- Email: [email protected]
A search for brain damage in a rat model of alcoholic sleep apnea
EXPERIMENTAL
NEUROLOGY
84,
2 19-224 ( 1984)
A Search for Brain Damage in a Rat Model of Alcoholic Sleep Apnea BRIAN CRAGG AND STEPHEN PHILLIPS Department of Physiology, Monash University, Australia 3168 Received November IO, I983 To determine whether or not brain damage is likely to occur in the repeated apneic episodes experienced by some alcoholics during intoxicated sleep, rats were anesthetized with alcohol and subjected to repeated episodes to asphyxia. Each episode lasted 90 s, arterial PO2 was less than 50 mm Hg for 60 s, and 30 episodes of asphyxia occupied 45 min in the first h of anesthesia. In other rats asphyxic episodes occupied more than half of the first 2 or 3 h of anesthesia, or I h of asphyxic episodes was repeated on 5 successivedays. In none of these rats was there evidence of neuronal damage by electron microscopy, or axonal degeneration by Fink-Heimer staining, or Purkinje cell loss by counting. It was found that even in normal nervous tissue about 3% of mitochondria were in condensed forms suggestive of a degenerative phase in their life cycle.
INTRODUCTION A novel mechanism by which human abuse of alcohol might cause brain damage has come to light in the finding (2) that some patients, after consuming alcohol, experience frequent periods of apnea during sleep. In the first 1 to 2 h of sleep after alcohol intake, the oxygen saturation of arterial blood was reduced in all subjects, and the time spent in apnea was as much as 45 min in the first 60 min of sleep in the most affected subject. The duration of apnea exceeded 1 min in some episodes in some subjects. The mean oxygen saturation of arterial blood was only 52% in the most affected subject, and this estimate understates the deviation because oxygen readings of less than 50% were reported as 50% because of inaccuracy below that value (2). We lack evidence as to whether the resulting biochemical disturbances would result in the irreversible event of the loss of neurons. Experimental work on rats permits control of the blood concentration of alcohol and of asphyxic episodes, and enables the brain tissue to be examined histologically 219 0014-4886/84 $3.00 Copyright 0 1984 by Academic Prcs Inc. All rights of rcpmduction in any form reserved.
220
CRACG AND PHILLIPS
at the optimal survival times for detecting ultrastructural damage by electron microscopy or axonal degeneration by specific silver impregnation (FinkHeimer method), or neuronal loss by counting nerve cells. MATERIALS
AND
METHODS
Male Wistar rats of about 300 g were anesthetized with ethanol by i.p. injection of 10% (w/v) cane spirit (CSR) in saline to a dose of 4 g/kg. The duration of anesthesia was at least 4 h, and all rats recovered and were able to right themselves after 6 h. Asphyxia was induced by applying a polythene cup that fitted the snout and mouth. The inner rim of the cup was heavily covered with petroleum jelly which allowed a very restricted air flow through the fur. The jelly was renewed or moved up from the bottom of the cup to the rim as necessary, to avoid blocking the nares. The volume of air contained in the cup when in place on a rat was 2 ml. The rat continued to rebreath this volume throughout the 90 s of asphyxia. In some rats, a cannula was inserted in the descending aorta, and blood samples taken to measure PO2 and pH on a Coming pHIBlood Gas Meter, model 165/2. Episodes of 90-s asphyxia were separated by 30-s recovery periods of unobstructed breathing, so that in 1 h, 45 min were spent in asphyxia. One day after the asphyxic episodes, the rats had cleaned off the petroleum jelly and recovered normal behavior. After 1 to 28 days, the rats were deeply anesthetized with chloroform and perfused through the left cardiac ventricle with 4% pamformaldehyde in phosphate buffer. Samples of brain tissue for electron microscopy were further fixed in the same fixative to which 1% glutaraldehyde had been added. After more than 1 week of fixation, the remaining brain was sectioned parasagitally at 30 pm on a freezing microtome, and sections stained by a modified Fink-Heimer method (1). For counting Purkinje cells, midsagittal sections of cerebellar vermis cut at 30 pm were stained by rocking overnight in 0.001% toluidine blue plus 0.001% thionin in 0.1 M sodium phosphate buffer at pH 7.3. Sections were then mounted on glass slides from phosphate buffer, dried, and dipped briefly in xylol before coverslipping with DPX. Purkinje cells were very prominent after this treatment, and were counted under a 10X objective using a hand-held counter. The length of the Purkinje cell line was measured on a Zeiss MOP electronic tablet by projection from a Leitz Neopromar microscope at a magnification of X42. For electron microscopy, plates of tissue less than 0.5 mm thick were immersed in 1% osmium tetroxide in phosphate buffer overnight, stained with 2% uranyl acetate solution, dehydrated in alcohol and epoxypropane, and embedded in Araldite. Sections of silver interference color were stained with Reynold’s lead citrate solution and examined in a JEOL 100s microscope. To count abnormal mitochondria, a fixed magnification of 40,000 was used,
ASPHYXIA
221
AND BRAIN DAMAGE
and fields were selected at random by flicking the specimen movement controls. About 60 fields representing 9.1 pm2 each were counted and all unusual structures recorded in each specimen. RESULTS Rats under alcohol anesthesia had a blood oxygen tension of about 90 mm Hg and a pH close to 7.4 (Table 1). Five periods of 90-s asphyxia were administered, each separated by 30-s intervals of free breathing. During a sixth period of asphyxia blood samples were taken after 30 and 90 s. The oxygen tension was less than 50 mm Hg for the last 60 s of the 90-s period, and the pH was shifted toward acidity (Table 1). Rats under alcohol anesthesia were then subjected to periods of 90-s asphyxia interspersed with periods of free breathing to a total of 1 h. In the earlier experiments the free breathing had to be increased beyond 30 s on some occasions to allow the rat to recover spontaneous respiration, and the total time in asphyxia was 31.5 to 34.5 min in 1 h. In later experiments the chest of the rat was manually pumped if respiration ceased, and it was possible to keep the periods of free breathing to 30 s so that 45 min in 1 h was spent in asphyxia. Altogether five rats were studied after 7 days of survival, three after 14 days, and four after 28 days. Those times were chosen for detecting neuronal degeneration, but two rats were also studied after 24-h survival for any acute mitochondrial changes. In other rats the duration of asphyxia was increased to 60 min in 2 h or 101 min in 3 h. In three further rats asphyxia for 30 min in 1 h was repeated under alcohol anesthesia on 5 successive days. These five rats had 7 days survival. The results of all the experiments were uniform and will be described together. Cells and neuropil in the cerebellum, cerebral cortex, and hippocampus were searched for abnormalities by electron microscopy. The great majority of the mitochondria appeared normal, but some condensed and intensely TABLE
1
Oxygen Tension and Acid-Base Balance of Aortic Blood in Rata under Alcohol Anesthesia and during Asphyxia*
Anesthesia Asphyxia, 30 s Asphyxia, 90 s
90.0 * 2.1 N=6 44.1 f 4.7 30.3 + 3.2
u PO2 in mm Hg, R f. SE, N = 4 unless otherwise stated.
7.39 2 .Ol N=S 7.21 + .04 7.14 * .03
CRAGG AND PHILLIPS
FIGS. l-6. Osmiophilic bodies with remnants of mitochondrial structure in neurons of cerebral cortex 7 days (Figs. I, 2), 1 day (Figs. 3,4,6), or 28 days (Fig. 5) after I h of asphyxic episodes under alcohol anesthesia. Bars are 0.5 pm.
osmiophilic forms were found in the neuropil (Figs. l-6). These were distinct from lysosomes found in cell bodies, and often included fragments of double membranes or cristae. However, condensed mitochondria could also be found by searching normal control material, and no peculiar forms were found in the asphyxiated brains that were not present in the controls. Counts were made on brain samples in which abnormal mitochondria appeared prominent, together with control samples from normal rats prepared concurrently. Dark-
ASPHYXIA
223
AND BRAIN DAMAGE
ened mitochondria as in Figs. l-6 were present at about one per 20 pm2 in both control and asphyxiated brain equally. There were about 30 normal mitochondria in the same area in both control and asphyxiated brains. The abnormal forms thus amounted to about 3% of all mitochondria. Pale expanded mitochondria were infrequent, and no ultrastructural features were found that could not be matched in the control material. To search larger areas for degeneration, Araldite sections of 0.5 pm thickness were stained with toluidine blue and scanned under oil immersion, looking particularly for degenerated axons in the cerebral and cerebellar white matter, where the Fink-Heimer method is often difficult to interpret. Degenerated axons, cell bodies, or glial reactions were not present. Parasagittal sections of whole brain from olfactory bulb to brain stem were stained by the FinkHeimer method and scanned, together with normal control sections, and positive control sections that had surgical lesions in the cerebral cortex. The latter gave clear staining of degenerated axons at both the shortest (7 days) and longest (28 days) survival times studied in the asphyxiated brains. There were no convincing signs of axonal degeneration in the asphyxiated brains. To look for a loss in the more severely asphyxiated brains, Purkinje cells in the cerebellar cortex were then counted. The total counts for ail lobes in single parasagittal sections from each brain are shown in Table 2. The count was divided by the length of the Purkinje ccl1 line to obtain the density in cells per millimeter as shown. There was no suggestion of loss due to asphyxia in any of those measurements, and the scanning of the entire Purkinje cell line during counting did not show any evidence of cell debris or glial reactions. DISCUSSION The changes in blood gases (Table 1) are similar to those of the more extreme human cases, who experience the largest deviations in the Crst hour after taking alcohol (2). Increasing the duration of asphyxia to 2 and 3 h did TABLE 2 Total Number and Linear Density of Purkinje Cells in Midline 30-pm Sections of Cerebellum in Control Bats and in Three Groups of Rats Subjected to Asphyxiz’
Controls, N = 4 45 min Asphyxia, in 1st h, 28 days survival, N = 4 Asphyxia repeated on 5 successivedays, 7 days survival, N=3 Asphyxic episodes for 2 to 3 h, 7 days survival, N = 2
Cell number
Cells/mm
1829 f 120 1824 Z!Y104
25.3 + 0.6 26.5 + 0.5
1876 + 9 1870 f 89
25.5 k 0.2 25.8 + 1.6
224
CRAGG
AND
PHILLIPS
not produce damage in our rats. The alcoholic human might experience apneic sleep on more than the 5 successive days experienced by our rats, but it is not clear what biochemical disturbances could accumulate from day to day. The alcohol dose that we administered (4 g/kg) produced surgical anesthesia within 15 min, and this could hardly be exceeded by an alcoholic human. It therefore seems unlikely that apneic sleep after alcohol consumption would produce a loss of neurons unless human nerve cells are more easily damaged than those of the rat. We do not know of evidence that bears on this question. One human brain that had experienced alcoholic apnea in sleep (2) did not show histological evidence of hypoxic encephalopathy (personal communication from Dr. Clive Harper, Royal Perth Hospital). Short of a loss of nerve cells, there are other possibilities of neuronal damage that may be less irreversible. It is possible for dendrites and the synaptic contacts upon them to retract without degeneration and subsequently be reformed (4, 5). We have not seen anything unusual in the ultrastructure of dendrites or synapses, but we have not counted the latter. A nondegenerative retraction of synapses, or a loss of neurotransmitter binding sites, might cause psychometric deficits in alcoholic humans without structural evidence of damage. The condensed mitochondrial forms that we have counted are presumably part of the natural life cycle of mitochondria, but their presence in normal material does not seem to have been noticed or quantified previously (3, 6).
REFERENCES 1. ANKER, R. L., AND B. G. CRACG. 1974. Development of the extrinsic connections of the visual cortex in the cat. J. Comp. Neural. 154: 29-42. 2. ISSA, F. G., AND C. E. SULLIVAN. 1982. Alcohol, snoring and sleep apnoea. .I. Neural. Neurosurg. Psychiatry 45: 353-359. 3. PETERS,A., S. L. PALAY, AND H. DE F. WEBSTER. 1976. The Fine Structure of the Nervous System: The Neurons and Supporting Cells. Saunders, London. 4. SUMNER, B. E. H., AND W. E. WATSON. 197 1. Retraction and expansion of the dendritic tree of motor neurones of adult rats produced in vivo. Nature (London) 233: 273-275. 5. SUMNER, B. E. H. 1975. A quantitative analysis of houtons with different types of synapses in normal and injured hypoglossal nuclei. Exp. Neural. 49: 406-17. 6. VANNESTE, J., AND P. VAN DEN BOSCH DE AGUILAR. 1981. Mitochondrial alterations in the spinal ganglion neurons in ageing rats. Acta Neuropathol. (Berlin) 54: 83-87.
NEUROLOGY
84,
2 19-224 ( 1984)
A Search for Brain Damage in a Rat Model of Alcoholic Sleep Apnea BRIAN CRAGG AND STEPHEN PHILLIPS Department of Physiology, Monash University, Australia 3168 Received November IO, I983 To determine whether or not brain damage is likely to occur in the repeated apneic episodes experienced by some alcoholics during intoxicated sleep, rats were anesthetized with alcohol and subjected to repeated episodes to asphyxia. Each episode lasted 90 s, arterial PO2 was less than 50 mm Hg for 60 s, and 30 episodes of asphyxia occupied 45 min in the first h of anesthesia. In other rats asphyxic episodes occupied more than half of the first 2 or 3 h of anesthesia, or I h of asphyxic episodes was repeated on 5 successivedays. In none of these rats was there evidence of neuronal damage by electron microscopy, or axonal degeneration by Fink-Heimer staining, or Purkinje cell loss by counting. It was found that even in normal nervous tissue about 3% of mitochondria were in condensed forms suggestive of a degenerative phase in their life cycle.
INTRODUCTION A novel mechanism by which human abuse of alcohol might cause brain damage has come to light in the finding (2) that some patients, after consuming alcohol, experience frequent periods of apnea during sleep. In the first 1 to 2 h of sleep after alcohol intake, the oxygen saturation of arterial blood was reduced in all subjects, and the time spent in apnea was as much as 45 min in the first 60 min of sleep in the most affected subject. The duration of apnea exceeded 1 min in some episodes in some subjects. The mean oxygen saturation of arterial blood was only 52% in the most affected subject, and this estimate understates the deviation because oxygen readings of less than 50% were reported as 50% because of inaccuracy below that value (2). We lack evidence as to whether the resulting biochemical disturbances would result in the irreversible event of the loss of neurons. Experimental work on rats permits control of the blood concentration of alcohol and of asphyxic episodes, and enables the brain tissue to be examined histologically 219 0014-4886/84 $3.00 Copyright 0 1984 by Academic Prcs Inc. All rights of rcpmduction in any form reserved.
220
CRACG AND PHILLIPS
at the optimal survival times for detecting ultrastructural damage by electron microscopy or axonal degeneration by specific silver impregnation (FinkHeimer method), or neuronal loss by counting nerve cells. MATERIALS
AND
METHODS
Male Wistar rats of about 300 g were anesthetized with ethanol by i.p. injection of 10% (w/v) cane spirit (CSR) in saline to a dose of 4 g/kg. The duration of anesthesia was at least 4 h, and all rats recovered and were able to right themselves after 6 h. Asphyxia was induced by applying a polythene cup that fitted the snout and mouth. The inner rim of the cup was heavily covered with petroleum jelly which allowed a very restricted air flow through the fur. The jelly was renewed or moved up from the bottom of the cup to the rim as necessary, to avoid blocking the nares. The volume of air contained in the cup when in place on a rat was 2 ml. The rat continued to rebreath this volume throughout the 90 s of asphyxia. In some rats, a cannula was inserted in the descending aorta, and blood samples taken to measure PO2 and pH on a Coming pHIBlood Gas Meter, model 165/2. Episodes of 90-s asphyxia were separated by 30-s recovery periods of unobstructed breathing, so that in 1 h, 45 min were spent in asphyxia. One day after the asphyxic episodes, the rats had cleaned off the petroleum jelly and recovered normal behavior. After 1 to 28 days, the rats were deeply anesthetized with chloroform and perfused through the left cardiac ventricle with 4% pamformaldehyde in phosphate buffer. Samples of brain tissue for electron microscopy were further fixed in the same fixative to which 1% glutaraldehyde had been added. After more than 1 week of fixation, the remaining brain was sectioned parasagitally at 30 pm on a freezing microtome, and sections stained by a modified Fink-Heimer method (1). For counting Purkinje cells, midsagittal sections of cerebellar vermis cut at 30 pm were stained by rocking overnight in 0.001% toluidine blue plus 0.001% thionin in 0.1 M sodium phosphate buffer at pH 7.3. Sections were then mounted on glass slides from phosphate buffer, dried, and dipped briefly in xylol before coverslipping with DPX. Purkinje cells were very prominent after this treatment, and were counted under a 10X objective using a hand-held counter. The length of the Purkinje cell line was measured on a Zeiss MOP electronic tablet by projection from a Leitz Neopromar microscope at a magnification of X42. For electron microscopy, plates of tissue less than 0.5 mm thick were immersed in 1% osmium tetroxide in phosphate buffer overnight, stained with 2% uranyl acetate solution, dehydrated in alcohol and epoxypropane, and embedded in Araldite. Sections of silver interference color were stained with Reynold’s lead citrate solution and examined in a JEOL 100s microscope. To count abnormal mitochondria, a fixed magnification of 40,000 was used,
ASPHYXIA
221
AND BRAIN DAMAGE
and fields were selected at random by flicking the specimen movement controls. About 60 fields representing 9.1 pm2 each were counted and all unusual structures recorded in each specimen. RESULTS Rats under alcohol anesthesia had a blood oxygen tension of about 90 mm Hg and a pH close to 7.4 (Table 1). Five periods of 90-s asphyxia were administered, each separated by 30-s intervals of free breathing. During a sixth period of asphyxia blood samples were taken after 30 and 90 s. The oxygen tension was less than 50 mm Hg for the last 60 s of the 90-s period, and the pH was shifted toward acidity (Table 1). Rats under alcohol anesthesia were then subjected to periods of 90-s asphyxia interspersed with periods of free breathing to a total of 1 h. In the earlier experiments the free breathing had to be increased beyond 30 s on some occasions to allow the rat to recover spontaneous respiration, and the total time in asphyxia was 31.5 to 34.5 min in 1 h. In later experiments the chest of the rat was manually pumped if respiration ceased, and it was possible to keep the periods of free breathing to 30 s so that 45 min in 1 h was spent in asphyxia. Altogether five rats were studied after 7 days of survival, three after 14 days, and four after 28 days. Those times were chosen for detecting neuronal degeneration, but two rats were also studied after 24-h survival for any acute mitochondrial changes. In other rats the duration of asphyxia was increased to 60 min in 2 h or 101 min in 3 h. In three further rats asphyxia for 30 min in 1 h was repeated under alcohol anesthesia on 5 successive days. These five rats had 7 days survival. The results of all the experiments were uniform and will be described together. Cells and neuropil in the cerebellum, cerebral cortex, and hippocampus were searched for abnormalities by electron microscopy. The great majority of the mitochondria appeared normal, but some condensed and intensely TABLE
1
Oxygen Tension and Acid-Base Balance of Aortic Blood in Rata under Alcohol Anesthesia and during Asphyxia*
Anesthesia Asphyxia, 30 s Asphyxia, 90 s
90.0 * 2.1 N=6 44.1 f 4.7 30.3 + 3.2
u PO2 in mm Hg, R f. SE, N = 4 unless otherwise stated.
7.39 2 .Ol N=S 7.21 + .04 7.14 * .03
CRAGG AND PHILLIPS
FIGS. l-6. Osmiophilic bodies with remnants of mitochondrial structure in neurons of cerebral cortex 7 days (Figs. I, 2), 1 day (Figs. 3,4,6), or 28 days (Fig. 5) after I h of asphyxic episodes under alcohol anesthesia. Bars are 0.5 pm.
osmiophilic forms were found in the neuropil (Figs. l-6). These were distinct from lysosomes found in cell bodies, and often included fragments of double membranes or cristae. However, condensed mitochondria could also be found by searching normal control material, and no peculiar forms were found in the asphyxiated brains that were not present in the controls. Counts were made on brain samples in which abnormal mitochondria appeared prominent, together with control samples from normal rats prepared concurrently. Dark-
ASPHYXIA
223
AND BRAIN DAMAGE
ened mitochondria as in Figs. l-6 were present at about one per 20 pm2 in both control and asphyxiated brain equally. There were about 30 normal mitochondria in the same area in both control and asphyxiated brains. The abnormal forms thus amounted to about 3% of all mitochondria. Pale expanded mitochondria were infrequent, and no ultrastructural features were found that could not be matched in the control material. To search larger areas for degeneration, Araldite sections of 0.5 pm thickness were stained with toluidine blue and scanned under oil immersion, looking particularly for degenerated axons in the cerebral and cerebellar white matter, where the Fink-Heimer method is often difficult to interpret. Degenerated axons, cell bodies, or glial reactions were not present. Parasagittal sections of whole brain from olfactory bulb to brain stem were stained by the FinkHeimer method and scanned, together with normal control sections, and positive control sections that had surgical lesions in the cerebral cortex. The latter gave clear staining of degenerated axons at both the shortest (7 days) and longest (28 days) survival times studied in the asphyxiated brains. There were no convincing signs of axonal degeneration in the asphyxiated brains. To look for a loss in the more severely asphyxiated brains, Purkinje cells in the cerebellar cortex were then counted. The total counts for ail lobes in single parasagittal sections from each brain are shown in Table 2. The count was divided by the length of the Purkinje ccl1 line to obtain the density in cells per millimeter as shown. There was no suggestion of loss due to asphyxia in any of those measurements, and the scanning of the entire Purkinje cell line during counting did not show any evidence of cell debris or glial reactions. DISCUSSION The changes in blood gases (Table 1) are similar to those of the more extreme human cases, who experience the largest deviations in the Crst hour after taking alcohol (2). Increasing the duration of asphyxia to 2 and 3 h did TABLE 2 Total Number and Linear Density of Purkinje Cells in Midline 30-pm Sections of Cerebellum in Control Bats and in Three Groups of Rats Subjected to Asphyxiz’
Controls, N = 4 45 min Asphyxia, in 1st h, 28 days survival, N = 4 Asphyxia repeated on 5 successivedays, 7 days survival, N=3 Asphyxic episodes for 2 to 3 h, 7 days survival, N = 2
Cell number
Cells/mm
1829 f 120 1824 Z!Y104
25.3 + 0.6 26.5 + 0.5
1876 + 9 1870 f 89
25.5 k 0.2 25.8 + 1.6
224
CRAGG
AND
PHILLIPS
not produce damage in our rats. The alcoholic human might experience apneic sleep on more than the 5 successive days experienced by our rats, but it is not clear what biochemical disturbances could accumulate from day to day. The alcohol dose that we administered (4 g/kg) produced surgical anesthesia within 15 min, and this could hardly be exceeded by an alcoholic human. It therefore seems unlikely that apneic sleep after alcohol consumption would produce a loss of neurons unless human nerve cells are more easily damaged than those of the rat. We do not know of evidence that bears on this question. One human brain that had experienced alcoholic apnea in sleep (2) did not show histological evidence of hypoxic encephalopathy (personal communication from Dr. Clive Harper, Royal Perth Hospital). Short of a loss of nerve cells, there are other possibilities of neuronal damage that may be less irreversible. It is possible for dendrites and the synaptic contacts upon them to retract without degeneration and subsequently be reformed (4, 5). We have not seen anything unusual in the ultrastructure of dendrites or synapses, but we have not counted the latter. A nondegenerative retraction of synapses, or a loss of neurotransmitter binding sites, might cause psychometric deficits in alcoholic humans without structural evidence of damage. The condensed mitochondrial forms that we have counted are presumably part of the natural life cycle of mitochondria, but their presence in normal material does not seem to have been noticed or quantified previously (3, 6).
REFERENCES 1. ANKER, R. L., AND B. G. CRACG. 1974. Development of the extrinsic connections of the visual cortex in the cat. J. Comp. Neural. 154: 29-42. 2. ISSA, F. G., AND C. E. SULLIVAN. 1982. Alcohol, snoring and sleep apnoea. .I. Neural. Neurosurg. Psychiatry 45: 353-359. 3. PETERS,A., S. L. PALAY, AND H. DE F. WEBSTER. 1976. The Fine Structure of the Nervous System: The Neurons and Supporting Cells. Saunders, London. 4. SUMNER, B. E. H., AND W. E. WATSON. 197 1. Retraction and expansion of the dendritic tree of motor neurones of adult rats produced in vivo. Nature (London) 233: 273-275. 5. SUMNER, B. E. H. 1975. A quantitative analysis of houtons with different types of synapses in normal and injured hypoglossal nuclei. Exp. Neural. 49: 406-17. 6. VANNESTE, J., AND P. VAN DEN BOSCH DE AGUILAR. 1981. Mitochondrial alterations in the spinal ganglion neurons in ageing rats. Acta Neuropathol. (Berlin) 54: 83-87.