respiratory chain). Methods Enzymol 1967;10:235-9. [16] Macider B. Microsomal DPNH-cytochrome e reductase. Methods Enzymol 1967; 10:551-3. [17] Hatefi Y, Stiggal DL. Preparation and properties of succinate: Ubiquinone oxidoreductase (complex II). Methods Enzymol 1978;53: 21-7. [18] Shimomura Y, Nishikimi M, Ozawa T. Isolation and reconstitution of the iron-sulfur protein in ubiquinol-cytochrome c oxidoreductase complex. J Biol Chem 1984;259:14059-63.
332
PEDIATRIC NEUROLOGY
Vol. 8 No. 5
[19] Orii Y, Okunuki K. Studies on cytochrome ft. ~'yt