- Email: [email protected]
Detrimental cerebrometabolic effects of hyperoxia in newborn rats
Neurochem. Int. Vol. 10, No. 3, pp. 355-359, 1987 Printed in Great Britain. All rights reserved
0197-0186/87 $3.00 + 0.00 © 1987 Pergamon Journals Ltd
DETRIMENTAL CEREBROMETABOLIC EFFECTS OF HYPEROXIA IN NEWBORN RATS EDWIN M. NEMOTO,*~f MARINA R. LIN,t MAMDOUHA AHDAB=BARMADA,~ JOHN MOOSSY~§ and PETER M. WINTER~" The 1"Departments of Anesthesiology and Critical Care Medicine, and J~Pathology (Neuropathology) and §Neurology, University of Pittsburgh School of Medicine, Pittsburgh, PA 15261, U.S.A. (Received 11 August 1986; accepted 29 October 1986)
Abstract--To investigate the pathogenesis of oxygen toxicity in the newborn brain, we exposed one-day-old Spragne-Dawley albino rats to 100% 02 and measured whole-brain high-energy phosphates, glucose, lactate, and free fatty acids (FFA) after 0, 15, 30, 60 and 120min. Whole-brain adenosine triphosphate and creatine phosphate fell significantly from about 4.5 to 2.5/*mol-mg -l protein. Brain lactate remained at about 0.3/~mol-mg -l protein in hyperoxic rats, but increased in normoxic rats, from 0.3 to 1.3/~mol- mg - l protein at 120 min. Total FFA decreased from 30 to 15 nmol. mg ~ ~protein during normoxia, but increased to 40 nmol.mg -~ protein during hyperoxia. Undetectable in normoxic rats, arachidonic acid increased to between 4 and 6 nmol.mg-~ protein during hyperoxia while oleic acid increased by two to threefold. In normoxia, palmitate decreased by 700 from 12 to 4 nmol.mg-~ protein whereas in hyperoxia it remained at 10nmol-mg -I protein. Normobaric 100% 02 has detrimental metabolic effects on the neonatal brain which cannot be attributed to cerebral vasospasm or seizureinduced cerebral anoxia because lactic acidosis was not observed. FFA changes suggest that a likely explanation is membrane lipid peroxidation from O2-induced free radicals.
In the newborn, the brain is highly sensitive to the toxic effects of hyperoxia (Haugaard, 1964). Yet, inadvertent hyperoxia with PaO2 exceeding 100 torr may occur in premature infants during general anesthesia or supplemental O2 therapy for respiratory distress syndrome. In a retrospective review of neonatal autopsies, A h d a b - B a r m a d a et al. (1980) found that all 37 patients in w h o m PaO2 had been greater than 150 torr for three hours had pontosubicular necrosis (PSN) at autopsy, whereas 27 who had not been hyperoxic did not have PSN. PSN is an abnormal pattern of selective neuronal necrosis that occurs primarily in the pontine base neurons and the subiculum of the hippocampus (Friede, 1972). Histopathological features of PSN differ from those described with hypoxia or ischemia in the newborn. Thus, there appears to be a strong correlation between PSN lesions in the central nervous system and the toxic effects of hyperoxia in the newborn.
Our aim in the present study was to produce in the rat central nervous system lesions similar to those described in PSN and to study the associated biochemical effects of hyperoxia. The rat was chosen as a model because the rapid maturation schedule o f the newborn rat brain is similar to that of the developing human brain (Donaldson et al., 1931; Dobbing and Smart, 1974). EXPERIMENTAL PROCEDURES
*Address all correspondence to: Edwin M. Nemoto, Ph.D., University of Pittsburgh School of Medicine, 1081 Scaife Hall, Pittsburgh, PA 15261, U.S.A. Tel: (412) 648-9869. Supported in part by the United Way, HRSF Grant No. X-31 (MAB) and the American Heart Association, Dallas, Texas, Grant No. 84-1138 (EMN). 355
Timed, pregnant Spragne-Dawley rats were received three to four days before term. Anywhere between 4 and 16 h after birth, the newborn rats were randomly assigned to the control or experimental group, weighed, numbered, and placed into one of two 5.3 1glass jars containing an inch of sawdust. For the first 45 rain, both jars were insuffiated with humidified room air at 61. mln- l Thereafter, one jar was insufflated with humidified air (control group) and the other with humidified 100% oxygen (experimental group) both at 6 l.mln -l. After 0, 15, 30, 60 and 120 rain, the rats were removed from the jars and decapitated so that their heads fell directly into liquid nitrogen. The frozen brains were removed from the calvaria at - 1 5 to -20°C in a modified cryostat microtome chamber and stored at -95°C until biochemical analysis. For determination of whole brain adenosine triphosphate (ATP), creatine phosphate (CP), glucose and lactate levels, the brains were ground under liquid nitrogen and deproteinized with 6% perchloric acid. Brain ATP, Cp, glucose and
356
EDWIN M. NEMOTOet al.
lactate were determined spetrofluorometrically (Farrand Optical Mark II) using an enzymecycling technique (Lowry et al., 1972). Whole brain free fatty acids (FFA) were separated by thin-layer chromatography after chloroform/ methanol extraction, and quantitated by gas-liquid chromatography (Varian model 3700) as previously described (Shiu et al., 1983). Brain tissue protein was measured by Comassie blue binding with serum albumin standards in the perchloric acid-precipitated pellet of the samples for ATP, CP, glucose and lactate analyses and in the extracted sediment of the chloroform/methanol-extracted samples. All brain metabolite values are expressed per mg protein. Statistical analysis of the data was by two-way analysis of variance. The Student-Newman-Keuls test was used to test for significant differences between group means (unpaired analysis), with a maximum significant P value of 0.05.
lower after 120 min of hyperoxia compared with the normoxic group. Thus, during hyperoxia, the decline in brain ATP was not accompanied by brain lactic acidosis. Brain F F A differed markedly between the hyperoxic and the normoxic groups (Fig. 2). Arachidonic acid was not detectable (ND) in the normoxic rat brain, but ranged between 4 and 6 nmol.mg 1 protein during hyperoxia. Oleic acid, unchanged in normoxia, was significantly increased after 15 min or more of hyperoxia. Brain palmitate levels were reduced after 30 min or more of normoxia, and unchanged during hyperoxia. In the normoxic group total FFA progressively decreased between 30 and 120 min, whereas in the hyperoxic group it increased. Thus, brain FFA levels after 15 min of hyperoxia were consistently higher compared with normoxia levels. Figure 3 shows the individual FFA as percent of total FFA. After 30 min or more of normoxia, the percentage of palmitate decreased, stearate increased, oleate did not change, and arachidonate was not detectable. In contrast, during hyperoxia the contribution of both palmitate and stearate to the FFA pool decreased. After 15 min or more of hyperoxia, arachidonic acid levels represented between 10% and 15% of total FFA. The contribution of oleate also
RESULTS
After 60 and 120 min of normobaric 100% 02, whole-brain ATP was significantly reduced compared with values in the normoxic control group (Fig. 1); at 15 and 60 min it was lower than the 0-min control value. CP was similar between the normoxic and hyperoxic groups, except for a decrease at 60 and 120 min in the hyperoxic group compared with the 0-min control value. Brain glucose was unaffected by hyperoxia. Brain lactate level was elevated after 30min or more of normoxia. It was significantly
~] 7-
NORMOXIA HYPEROXIA = SD
0.S
C.8
0.7 ~
~,o.~[
0.6 0.5 0.4
03tll
0.3 tu 0.20 O
o.,FN
0g
0Ltl-I 0
15
30 60 MINUTES
120
0
15
30 60 MINUTES
120
1.4 ~" 1.2 1.0 0.8 0.6 0.4 ,~ 0.2~ 0-J
Fig. 1. Whole brain adenosine triphosphate (ATP), creatine phosphate (CP), glucose, and lactate in newborn rats after 45 min in room air followed by 120min in either room air or 100% 02 in newborn rats. Number of rats studied indicated on each bar. *P < 0.05 compared with corresponding normoxia value. ÷ P < 0.05 compared with 0-min control value.
Oxygen toxicity in newborn brain
357 NORMOXIA
HYPEROXIA
"
t21
ND = NOT DETECTABLE - 3 0
6
~ - = SD
o +*
_20 -°''
.
o£
< 0
0
15
15 a'E"
oE
lo _oE
10
LU o
s
0
15
30 60 MINUTES
120
0
15
30 60 MINUTES
120
+*
50 O.
~,4o
F-
0
15
30 60 MINUTES
120
Fig. 2. Whole brain free fatty acids (FFA) in newborn rats after 45 min in room air followed by 120 min in either room air or 100% 02. Number of rats studied indicated on each bar. *P < 0.05 vs normoxia value. + P < 0.05 vs 0-min value.
increased, from about 25% of total F F A at 0-min to 36% after 2 h of 100% 02. Thus, during hyperoxic exposure, the proportion of saturated F F A tended to decrease and that of unsaturated FFA, to increase in the total F F A pool. DISCUSSION
Exposure of one-day old rats to hyperoxia causes a significant reduction in brain tissue high-energy phosphate and an increase in F F A levels. The mechanisms of these changes are unknown, but three possible explanations, in order of decreasing likelihood, are: (1) Membrane lipid peroxidation; (2) inhibition of gamma-aminobutryric acid (GABA) and cerebral metabolic activation; and (3) Ozinduced cerebrovascular constriction and cerebral ischemia-anoxia. Our results do not support the notion of an intense,
O2-induced cerebral vasoconstriction resulting in cerebral ischemia-anoxia, because brain lactic acidosis did not accompany the reduction in ATP and CP. Brain lactate was lower in the hyperoxic compared with the normoxic rats namely, 0.4 vs 1.3 #mol. mg -~ protein at 120 min, respectively. Normal adult rat brain lactate values range between 1.2 and 1.7#mol.g -~ brain or 0.4 to 0.6/~mol-mg -~ protein, based upon 3 mg protein, g - ~brain (Veech et al., 1973). However, Vanucci et aL, (1978) reported lactate values of 0.87_+ 11 #mol.g -~ brain or 0.29 #mols.mg -~ protein in 24h postnatal rats. The reason for the marked differences in brain lactate concentration between the normoxic and hyperoxic rats in our study is unknown and cannot be resolved with the data obtained. However, plasma lactate levels in rats one day before birth and up until birth, are about 10 to 12 mM and on room air, decreases to 1 to 2 mM within 2 h postnatal (Medina,
358
EDWIN M. NEMOTOet al.
r/ ~~ =
PALMITATE (16:0) = ARACHIDONATE(20:4)
NORMOXIA
~ ' ~ = OLEATE ~= STEARATE
(18:1) (t :0)
ND = NOTDETECTABLE
8+
~o
0
15
30 60 MIN EXPOSURE
120
Fig. 3. Whole brain free fatty acids (FFA) as percent of total FFA in newborn rats after 45 min in room air followed by 120min in either room air or 100% 02. Number of rats studied at each time of air or 02 exposure is in parentheses above bars. *P < 0.05 vs normoxia value. +P < 0.05 vs 0-min value.
1985). On 100% 02, plasma lactate falls more rapidly and within one hour postnatal, decreases to about 2 raM. Because the permeability of the neonatal rat blood brain barrier to lactate is much greater than that of the adult rat (Hellman et al., 1982) and lactate is easily oxidized as a fuel substrate (Hellman et al., 1982; Arizmendi et al., 1983; Medina, 1985) the changes we observed in brain lactate may be a reflection of the changes occurring in plasma lactate levels. Although there is no evidence that 100% 02 at normal pressure induces seizures in the newborn, in a speculative vein, some of the metabolic changes we observed are similar to those occurring in the postictal state following bicuculline-induced seizures (Blennow et al., 1979). Visual observations of these rats by Ahdah-Barmada et al. (1986) suggest hyperactivity early after exposure to 100% 02 followed by somnolence.
The levels of whole-brain F F A we observed in the newborn rat were about one-third the levels reported by Rodriquez de Turco and Bazan (1983) in the newborn mouse. However, they also reported significant changes in brain F F A in the first few hours of life. Within 60 min after birth, palmitate decreased by 25%, stearate by 10%, oleate by 50% and total F F A by 15%. Over a period of two hours, we found a 60% decrease in palmitate, a 26% decrease in stearate, a 50% decrease in oleate and a 50% decrease in total FFA. The reason for these changes in F F A are unknown. Hyperoxia reduced the proportion in the total F F A of palmitate and stearate, the saturated FFA, and increased arachidonic and oleic acids. This relative increase in arachidonic and oleic acids may indicate a differential activation of phospholipase A2 relative to phospholipase A~. By a process of elimination and based upon the observed increases in F F A during hyperoxia, we speculate that the cerebrometabolic effects are likely attributable to lipid peroxidation by oxygen free radicals such as the superoxide anion. The potential involvement of oxygen-induced free radical damage in the vulnerability of the newborn brain to hyperoxia is supported by the observation that superoxide dismutase (SOD) level in the newborn brain is onethird to one-fourth the level in the 2-month-old rat (Mavelli et al., 1978). Similarly SOD levels in the liver appears to increase rapidly (Utsumi et al., 1977), and the rate of oxygen-induced free radical lipid peroxidation is increased in the newborn compared with adult. Acknowledgements--The authors gratefully acknowledge
the comments and suggestions of Drs Etsuro Motoyama and Robert Guthrie, the editorial assistance of Ms Lisa Cohn, and the secretarial assistance of Ms Betty Jacobson and Ms Barb Burgrnan. REFERENCES
Abdab-Barmada M., Moossy J., Nemoto E. M. and Lin M. R. (1986) Hyperoxia produces neuronal necrosis in the rat. J. Neuropathol. exp. NeuroL 45, 233-246. Ahdab-Barmada M., Moossy J. and Painter M. 0980) Pontosubicular necrosis and hyperoxemia. Pediatrics 66, 840-847. Blennow G., Fobergrova J., Nilsson B. and Siesjo B. K. (1979) Effects of bicuculline-induced seizures on cerebral metabolism and circulation of rats rendered hypoglycemic by starvation. Ann. Neurol. 5, 139-151. Arizmendi C. and Medina J. M. (1983) Lactate as an oxidizable substrate for rat brain in vitro during the perinatal period. Biochem. J. 214, 633~535. Dobbing J. and Smart J. L. (1974) Vulnerability of Developing brain and behaviour. Br. Med. Bull. 30, 164-168.
Oxygen toxicity in newborn brain Donaldson H. H. and Hatai S. (1931) On the weight of the parts of the brain and on the percentage of water in them according to brain weight and to age, in albino and wild Norway rats. J. Comp. Neurol. 53, 263-307. Friede R. L. (1972) Ponto-subicular lesions in perinatal anoxia. Arch. Path. Labs. Med. 94, 343-354. Haugaard N. (1964-65) Poisoning of cellular reactions by oxygen. Ann. N.Y. Acad. Sci. 117, 736-744. Hellman J., Vannucci R. C. and Nardis E. E. (1982) Blood-brain barrier permeability to lactic acid in the newborn dog: lactate as a cerebral metabolic fuel. Pediatr. Res. 16, 40~t4. James L. S. and Lanman J. T., eds (1976) History of oxygen therapy and retrolental fibroplasia. Pediatrics 57, 591-598. Johnson P. C. and Bean J. W. (1957) Effect of sympathetic blocking agents on the toxic action o f O 2 at high pressure. Am. J. Physiol. 188, 593-598. Kennedy C., Grave G. D. and Jehle J. W. (1971) Effect of hyperoxia on the cerebral circulation of the newborn puppy. Pediatr. Res. 5, 659-667. Lowry O. H. and Passonneau J. V. (1972) A Flexible System of Enzymatic Analysis, pp. 152-155. Academic Press, New York. Mavelli I., Mondovi B., Federico R. and Rotilio G. (1978) Superoxide dismutase activity in developing rat brain. J. Neurochem. 31, 363-364.
359
Medina J. M. (1985) The role of lactate as an energy substrate for the brain during the early neonatal period. Biol. Neonate 48, 237-244. Rodriguez de Turco E. B. and Bazan N. G. (1983) Changes in free fatty acids and diglycerides in mouse brain at birth and during anoxia. J. Neurochem. 41, 794--800. Shiu G. K., Nemmer J. P. and Nemoto E. M. (1983) Reassessment of brain free fatty acid liberation during global ischemia and its attenuation by barbiturate anesthesia. J. Neurochem. 40, 880-884. Torbati D. and Lambertsen C. J. (1983) Regional cerebral metabolic rate for glucose during hyperbolic oxygeninduced convulsions. Brain Res. 279, 382-386. Torbati D., Lambertsen C. J. and Greenberg J. (1984) Regional cerebral glucose utilization rates in rats during asymptomatic period of exposures to 1, 2 and 3 atmospheres absolute of oxygen. Neuroscience 11, 947-950. Utsumi K., Yoshioka T., Yamanaka N. and Nakazawa T. (1977) Increase in superoxide dismutase activity concomitant with a decrease in lipid peroxidation of liver during post partum development. FEBS Lett. 79, 1 3. Vannucci R. C. and Vannucci S. J. (1978) Cerebral carbohydrate metabolism during hypoglycemia and anoxia in newborn rats. Ann. Neurol. 4, 73-79. Veech R. L., Harris R. L., Veloso D. and Veech E. H. (1973) Freeze-blowing: a new technique for the study of brain in vivo. J. Neurochem. 20, 183-188.
0197-0186/87 $3.00 + 0.00 © 1987 Pergamon Journals Ltd
DETRIMENTAL CEREBROMETABOLIC EFFECTS OF HYPEROXIA IN NEWBORN RATS EDWIN M. NEMOTO,*~f MARINA R. LIN,t MAMDOUHA AHDAB=BARMADA,~ JOHN MOOSSY~§ and PETER M. WINTER~" The 1"Departments of Anesthesiology and Critical Care Medicine, and J~Pathology (Neuropathology) and §Neurology, University of Pittsburgh School of Medicine, Pittsburgh, PA 15261, U.S.A. (Received 11 August 1986; accepted 29 October 1986)
Abstract--To investigate the pathogenesis of oxygen toxicity in the newborn brain, we exposed one-day-old Spragne-Dawley albino rats to 100% 02 and measured whole-brain high-energy phosphates, glucose, lactate, and free fatty acids (FFA) after 0, 15, 30, 60 and 120min. Whole-brain adenosine triphosphate and creatine phosphate fell significantly from about 4.5 to 2.5/*mol-mg -l protein. Brain lactate remained at about 0.3/~mol-mg -l protein in hyperoxic rats, but increased in normoxic rats, from 0.3 to 1.3/~mol- mg - l protein at 120 min. Total FFA decreased from 30 to 15 nmol. mg ~ ~protein during normoxia, but increased to 40 nmol.mg -~ protein during hyperoxia. Undetectable in normoxic rats, arachidonic acid increased to between 4 and 6 nmol.mg-~ protein during hyperoxia while oleic acid increased by two to threefold. In normoxia, palmitate decreased by 700 from 12 to 4 nmol.mg-~ protein whereas in hyperoxia it remained at 10nmol-mg -I protein. Normobaric 100% 02 has detrimental metabolic effects on the neonatal brain which cannot be attributed to cerebral vasospasm or seizureinduced cerebral anoxia because lactic acidosis was not observed. FFA changes suggest that a likely explanation is membrane lipid peroxidation from O2-induced free radicals.
In the newborn, the brain is highly sensitive to the toxic effects of hyperoxia (Haugaard, 1964). Yet, inadvertent hyperoxia with PaO2 exceeding 100 torr may occur in premature infants during general anesthesia or supplemental O2 therapy for respiratory distress syndrome. In a retrospective review of neonatal autopsies, A h d a b - B a r m a d a et al. (1980) found that all 37 patients in w h o m PaO2 had been greater than 150 torr for three hours had pontosubicular necrosis (PSN) at autopsy, whereas 27 who had not been hyperoxic did not have PSN. PSN is an abnormal pattern of selective neuronal necrosis that occurs primarily in the pontine base neurons and the subiculum of the hippocampus (Friede, 1972). Histopathological features of PSN differ from those described with hypoxia or ischemia in the newborn. Thus, there appears to be a strong correlation between PSN lesions in the central nervous system and the toxic effects of hyperoxia in the newborn.
Our aim in the present study was to produce in the rat central nervous system lesions similar to those described in PSN and to study the associated biochemical effects of hyperoxia. The rat was chosen as a model because the rapid maturation schedule o f the newborn rat brain is similar to that of the developing human brain (Donaldson et al., 1931; Dobbing and Smart, 1974). EXPERIMENTAL PROCEDURES
*Address all correspondence to: Edwin M. Nemoto, Ph.D., University of Pittsburgh School of Medicine, 1081 Scaife Hall, Pittsburgh, PA 15261, U.S.A. Tel: (412) 648-9869. Supported in part by the United Way, HRSF Grant No. X-31 (MAB) and the American Heart Association, Dallas, Texas, Grant No. 84-1138 (EMN). 355
Timed, pregnant Spragne-Dawley rats were received three to four days before term. Anywhere between 4 and 16 h after birth, the newborn rats were randomly assigned to the control or experimental group, weighed, numbered, and placed into one of two 5.3 1glass jars containing an inch of sawdust. For the first 45 rain, both jars were insuffiated with humidified room air at 61. mln- l Thereafter, one jar was insufflated with humidified air (control group) and the other with humidified 100% oxygen (experimental group) both at 6 l.mln -l. After 0, 15, 30, 60 and 120 rain, the rats were removed from the jars and decapitated so that their heads fell directly into liquid nitrogen. The frozen brains were removed from the calvaria at - 1 5 to -20°C in a modified cryostat microtome chamber and stored at -95°C until biochemical analysis. For determination of whole brain adenosine triphosphate (ATP), creatine phosphate (CP), glucose and lactate levels, the brains were ground under liquid nitrogen and deproteinized with 6% perchloric acid. Brain ATP, Cp, glucose and
356
EDWIN M. NEMOTOet al.
lactate were determined spetrofluorometrically (Farrand Optical Mark II) using an enzymecycling technique (Lowry et al., 1972). Whole brain free fatty acids (FFA) were separated by thin-layer chromatography after chloroform/ methanol extraction, and quantitated by gas-liquid chromatography (Varian model 3700) as previously described (Shiu et al., 1983). Brain tissue protein was measured by Comassie blue binding with serum albumin standards in the perchloric acid-precipitated pellet of the samples for ATP, CP, glucose and lactate analyses and in the extracted sediment of the chloroform/methanol-extracted samples. All brain metabolite values are expressed per mg protein. Statistical analysis of the data was by two-way analysis of variance. The Student-Newman-Keuls test was used to test for significant differences between group means (unpaired analysis), with a maximum significant P value of 0.05.
lower after 120 min of hyperoxia compared with the normoxic group. Thus, during hyperoxia, the decline in brain ATP was not accompanied by brain lactic acidosis. Brain F F A differed markedly between the hyperoxic and the normoxic groups (Fig. 2). Arachidonic acid was not detectable (ND) in the normoxic rat brain, but ranged between 4 and 6 nmol.mg 1 protein during hyperoxia. Oleic acid, unchanged in normoxia, was significantly increased after 15 min or more of hyperoxia. Brain palmitate levels were reduced after 30 min or more of normoxia, and unchanged during hyperoxia. In the normoxic group total FFA progressively decreased between 30 and 120 min, whereas in the hyperoxic group it increased. Thus, brain FFA levels after 15 min of hyperoxia were consistently higher compared with normoxia levels. Figure 3 shows the individual FFA as percent of total FFA. After 30 min or more of normoxia, the percentage of palmitate decreased, stearate increased, oleate did not change, and arachidonate was not detectable. In contrast, during hyperoxia the contribution of both palmitate and stearate to the FFA pool decreased. After 15 min or more of hyperoxia, arachidonic acid levels represented between 10% and 15% of total FFA. The contribution of oleate also
RESULTS
After 60 and 120 min of normobaric 100% 02, whole-brain ATP was significantly reduced compared with values in the normoxic control group (Fig. 1); at 15 and 60 min it was lower than the 0-min control value. CP was similar between the normoxic and hyperoxic groups, except for a decrease at 60 and 120 min in the hyperoxic group compared with the 0-min control value. Brain glucose was unaffected by hyperoxia. Brain lactate level was elevated after 30min or more of normoxia. It was significantly
~] 7-
NORMOXIA HYPEROXIA = SD
0.S
C.8
0.7 ~
~,o.~[
0.6 0.5 0.4
03tll
0.3 tu 0.20 O
o.,FN
0g
0Ltl-I 0
15
30 60 MINUTES
120
0
15
30 60 MINUTES
120
1.4 ~" 1.2 1.0 0.8 0.6 0.4 ,~ 0.2~ 0-J
Fig. 1. Whole brain adenosine triphosphate (ATP), creatine phosphate (CP), glucose, and lactate in newborn rats after 45 min in room air followed by 120min in either room air or 100% 02 in newborn rats. Number of rats studied indicated on each bar. *P < 0.05 compared with corresponding normoxia value. ÷ P < 0.05 compared with 0-min control value.
Oxygen toxicity in newborn brain
357 NORMOXIA
HYPEROXIA
"
t21
ND = NOT DETECTABLE - 3 0
6
~ - = SD
o +*
_20 -°''
.
o£
< 0
0
15
15 a'E"
oE
lo _oE
10
LU o
s
0
15
30 60 MINUTES
120
0
15
30 60 MINUTES
120
+*
50 O.
~,4o
F-
0
15
30 60 MINUTES
120
Fig. 2. Whole brain free fatty acids (FFA) in newborn rats after 45 min in room air followed by 120 min in either room air or 100% 02. Number of rats studied indicated on each bar. *P < 0.05 vs normoxia value. + P < 0.05 vs 0-min value.
increased, from about 25% of total F F A at 0-min to 36% after 2 h of 100% 02. Thus, during hyperoxic exposure, the proportion of saturated F F A tended to decrease and that of unsaturated FFA, to increase in the total F F A pool. DISCUSSION
Exposure of one-day old rats to hyperoxia causes a significant reduction in brain tissue high-energy phosphate and an increase in F F A levels. The mechanisms of these changes are unknown, but three possible explanations, in order of decreasing likelihood, are: (1) Membrane lipid peroxidation; (2) inhibition of gamma-aminobutryric acid (GABA) and cerebral metabolic activation; and (3) Ozinduced cerebrovascular constriction and cerebral ischemia-anoxia. Our results do not support the notion of an intense,
O2-induced cerebral vasoconstriction resulting in cerebral ischemia-anoxia, because brain lactic acidosis did not accompany the reduction in ATP and CP. Brain lactate was lower in the hyperoxic compared with the normoxic rats namely, 0.4 vs 1.3 #mol. mg -~ protein at 120 min, respectively. Normal adult rat brain lactate values range between 1.2 and 1.7#mol.g -~ brain or 0.4 to 0.6/~mol-mg -~ protein, based upon 3 mg protein, g - ~brain (Veech et al., 1973). However, Vanucci et aL, (1978) reported lactate values of 0.87_+ 11 #mol.g -~ brain or 0.29 #mols.mg -~ protein in 24h postnatal rats. The reason for the marked differences in brain lactate concentration between the normoxic and hyperoxic rats in our study is unknown and cannot be resolved with the data obtained. However, plasma lactate levels in rats one day before birth and up until birth, are about 10 to 12 mM and on room air, decreases to 1 to 2 mM within 2 h postnatal (Medina,
358
EDWIN M. NEMOTOet al.
r/ ~~ =
PALMITATE (16:0) = ARACHIDONATE(20:4)
NORMOXIA
~ ' ~ = OLEATE ~= STEARATE
(18:1) (t :0)
ND = NOTDETECTABLE
8+
~o
0
15
30 60 MIN EXPOSURE
120
Fig. 3. Whole brain free fatty acids (FFA) as percent of total FFA in newborn rats after 45 min in room air followed by 120min in either room air or 100% 02. Number of rats studied at each time of air or 02 exposure is in parentheses above bars. *P < 0.05 vs normoxia value. +P < 0.05 vs 0-min value.
1985). On 100% 02, plasma lactate falls more rapidly and within one hour postnatal, decreases to about 2 raM. Because the permeability of the neonatal rat blood brain barrier to lactate is much greater than that of the adult rat (Hellman et al., 1982) and lactate is easily oxidized as a fuel substrate (Hellman et al., 1982; Arizmendi et al., 1983; Medina, 1985) the changes we observed in brain lactate may be a reflection of the changes occurring in plasma lactate levels. Although there is no evidence that 100% 02 at normal pressure induces seizures in the newborn, in a speculative vein, some of the metabolic changes we observed are similar to those occurring in the postictal state following bicuculline-induced seizures (Blennow et al., 1979). Visual observations of these rats by Ahdah-Barmada et al. (1986) suggest hyperactivity early after exposure to 100% 02 followed by somnolence.
The levels of whole-brain F F A we observed in the newborn rat were about one-third the levels reported by Rodriquez de Turco and Bazan (1983) in the newborn mouse. However, they also reported significant changes in brain F F A in the first few hours of life. Within 60 min after birth, palmitate decreased by 25%, stearate by 10%, oleate by 50% and total F F A by 15%. Over a period of two hours, we found a 60% decrease in palmitate, a 26% decrease in stearate, a 50% decrease in oleate and a 50% decrease in total FFA. The reason for these changes in F F A are unknown. Hyperoxia reduced the proportion in the total F F A of palmitate and stearate, the saturated FFA, and increased arachidonic and oleic acids. This relative increase in arachidonic and oleic acids may indicate a differential activation of phospholipase A2 relative to phospholipase A~. By a process of elimination and based upon the observed increases in F F A during hyperoxia, we speculate that the cerebrometabolic effects are likely attributable to lipid peroxidation by oxygen free radicals such as the superoxide anion. The potential involvement of oxygen-induced free radical damage in the vulnerability of the newborn brain to hyperoxia is supported by the observation that superoxide dismutase (SOD) level in the newborn brain is onethird to one-fourth the level in the 2-month-old rat (Mavelli et al., 1978). Similarly SOD levels in the liver appears to increase rapidly (Utsumi et al., 1977), and the rate of oxygen-induced free radical lipid peroxidation is increased in the newborn compared with adult. Acknowledgements--The authors gratefully acknowledge
the comments and suggestions of Drs Etsuro Motoyama and Robert Guthrie, the editorial assistance of Ms Lisa Cohn, and the secretarial assistance of Ms Betty Jacobson and Ms Barb Burgrnan. REFERENCES
Abdab-Barmada M., Moossy J., Nemoto E. M. and Lin M. R. (1986) Hyperoxia produces neuronal necrosis in the rat. J. Neuropathol. exp. NeuroL 45, 233-246. Ahdab-Barmada M., Moossy J. and Painter M. 0980) Pontosubicular necrosis and hyperoxemia. Pediatrics 66, 840-847. Blennow G., Fobergrova J., Nilsson B. and Siesjo B. K. (1979) Effects of bicuculline-induced seizures on cerebral metabolism and circulation of rats rendered hypoglycemic by starvation. Ann. Neurol. 5, 139-151. Arizmendi C. and Medina J. M. (1983) Lactate as an oxidizable substrate for rat brain in vitro during the perinatal period. Biochem. J. 214, 633~535. Dobbing J. and Smart J. L. (1974) Vulnerability of Developing brain and behaviour. Br. Med. Bull. 30, 164-168.
Oxygen toxicity in newborn brain Donaldson H. H. and Hatai S. (1931) On the weight of the parts of the brain and on the percentage of water in them according to brain weight and to age, in albino and wild Norway rats. J. Comp. Neurol. 53, 263-307. Friede R. L. (1972) Ponto-subicular lesions in perinatal anoxia. Arch. Path. Labs. Med. 94, 343-354. Haugaard N. (1964-65) Poisoning of cellular reactions by oxygen. Ann. N.Y. Acad. Sci. 117, 736-744. Hellman J., Vannucci R. C. and Nardis E. E. (1982) Blood-brain barrier permeability to lactic acid in the newborn dog: lactate as a cerebral metabolic fuel. Pediatr. Res. 16, 40~t4. James L. S. and Lanman J. T., eds (1976) History of oxygen therapy and retrolental fibroplasia. Pediatrics 57, 591-598. Johnson P. C. and Bean J. W. (1957) Effect of sympathetic blocking agents on the toxic action o f O 2 at high pressure. Am. J. Physiol. 188, 593-598. Kennedy C., Grave G. D. and Jehle J. W. (1971) Effect of hyperoxia on the cerebral circulation of the newborn puppy. Pediatr. Res. 5, 659-667. Lowry O. H. and Passonneau J. V. (1972) A Flexible System of Enzymatic Analysis, pp. 152-155. Academic Press, New York. Mavelli I., Mondovi B., Federico R. and Rotilio G. (1978) Superoxide dismutase activity in developing rat brain. J. Neurochem. 31, 363-364.
359
Medina J. M. (1985) The role of lactate as an energy substrate for the brain during the early neonatal period. Biol. Neonate 48, 237-244. Rodriguez de Turco E. B. and Bazan N. G. (1983) Changes in free fatty acids and diglycerides in mouse brain at birth and during anoxia. J. Neurochem. 41, 794--800. Shiu G. K., Nemmer J. P. and Nemoto E. M. (1983) Reassessment of brain free fatty acid liberation during global ischemia and its attenuation by barbiturate anesthesia. J. Neurochem. 40, 880-884. Torbati D. and Lambertsen C. J. (1983) Regional cerebral metabolic rate for glucose during hyperbolic oxygeninduced convulsions. Brain Res. 279, 382-386. Torbati D., Lambertsen C. J. and Greenberg J. (1984) Regional cerebral glucose utilization rates in rats during asymptomatic period of exposures to 1, 2 and 3 atmospheres absolute of oxygen. Neuroscience 11, 947-950. Utsumi K., Yoshioka T., Yamanaka N. and Nakazawa T. (1977) Increase in superoxide dismutase activity concomitant with a decrease in lipid peroxidation of liver during post partum development. FEBS Lett. 79, 1 3. Vannucci R. C. and Vannucci S. J. (1978) Cerebral carbohydrate metabolism during hypoglycemia and anoxia in newborn rats. Ann. Neurol. 4, 73-79. Veech R. L., Harris R. L., Veloso D. and Veech E. H. (1973) Freeze-blowing: a new technique for the study of brain in vivo. J. Neurochem. 20, 183-188.