Respiratory chain failure in adult muscle fibres: relationship with ageing and possible implications for the neuronal pool
E d w a r d Byrne a a n d X e n i a D e n n e t t b "Department of Neurological Sciences, St. Fincent's Hospital, Melbourne, Vic. 3065, Australia and h State Neuropathology Serrice.
C / o Unicersity of Melbourne, Parkt'ille, V[c. 3052, Australia
(Received 9 March 1992) (Revision received 12 May 1992)
(Accepted 19 May 1992)
Keywords: Respiratory chain failure; Human muscle
Summary
A histochemical analysis of mitochondrial enzyme activity was carried out in 103 human diaphrag- matic skeletal muscles from 49 subjects of different ages, obtaine,' -ither at the time of abdominal surgery or at necropsy. Evidence of respiratory failure (cytochrom~. oxidase negativity) was seen in occasional fibres from the fourth decade on with an approximate 10-fold increase between the fourth and ninth decade (0.16% to 2.85%). A similar incidence of mitochondrial failure in CNS neurones to that documented in skeletal muscle could easily account for attrition of 25% of neurones over a 50-year period as reported in the literature. Possible theoretical relationships between morphological markers of mitochondrial failure and cell attrition are explored. While the projections from muscle to neurone are somewhat speculative, it is clear that if a similar extent of mitochondrial pathology exists in the brain to that documented in skeletal muscle, this could easily account for neuronal loss in the ageing brain.
The ageing phenomenon is a highly complex one involving many factors both genetic and envi- ronmental. Even where major acquired illnesses and environmental stresses are avoided, however, it is clear that there is a steady deterioration in function in certain tissues from middle age on- wards. This is most marked in tissues which lack regenerative capacity, either partially, as in the case of skeletal muscle, or totally, as in the case of CNS neurones. The CNS neuronal pool is fully
Correspondence: Assoc. Professor E. Byrne, Department of Neurological Sciences, St. Vincent's Hospital, Melbourne, Vic. 3065, Australia.
established by the age of 2 years (and is therefore a fixed post-mitotic cell population). Although the time of onset varies greatly, with some indi- viduals maintaining a vigorous intellect in ad- vanced old age, it is likely that evtmtually certain CNS functions begin to fail in everyone if they live long enough, in people who develop intellec- tual failure relatively early, this is one of the most troublesome features of ageing. This general global intellectual decline is probably distinct from the severe dementia of Aizheimer type where an additional genetic or environmental in- sult may operate.
A number of studies have investigated the rate of fall-out of CNS neurones with ageing. Early
studies suggested a loss of between 15 and 25% of cortical neurones from early to late adult life (Brody, 1955). The difficulties in measuring neu- ronal numbers accurately are considerable but more recent work tends to confirm that a neu- ronal loss of this order occurs although with a regional emphasis (Hanley, 1974). A fall-out of approximately 2.5% per decade has been demon- strated in cerebellar Purkinje cells (Hall et al., 1975) and approximately 25% of spinal cord mo- tor neurones were )ost in the tenth decade com- pared to the second decade as demonstrated by Tomlinson and Irving (1977). It would appear that a fall-out of approximately 25% of CNS neurones in certain areas occurs between the ages of 40 and 90 years, an attrition rate of approximately 0.5% per year, if distributed uni- formly over this period, although it is likely that lower rates occur in the younger decades with an acceleration in later years. The extreme slowness of this process leads to several important observa- tions.
Only a small proportion of neurones will show evidence of degeneration at any one time, the exact proportion depending on the time interval between the first morphological evidence of dys- function to the death and disappearance of the cell. If. for example, the process of cell death is extremely slow, say over 1 year, then 0,5% of nerve cells will show morphological changes at any one time, if 0,5% of nerve cells die over that year and morphological changes are evident very early in the cell failure process. The crucial point is that there is every likelihood that morphologi- cal markers which either identify degenerate cells or give clues to the underlying process, will be evident in only a very small proportion of cells at any one time. This theme is developed further in Table 1, where the time interval between the development of histochemical evidence of respi- ratory failure to cell death and removal is calcu- lated for a range of percentages of cytochrome oxidase negative fibres and for a range of annual cell attrition rates.
Considerable evidence has accumulated that a failure of mitochondrial energy generation may be important in the ageing process. A significant fall in the efficiency of state I11 mitochondrial respiration has been demonstrated in intact mite-
TABLE 1
THEORETICAL CONSTRUCT LINKING PROPORTION OF CYTOCHROME OXIDASE NEGATIVE FIBRES, AN- NUAL RATE OF CELL ATTRITION AND SURVIVAL TIME OF INDIVIDUAL CELLS ENTERING ENERGY FAILURE FROM THE TIME OF ONSET OF ENERGY FAILURE DEMONSTRATED MORPHOLOGICALLY TO CELL FALL-OUT (DAYS)
chondria isolated from human skeletal muscle with rates from patients in their 80s being about half those of younger subjects (Trounce et al., 1989) and with the identification of occasional striated muscle fibres having a severe respiratory chain failure as evidenced by lack of cytochrome oxidase activity in both cardiomyocytes and hu- man diaphragm (Mtiller-H6cker, 1989, 1990; Byrne et al., 1991). Mtiller-H6cker (1989) demon- strated mitochondrial abnormalities in the human heart as early as the second decade, whereas in limb muscle the defects appeared in the third decade. A lO-fold increase in the incidence of defective fibres was noted for those in the eighth to ninth decade when compared with those be- tween the third and sixth decade, in parallel with these findings of phenotypic evidence of defective mitochondrial function, a number of groups have now identified an age related accumulation of large deletions in the human mitochondrial genome, the first observations having been made by Linnane's group (Linnane et al., 1991; Cor- topassi and Arnheim, 1990) and in rats by Torii et al. (1992). The significance of these findings has been queried as the deletions are only detected after polymerase chain reaction enhancement and presumably are present as a low proportion of total mitochondrial DNA content. The same con- siderations can be applied to the histochemicai studies demonstrating cytochrome oxidase nega- tive fibres, as the proportion of these, especially
in the middle adult years, is rather small. These arguments, as pointed out in the earlier discus- sion, do not hold water if the overall rate of cell attrition is very slow, as in the ageing process, as a very low incidence of energy deficient/degener- ating cells would be predicted at any one time point.
In this study, mitochondrial function in human diaphragm from subjects of different ages with no known neuromuscular disease was assessed in individual fibres using a histochemical battery. The proportion of fil-,'es with respiratory chain failure at various ,,gcs ~d~ .,~ -o~ed T~,-se obser- vations represent, in part, an effort to confirm the critical observations of Miiller-H~icker. In addi- tion, a parallel is drawn betweer, skeletal muscle and the human neuronal pool. As neurones share with skeletal muscle a high dependence on en- ergy derived through oxidative phosphorylation but differ in having no regenerative capacity in adult life, it is reasonable to predict that they would be particularly vulnerable to mitochondrial failure. The concept is developed that the mito- chondrial pathology of skeletal muscle may pro- vide a window on CNS events in ageing.
Methodology
Subjects with no evidence of a neuromuscular problem were selected for study. Biopsies were taken from the greater erus of the diaphragm on either the right, left or both sides. The patients included 39 patients having routine upper abdom- inal surgery (13 hiatus hernia repair, 16 surgery for gastro-oesophageal reflux and 10 other condi- tions including achalasia and gastric neoplasia) and in 10 eases specimens were obtained at necropsy from patients with sudden death, usu- ally from cardiac causes. A total of 103 specimens were obtained from these 49 patients. Consent was sought from all patients who agreed to give intraoperativ¢ biopsies. Specimens from autopsy cases were obtained from the Victorian Institute of Forensic Pathology following the granting of approval from their Ethics Committee on 22 February 1990.
Tr immed blocks of muscle tissue were quenched in iso-pentane pre-cooled in liquid ni- trogen. A standardised histochemical battery was
127
performed on serial cryostat sections including cytochrome oxidase, NADH tetrazolium reduc- tase, lactate dehydrogenase, and Ca 2 + dependent myofibrillary ATPase, including acid pre-incuba- tion at pH 4.3 and 4.6.
In most cases the percentage of cytochrome oxidase negative fibres was obtained after count- ing 1000 contiguous muscle fibres. Fifteen speci- mens contained less than 1000 fibres. Data were analysed using the Minitab Release 7.1 (1989) statistical program.
Results
Muscle histochemistry Abnormalities were confined to respiratory
complex histochemical probes with no evidence of abnormalities of fibre type. Myosin ATPase staining revealed a normal reaction in cy- tochrome oxidase negative fibres as did NADH- TR, LDH and haematoxylin and eosin indicating that cytochrome oxidase negativity was a selective finding and one not simply attributable to general enzymatic failure as part of cell necrosis.
Percentages of cytochrome oxidase negative fibres are shown in Fig. 1, and the mean results for each decade are shown in Table 2. It can be seen that cytochrome negative fibres are not found in the first three decades, appearing in the fourth decade, with a fairly low prevalence until later in life. There is then a marked, and statisti- cally significant increase in the seventh, eighth and ninth decades when compared with the fourth decade (all p values ~ 0.0028).
4.0
24
0
Fig. 1. Percentages of cytochrome oxidase negative fibres from diaphragm muscles plotted against age of patients in years.
• • 2 •
_..,:,'-. = . P . ~ I | s ' | , 20 50 80
AGE In YEARS
128
TABLE 2
A THEORETICAL CONSTRUCT LINKING PERCENT- AGE OF CYTOCHROME OXIDASE NEGATIVE FIBRES IN DIFFERENT AGE GROUPS AND ANNUAL RATE OF CELL A' iTRITION
Age in Observed % Predicted % years COX negative cell attrition
Extrapolation from mean observed incidence of respiratory deficient fibres in each decade to rate of neuronal attrition. These projections are based on the assumptions that ( i ) a similar rate of accumulation of respiratory deficient cells operates in the CNS neuronal pool as in skeletal muscle, (2) there is a 25~ loss in the neuronal pool with ageing and (3) the prevalence of respiratory deficient fibres is directly linked to the rate of cell attrition,
The percentage of cytochrome oxidase nega- tive fibres increases significantly in an exponen- tial manner with increasing age (r -- 0.908). There was no difference between age-matched patients with various conditions and those with sudden death. The extent of the rise in the later decades fits an exponential curve. Areas of cytochromc oxidase inactivity were demonstrated in all three fibre types, but were most commonly seen in type I fibres. As well as specific COX defects other fibres contained abnormalities of mitochondrial staining including moth-eaten and core-like le- sions, ring fibres and sarcoplasmic masses (Fig. 2) and again these increased with age.
The cytochrome oxidase stain is probably the only histochemical reaction which is a pure respi- ratory chain probe. If one makes the assumption that the time course from onset of energy failure to cell attrition is likely to be similar at different ages, the finding of an increase in the number of cytochrome oxidase negative fibres with ageing suggests that the rate of mitochondrial failure steadily increases with ageing. (The alternative explanation is that cells with mitochondrial fail- ure persist for very long times before dying, lead-
ing to an increased prevalence with a steady state annual incidence.)
Possible theoretical implications for CNS neurones The putative factors which lead to mitochon-
drial failure, namely an accumulation of mtDNA mutations, are likely to affect mitochondria in all tissues at approximately the same rate and the resultant effects on cell metabolism will depend on number of mitochondria/mitochondriai DNA copies per cell, relative dependence on oxidative and glycolytic energy generation and regenerative capacity of the tissue.
The rate of development of mitochondrial fail- ure should be broadly similar in skeletal muscle and neuronal mitochondria and the prevalence of mitochondrial deficient cells in skeletal muscle at different ages gives a rough indication of the extent of neuronal mitochondrial pathology. If one extrapolates from the percentage of muscle cells with respiratory failure observed at different ages to neurones and takes a neuronal fall-out rate of 25% over the last 60 years of life (litera- ture figure) one can calculate approximate annual rates of cell fall-out. This calculation is based on several assumptions including (1)cytochrome oxi- dase negativity is linked to cell attrition and (2) the time cause of energy failure to cell attrition is similar at different ages. While speculative, these are not unlikely assumptions. The very high inci- dence of cytochrom¢ oxidase negative muscle fi- bres in very elderly subjects (mean 2.85% in the ninth decade) if mirrored in the CNS neuronal pool could account for a very high annual neu- ronal attrition related to energy failure.
The theoretical considerations set out in Table 2 indicate that the percentage of cytochrome oxidase negative fibres at any one time is proba- bly greater than the annual rate of cell attrition and that cell failure is very slow ( ~ 700 days).
Discussion
The findings in this study confirm those of an important study by Miiller-H6cker (1990), indicat- ing that respiratory deficient fibres begin to accu- mulate in the fourth decade and there is an exponential increase in their numbers late in life. The observed numbers of such fibres range from
129
i i i •
i
k~
. • ~:?~ ~i~/: I ? . : ~ ~ . . . . ~ :~ ~,~
Fig. 2. (A-C) Sections of diaphragm from a 72-year-old woman, stained with NADH-TR, cytochrome oxidase and ATPase pit 4.3 pre-incubation respectively, original magnification 10x. (D-F) Similarly stained sections from a 70-year-old man. original
magnification 25 x . Type I fibres are indicated by arrows and type 2A by asterisks.
130
0.16% in the fourth decade to 2.85% in the ninth decade. We used these findings in skeletal muscle to develop an argument that a similar occurrence of mitochondrial failure in the human neuronal pool, another post-mitotic cell population, could easily account for the loss of 25% of neurones in the course of ageing. Although the argument is couched largely in terms of cell attrition, it is likely that lesser degrees of energy dysfunction could result in the other well characterised neu- ropathological neuronal changes of ageing, for example, contraction of the dendritic tree and loss of synapses (Scheibel et al., 1975).
The histochemical picture of adult muscle, with most fibres showing normal respiratory activity but with some inert fibres, indicates a wide range in the efficiency of mitochondrial energy genera- tion between different fibres, a situation that should lead to energy deficiency in respiratory deficient fibres unless glycolytically derived en- ergy can sustain cell needs. The argument in this paper that these changes may have a role in human ageing has been primarily directed to- wards the CNS, both because the more serious deleterious effects of ageing that cannot be largely attributed to another discrete process (i.e., atheroma, neoplasia) largely affect the CNS and because the high dependence of neurones on oxidative metabolism and the lack of a regenera- tive capacity makes the neuronal pool a very vulnerable target for mitochondrial failure, While such a projection from skeletal muscle to CNS neurones is somewhat speculative, there is every likelihood that CNS neurones will be at least as susceptible to the effects of mitochondrial failure as skeletal muscle cells.
The question as to whether mitochondrial fail- ure has an effect on skeletal muscle function with ageing also deserves closer attention, Skeletal muscle retains a regenerative capacity, not from the secondary myocyte which is a fixed post- mitotic cell but from the satellite cell population. There is a potential therefore for muscle fibres which fail to be replaced, and the effect of mito- chondrial failure on muscle function is likely to be limited to a loss of power related to the number of cells in energy crisis at any one time. As this and earlier studies have shown, the inci- dence is relatively small in younger patients but
in certain very old patients, where up to 4% of fibres show changes, it may well have some physi- ological sequelae.
The view that an evolving energy failure is central in CNS ageing in man represents an at- tractive new concept for which there is consider- able indirect support. Confirmation of this theory will require both studies on the CNS itself to confirm that a similar pathology develops to that seen in skeletal muscle and further exploration of the time course and interrelationship of energy failure and cell attrition. The ageing process in man is an extremely slow one and it is stressed that pathologic factors that may have a central role may be detectable in only a small proportion of cells in a single time frame. These considera- tions apply both to histochemical studies of respi- ratory chain function (with only a small percent- age of fibres showing cytochrome oxidase negativ- ity) and to the load of deleted mitochondrial genomes (with PCR necessary for detection).
Acknowledgements
We acknowledge Professor C,J, Martin, Dr P, Brooke, Professor S. Cordner and staff of the Victorian Institute of Forensic Pathology for as- sisting with the collection of muscle tissues, Dr R, Kirsner for statistical assistance and Mrs S. Tombs for technical assistance.
References
Brody, H, (1955) Organisation of the cerebral cortex, II1. A study of aging in the human cerebral cortex, L Cutup. Neurol., 102, 511-556,
Byrne, E., X. Dennett and l, Trounce (1991) Oxidative energy failure in fixed post-mitotic cells; a major factor in senes- cence, Rev, NeuroL, 147, 532-535,
Cortopossi, G.A, and N. Arnheim (1990) Detection of a specific mitochondrial DNA deletion in tissues of older humans, Nucleic Acids Res., 18, 6927-6933.
Hall, T.C., A.K.H. Miller and LA.N. Corsellis (1975) Varia- tions in the human Purkinje celt population according to age and sex, Neuropathol. Appl, Neurobiol., 1, 267-292,
Hanley, T. (1974) 'Neuronal fall-out' in the ageing brain: a critical review of the quantitative data, Age Ageing, 3, 133-151.
Linnane, A.W., A. Bauma, R.S. Maxwell, H. Preston, C.S. Zhang and S. Marzuki (1991) Mitochondrial gene muta- tion: the ageing process and degenerative disease, Biochem. Int., 22, 1067-1076.
Miiller-H6cker, J.D. (1989) Cytochrome-c-oxidase deficient cardiomyocytes in the human heart: an age-related phe- nomenon, Am. J. PathoL, 134, 1167-1173.
Miiller-Hficker, .I.D. (1990) Cytochrome c oxidase deficient fibres in the limb muscle and diaphragm of man without muscular disease: an age-related alteration, J. Neurol. Sci., 100, 14-21.
Scheibel, M.E., U. Tomiyasu and A.B. Scheibel (1977) The aging human Betz cell, Exp. Neurol., 56, 598-609.
Tomlinson, B.E. and D. Irving (1977) The numbers of limb
131
motor neurons in the human lumbosacral cord throughout life, J. Neurol. Sci., 34, 213-219.
Torii, K., S. Sugiyama, K. Takagi, T. Satake and T. Ozawa (1992) Age-related increase in respiratory muscle mito- chondrial function in rats, Am. J. Respir. Cell Mol. Biol., 6, 88-92.
Trounce, I., E. Byrne and S. Marzuki (1989) Decline in skeletal muscle respiratory chain function: possible factor in ageing, Lancet, i, 637-639.
