- Email: [email protected]
The effect of electrocortical state on cerebral carbohydrate metabolism in fetal sheep
Developmental Brain Research, 49 (1989) 1-5 Elsevier
BRESD 50944
1
Research Reports
The effect of electrocortical state on cerebral carbohydrate metabolism in fetal sheep Conrad R. Chao 1, A. Roger Hohimer 2 and John M. Bissonnette 2 Departments of Obstetrics and Gynecology, t Columbia University, New York, NY 10032 (U.S.A.) and 2Oregon Health Sciences University, Portland, OR 97201 (U.S.A.)
(Accepted 7 March 1989) Key words: Fetal brain; Electrocortical state; Brain metabolism; Glucose; Lactate; Oxygen
We measured hemispherical cerebral blood flow and arteriovenous differences across the cerebral cortex for glucose, oxygen, and lactate during the two primary electroencephalographic patterns (high and low voltage) in unanesthetized, near-term fetal sheep. Oxygen consumption was 127/~mol/min/100 g brain in high voltage and was 14% higher in low voltage. Glucose uptake was 19 itmol/min/100 g and was 37% higher in low voltage. Cerebral blood flow was 112 ml/min/100 g and was 29% higher in low voltage. The glucose:oxygen quotient increased from 0.91 in high voltage to 1.08 in low voltage. There was a net lactate efflux of 3.2 ~umol/min/100 g during low voltage compared to a net influx of 3.3/~mol/min/100 g in high voltage. During high voltage the fetal brain uses a small amount of lactate for oxidative metabolism. During low voltage, glucose uptake exceeds the oxygen uptake needed for completely aerobic consumption, and a portion of the energy utilized by the brain is produced anaerobically. INTRODUCTION Early studies of fetal cerebral carbohydrate metabolism reported the glucose:oxygen quotient (6 times the glucose arteriovenous difference/oxygen arteriovenous difference) to be approximately one, suggesting that glucose supplies all the aerobic energy requirements of the brain and that virtually all of that glucose is metabolized aerobically s. Studies of arteriovenous differences for various potential cerebral fuels found that only glucose is consumed; in particular, no net influx or efflux could be demonstrated for lactate 6. These studies suggest that glucose is the sole substrate for fetal brain energy production and that cerebral energy production is entirely aerobic in nature. However, these investigations did not examine the effects of electrocortical state on cerebral substrate fluxes. Recently, several investigators have shown that fetal cerebral blood flow, oxygen consumption and glucose utilization are different in the low voltage
state as compared to the high voltage state. Blood flow to most brain regions was higher in low voltage than high voltage 1~'12 and hemispherical oxidative metabolism is greater in low voltage 12 as is cerebral glucose utilization in many regional structures ~. The glucose:oxygen quotient varies significantly from approximately 1.07 in low voltage to 0.83 in high voltage 2. These data demonstrated that cerebral substrate fluxes may vary with electrocortical state and that glucose and oxygen may not be consumed in a precise 6:1 stoichiometry in either state. The finding of a glucose:oxygen quotient greater than one in the low voltage state suggested that a portion of the glucose might be anaerobically consumed; the glucose:oxygen quotient less than one in the high voltage state required that an additional substrate be oxidatively consumed. In order to further explore these questions, we measured glucose, lactate, and oxygen arteriovenous differences and cerebral blood flow in chronically catheterized near-term fetal sheep during the high
Correspondence: C.R. Chao, Department of Obstetrics and Gynecology, Columbia University, 622 W. 168th Street, New York, NY 10032, U.S.A.
0165-3806/89/$03.50 © 1989 Elsevier Science Publishers B.V. (Biomedical Division)
and low voltage electrocortical states. We hypothesized that significant differences in substrate fluxes and cerebral blood flow could be demonstrated as a function of state and that lactate consumption and production might account for the unexplained departure of the glucose:oxygen quotient from unity in both states. MATERIALS AND METHODS
Surgical procedures At 120-129 days of gestation, 23 ewes of mixed Western breed were anesthetized with 7-10 mg/kg i.v. thiamyl sodium followed by positive pressure ventilation with 1% halothane in 50% NO2/50% 02. Hysterotomy was performed and polyvinyl catheters inserted into both fetal brachial arteries and advanced until the tips approached the brachiocephalic artery. A hindlimb vein catheter was placed and advanced into the inferior vena cava. A trephine was used to remove a 1-cm circular portion of the skull in the midline to expose the sagittal sinus midway between the coronal and lambdoidal sutures. The sinus was then punctured with a needle and a polyvinyl catheter inserted caudally to the vicinity of the confluens (about 1 cm). Stainless steel electrodes (Cooner wire, Chatsworth, CA) were placed supradurally 1 cm lateral to the midline at the junction of the lambdoidal and sagittal sutures through small burr holes. All fetal incisions were closed and a polyvinyl catheter sewn to the fetal skin for amniotic fluid pressure measurement. The abdomen was closed and one million units of penicillin G were deposited in the amniotic fluid. Animals were allowed at least 4 days to recover postoperatively prior to any studies.
Physiological measurements Ewes were allowed food and water ad libitum throughout the study period. The peripheral arterial and sagittal sinus catheters were connected to strain gauges (Statham, Oxnard, CA) which had been calibrated with mercury manometers. Pressures were electronically averaged (0.3 Hz low-pass filter, Beckman type 9853 coupler) and recorded on a Beckman chart recorder. Fetal E E G signals were electronically filtered by a 30 Hz low-pass filter and recorded in analog form via a type 9853 coupler. An inde-
pendent 100/~V voltage source was used for calibration. Etectrocortical state was assessed by visual analysis; high voltage (50-200 t~V) and low voltage (<50/~V) epochs were easily distinguished.
Cerebral blood flow measurements Regional cerebral blood was measured using radionuclide labeled microspheres and the reference sample technique ~'13. For each blood flow determination a well dispersed suspension of approximately 2 x 106 15 /~m microspheres in 6% Dextran was injected over 15 s into the hindlimb vein followed by a flush of 3 ml saline. Reference sample blood was withdrawn from a brachial artery at a rate of 2.14 ml/min for a period beginning just prior to the injection and continuing for 3 rain post-injection. The isotopes utilized were ~4~Ce, 51Cr, 85Sr, 95Nb, and 465c. Following completion of the experiments the animal was painlessly killed with i.v. barbiturate and phenytoin. A postmortem examination was performed to verify catheter placement and to remove the brain. The brain tissues were counted with the reference samples in a Packard/Nuclear Data Multichannel Analyzer, Model 601 (Schaumburg, IL). Isotopic separation and blood flow calculation were performed by the methods of Makowski et al. s and Jones et al. 6.
Blood sampling and analysis Arterial and sagittal venous blood samples (3 ml) were anaerobically obtained immediately prior to the blood flow measurement. Sagittal sinus blood has been validated as a means of sampling exclusively cerebral venous drainage, as extracerebral contamination is negligible 1°. Blood gas and pH determinations were made at 39 ~C using a Radiometer Model 3BGS (Copenhagent. Oxygen contents were measured with a Lex-O2-Con (Lexington Instruments, Waltham. MA). Glucose and lactate were measured in duplicate in frozen-thawed whole blood using glucose oxidase and lactate oxidase methods (YSI 23A Glucose Analyzer, Yellow Springs Instruments. Yellow Springs, OH). This instrument utilizes the glucose oxidase-hydrogen peroxide and lactate oxidase-hydrogen peroxide methods with membrane-bound enzymes. Special care was taken to calibrate the instrument daily with standard solutions in the concentrations typical of
those found in fetal sheep blood in order to maximize the accuracy of the instrument in those ranges. Intra-replicate reproducibility was within 4%.
TABLE II
Calculations and data analysis For the purpose of this report, each animal had a single measurement of blood flow and arteriovenous differences in either the high (n = 13) or the low (n = 10) voltage state. Metabolic fluxes were calculated as the product of arteriovenous difference times the cerebral blood flow. The glucose-oxygen quotient was calculated as 6 times the glucose uptake divided by the oxygen uptake. Perfusion pressure was calculated as the peripheral arterial minus the sagittal sinus pressure. Cerebral vascular resistance was calculated as perfusion pressure divided by cerebral blood flow. Statistical significance was inferred by two-tailed unpaired t-tests.
0 2consumption ~mol/min/100g) Glucose consumption* (~mol/min/100g) Lactate flux** (,umol/min/100g) Glucose:oxygenquotient A-V oxygen content (mmol/l) A-V glucose (mmol/l) A-Vlactate* (mmol/l)
RESULTS
Physiological parameters The high and low voltage states were comparable with respect to arterial pO2, sagittal venous pO2, arterial glucose concentration, and arterial blood pressure as presented in Table I. Cerebral blood flow was significantly greater in the low voltage state and this was due to the difference in cerebral vascular resistance, as perfusion pressure (arterial minus sagittal sinus pressure) was unchanged.
Cerebral metabolic measurements during high and low voltage High voltage (n = 13)
Low voltage (n = 10)
127 + 11
145 + 11
19 _+2
26 __3
3.3 __ 1.4 -3.2 + 2.3 0.91 ___0.10 1.08 + 0.08 1.14__0.05 1.03 _+0.10 3.06 __+0.33 3.32 _+0.38 0.029 __0.010 -0.018 _.+_0.017
*P < 0.05; **P < 0.02.
Cerebral metabolic measurements Oxygen consumption (Table II) was greater during low voltage although the difference did not reach statistical significance. Glucose consumption was significantly greater during low voltage. During high voltage, the cerebral cortex was a net consumer of lactate; conversely, during low voltage, a net effiux of lactate was measured. In the case of glucose and oxygen, the changes in net fluxes were primarily due to the increase in cerebral blood flow, as arteriovenous differences were not changed significantly. On the other hand, the arteriovenous difference for lactate was positive in high voltage and negative in low voltage, and the difference between the means was significant. The glucose:oxygen quotient was 0.91 + 0.10 in high voltage and 1.08 + 0.08 in low voltage.
TABLE 1
Blood oxygen tensions and cerebral hemodynamic parameters during high and low voltage
A r t e r i a l p O 2 (Torr) VenouspO2 (Torr) Arterial glucose (mmol/l) Arterial blood pressure (mm Hg) Cerebral blood flow* (ml/min/100 g) Perfusion pressure (mm Hg) Cerebral vascular resistance* (mm Hg/ml/min/100 g) *P < 0.02.
High voltage (n = 13)
Low voltage (n = 10)
22.5 _+ 0.5 17.6 + 0.6 1.27 + 0.08
21.2 + 1.1 16.3 + 0.8 1.13 + 0.05
46 + 1
44 + 1
112 + 8 40 + 1
145 + 11 40 + 1
0.38 + 0.02
0.29 + 0.02
DISCUSSION
These data demonstrate that the high and low voltage states are metabolically distinct in the fetal sheep. High voltage was characterized by inadequate glucose uptake to account for the oxygen consumption, resulting in a glucose:oxygen quotient less than one. A net influx of lactate was measured in this state. During low voltage, glucose uptake exceeded that which could be aerobically consumed, and resulted in a glucose:oxygen quotient greater than one. The fetal brain was a net producer of lactate during this state. These findings confirm and expand on those of Clapp et al. 2 who found a similar
difference in glucose:oxygen quotients in the two states. Our finding of a higher hemispherical glucose uptake during low voltage versus high voltage activity is consistent with the work of Abrams et al. who reported that fetal brain glucose utilization as determined by the [14C]deoxyglucose method correlated with percentage time in REM sleep. (A close correspondence between the low voltage state and REM sleep has been demonstrated in the fetal sheepS.) Heiss et al. 4, in a small positron emission tomography study of adult humans, also reported an increase in regional cerebral glucose metabolism during dreaming, as opposed to a decrease during non-REM sleep. Our data expand on the findings of Richardson et al.12 who also demonstrated greater cerebral oxidative metabolism during low voltage despite an unchanged oxygen arteriovenous difference. Thus, the increase in metabolism was mediated by an increase in cerebral blood flow, rather than an increase in fractional oxygen extraction. Our experiments exhibited a non-significant decrease in oxygen arteriovenous difference such that while cerebral blood flow was 29% greater in low voltage, oxygen consumption increased by only 14%. Since venous oxygen tension was not different between the states, these data suggest that some factor other than maintenance of tissue oxygen tension may well determine the increment in cerebral blood flow between the states. The magnitude of the increase in cerebral blood flow during low voltage relative to high voltage (29%) more closely resembled the magnitude of the increase in glucose uptake (37%) than that of oxygen uptake (14%). A similar closer correspondence between glucose consumption as opposed to oxygen consumption and cerebral blood flow following functional stimulation has been noted in adult humans. Fox et al. 3, using positron emission tomography techniques, found that following visual stimulation oxygen metabolism increases by only 5%, while glucose metabolism increases by 51% and cerebral blood flow increases by 50%. It may be that cerebral blood flow is more closely coupled to glycolysis than to oxidative metabolism. Paulson and Newman 9 have suggested that cerebral blood flow may be regulated by the release of potassium from
astrocyte endfoot processes, a process that could be more closely linked to functional activation than either glycolysis, oxidative metabolism or maintenance of tissue oxygen partial pressures. One might speculate that the fetus, having a lower arterial pO 2 than the adult, might be closer to some cerebral 'anaerobic threshold'. To our knowledge no experiments have directly addressed this speculation. However, the hypoxic fetal sheep can increase cerebral blood flow to maintain oxygen delivery and consumption even at very low levels of arterial oxygenation 7. Although we have no direct measures of cerebral tissue or mitochondrial pO 2 and therefore cannot exclude the possibility that during low voltage there might be some degree of relative hypoxia which could result in lactate production, the sagittal venous pO 2 does not suggest any difference in oxygenation between the low voltage and high voltage electrocortical states. The finding of a net cerebral lactate efflux and a glucose:oxygen quotient greater than unity during the low voltage electrocortical state indicate anaerobic consumption of a portion of that glucose despite apparently adequate tissue oxygenation. This phenomenon, 'aerobic glycolysis', has been explored at the tissue metabolite level in the adult animal. Ueki et al. 14 reported that following somatosensory stimulation in rats cerebral blood flow increased by 33% and glucose metabolism increased by 57%. Tissue lactate-induced bioluminescence increased by 30% while tissue glucose-induced bioluminescence decreased by 21%. Our study together with those of these authors suggest that anaerobic cerebral glucose consumption may occur during some normal physiologic processes, not just during conditions of severe cerebral hypoxia. Perhaps, as Ueki et al. ~4 have speculated, an increase in ion transport due to the increased transmitter activity during low voltage may result in glycolytic activity to support membrane ion exchange pumps that are dependent on anaerobically produced ATP by virtue of their intracellular location. Although our measurements of lactate flux support the idea that primarily glucose and to a minor extent lactate are the only oxidatively utilized substrates during high voltage and that glucose is the sole substrate for both oxidative and anaerobic energy metabolism during low voltage, it is possible
that o t h e r as yet u n m e a s u r e d energy substrates are utilized for oxidation. If that were the case m o r e of the glucose taken up might actually be utilized for processes of n o r m a l growth and d e v e l o p m e n t rather than for energy production. In conclusion, we have d e m o n s t r a t e d that the c a r b o h y d r a t e metabolism of the fetal brain is dependent on electrocortical state. High voltage is characterized by a glucose:oxygen quotient less than one and a net u p t a k e of lactate. Low voltage exhibits a glucose:oxygen quotient greater than one and a net efflux of lactate. Cerebral blood flow increases in
low voltage and the increase exceeds the m e a s u r e d increase in oxygen consumption. Glucose consumption is also higher in low voltage and a p o r t i o n of that glucose appears to be c o n s u m e d anaerobically.
REFERENCES
8 Makowski, E.L., Schneider, J.M., Tsoulos, N.G., Colwill, J.R., Battaglia, F.C. and Meschia, G., Cerebral blood flow, oxygen consumption, and glucose utilization of fetal lambs in utero, Am. J. Obstet. Gynecol., 114 (1972) 292-303. 9 Paulson, O.B. and Newman, E.A., Does the release of potassium from astrocyte endfeet regulate cerebral blood flow?, Science, 237 (1987) 896-898. 10 Purves, M.J. and James, I.M., Observations on the control of cerebral blood flow in the sheep fetus and newborn lamb, Circ. Res., 25 (1969) 651-667. 11 Rankin, H.G., Landauer, M., Tian, Q. and Phenetton, T.M., Ovine fetal electrocortical activity and regional cerebral blood flow, J. Dev. Physiol., 9 (1987) 537-542. 12 Richardson, B.S., Patrick, J.E. and Abduljabbar, H., Cerebral oxidative metabolism in the fetal lamb: relationship to electrocortical state, Am. J. Obstet. Gynecol., 153 (1985) 426-431. 13 Rudolph, A.M. and Heymann, M.A., The circulation of the fetus in utero: methods for studying distribution of blood flow, cardiac output, and organ blood flow, Circ. Res., 21 (1967) 163-184. 14 Ueki, M., Linn, E and Hossman, K.-A., Functional activation of cerebral blood flow and metabolism before and after global ischemia of rat brain, J. Cereb. Blood Flow Metab., 8 (1988) 486-494.
1 Abrams, R.M., Hutchinson, A.A., Jay, T.M., Sokoloff, L. and Kennedy, C., Local cerebral glucose utilization nonselectively elevated in rapid eye movement sleep of the fetus, Dev. Brain Res., 40 (1988) 65-70. 2 Clapp, J.F., Szeto, H.H., Abrams, R., Larrow, R. and Mann, L.I., Physiologic variability and fetal electrocortical activity, Am. J. Obstet. Gynecol., 136 (1980) 1045-1050. 3 Fox, P.T., Raichle, M.E., Mintun, M.A. and Dence, C., Nonoxidative glucose consumption during focal physiologic neural activity, Science, 241 (1988) 462-464. 4 Heiss, W.D., Pawlik, G., Herholz, K., Wagner, R. and Wienhard, K., Regional cerebral glucose metabolism in man during wakefulness, sleep, and dreaming, Brain Res., 327 (1985) 362-366. 5 Ioffe, S., Jansen, A.H., Russell, B.J. and Chernick, V., Sleep, wakefulness, and the monosynaptic reflex in fetal and newborn lambs, Pfliigers Arch., 388 (1980) 149-157. 6 Jones, M.D., Burd, L.I., Makowski, E.L., Meschia, G. and Battaglia, EC., Cerebral metabolism in sheep: a comparative study of the adult, the lamb, and the fetus, Am. J. Physiol., 229 (1975) 235-239. 7 Jones, M.D., Sheldon, R.E., Peeters, L.L., Meschia, G., Battaglia, F.C. and Makowski, E.L., Fetal cerebral oxygen consumption at different levels of oxygenation, J. Appl. Physiol., 43 (1977) 1080-1084.
ACKNOWLEDGEMENTS This work was s u p p o r t e d by N I H G r a n t s H D 10034-12, H L 7596-02, and the O r e g o n H e a r t Association. The technical assistance of Sharon K n o p p , Neil N o t a r o b e r t o , Gail Willeke and Wally W i c k h a m is gratefully acknowledged.
BRESD 50944
1
Research Reports
The effect of electrocortical state on cerebral carbohydrate metabolism in fetal sheep Conrad R. Chao 1, A. Roger Hohimer 2 and John M. Bissonnette 2 Departments of Obstetrics and Gynecology, t Columbia University, New York, NY 10032 (U.S.A.) and 2Oregon Health Sciences University, Portland, OR 97201 (U.S.A.)
(Accepted 7 March 1989) Key words: Fetal brain; Electrocortical state; Brain metabolism; Glucose; Lactate; Oxygen
We measured hemispherical cerebral blood flow and arteriovenous differences across the cerebral cortex for glucose, oxygen, and lactate during the two primary electroencephalographic patterns (high and low voltage) in unanesthetized, near-term fetal sheep. Oxygen consumption was 127/~mol/min/100 g brain in high voltage and was 14% higher in low voltage. Glucose uptake was 19 itmol/min/100 g and was 37% higher in low voltage. Cerebral blood flow was 112 ml/min/100 g and was 29% higher in low voltage. The glucose:oxygen quotient increased from 0.91 in high voltage to 1.08 in low voltage. There was a net lactate efflux of 3.2 ~umol/min/100 g during low voltage compared to a net influx of 3.3/~mol/min/100 g in high voltage. During high voltage the fetal brain uses a small amount of lactate for oxidative metabolism. During low voltage, glucose uptake exceeds the oxygen uptake needed for completely aerobic consumption, and a portion of the energy utilized by the brain is produced anaerobically. INTRODUCTION Early studies of fetal cerebral carbohydrate metabolism reported the glucose:oxygen quotient (6 times the glucose arteriovenous difference/oxygen arteriovenous difference) to be approximately one, suggesting that glucose supplies all the aerobic energy requirements of the brain and that virtually all of that glucose is metabolized aerobically s. Studies of arteriovenous differences for various potential cerebral fuels found that only glucose is consumed; in particular, no net influx or efflux could be demonstrated for lactate 6. These studies suggest that glucose is the sole substrate for fetal brain energy production and that cerebral energy production is entirely aerobic in nature. However, these investigations did not examine the effects of electrocortical state on cerebral substrate fluxes. Recently, several investigators have shown that fetal cerebral blood flow, oxygen consumption and glucose utilization are different in the low voltage
state as compared to the high voltage state. Blood flow to most brain regions was higher in low voltage than high voltage 1~'12 and hemispherical oxidative metabolism is greater in low voltage 12 as is cerebral glucose utilization in many regional structures ~. The glucose:oxygen quotient varies significantly from approximately 1.07 in low voltage to 0.83 in high voltage 2. These data demonstrated that cerebral substrate fluxes may vary with electrocortical state and that glucose and oxygen may not be consumed in a precise 6:1 stoichiometry in either state. The finding of a glucose:oxygen quotient greater than one in the low voltage state suggested that a portion of the glucose might be anaerobically consumed; the glucose:oxygen quotient less than one in the high voltage state required that an additional substrate be oxidatively consumed. In order to further explore these questions, we measured glucose, lactate, and oxygen arteriovenous differences and cerebral blood flow in chronically catheterized near-term fetal sheep during the high
Correspondence: C.R. Chao, Department of Obstetrics and Gynecology, Columbia University, 622 W. 168th Street, New York, NY 10032, U.S.A.
0165-3806/89/$03.50 © 1989 Elsevier Science Publishers B.V. (Biomedical Division)
and low voltage electrocortical states. We hypothesized that significant differences in substrate fluxes and cerebral blood flow could be demonstrated as a function of state and that lactate consumption and production might account for the unexplained departure of the glucose:oxygen quotient from unity in both states. MATERIALS AND METHODS
Surgical procedures At 120-129 days of gestation, 23 ewes of mixed Western breed were anesthetized with 7-10 mg/kg i.v. thiamyl sodium followed by positive pressure ventilation with 1% halothane in 50% NO2/50% 02. Hysterotomy was performed and polyvinyl catheters inserted into both fetal brachial arteries and advanced until the tips approached the brachiocephalic artery. A hindlimb vein catheter was placed and advanced into the inferior vena cava. A trephine was used to remove a 1-cm circular portion of the skull in the midline to expose the sagittal sinus midway between the coronal and lambdoidal sutures. The sinus was then punctured with a needle and a polyvinyl catheter inserted caudally to the vicinity of the confluens (about 1 cm). Stainless steel electrodes (Cooner wire, Chatsworth, CA) were placed supradurally 1 cm lateral to the midline at the junction of the lambdoidal and sagittal sutures through small burr holes. All fetal incisions were closed and a polyvinyl catheter sewn to the fetal skin for amniotic fluid pressure measurement. The abdomen was closed and one million units of penicillin G were deposited in the amniotic fluid. Animals were allowed at least 4 days to recover postoperatively prior to any studies.
Physiological measurements Ewes were allowed food and water ad libitum throughout the study period. The peripheral arterial and sagittal sinus catheters were connected to strain gauges (Statham, Oxnard, CA) which had been calibrated with mercury manometers. Pressures were electronically averaged (0.3 Hz low-pass filter, Beckman type 9853 coupler) and recorded on a Beckman chart recorder. Fetal E E G signals were electronically filtered by a 30 Hz low-pass filter and recorded in analog form via a type 9853 coupler. An inde-
pendent 100/~V voltage source was used for calibration. Etectrocortical state was assessed by visual analysis; high voltage (50-200 t~V) and low voltage (<50/~V) epochs were easily distinguished.
Cerebral blood flow measurements Regional cerebral blood was measured using radionuclide labeled microspheres and the reference sample technique ~'13. For each blood flow determination a well dispersed suspension of approximately 2 x 106 15 /~m microspheres in 6% Dextran was injected over 15 s into the hindlimb vein followed by a flush of 3 ml saline. Reference sample blood was withdrawn from a brachial artery at a rate of 2.14 ml/min for a period beginning just prior to the injection and continuing for 3 rain post-injection. The isotopes utilized were ~4~Ce, 51Cr, 85Sr, 95Nb, and 465c. Following completion of the experiments the animal was painlessly killed with i.v. barbiturate and phenytoin. A postmortem examination was performed to verify catheter placement and to remove the brain. The brain tissues were counted with the reference samples in a Packard/Nuclear Data Multichannel Analyzer, Model 601 (Schaumburg, IL). Isotopic separation and blood flow calculation were performed by the methods of Makowski et al. s and Jones et al. 6.
Blood sampling and analysis Arterial and sagittal venous blood samples (3 ml) were anaerobically obtained immediately prior to the blood flow measurement. Sagittal sinus blood has been validated as a means of sampling exclusively cerebral venous drainage, as extracerebral contamination is negligible 1°. Blood gas and pH determinations were made at 39 ~C using a Radiometer Model 3BGS (Copenhagent. Oxygen contents were measured with a Lex-O2-Con (Lexington Instruments, Waltham. MA). Glucose and lactate were measured in duplicate in frozen-thawed whole blood using glucose oxidase and lactate oxidase methods (YSI 23A Glucose Analyzer, Yellow Springs Instruments. Yellow Springs, OH). This instrument utilizes the glucose oxidase-hydrogen peroxide and lactate oxidase-hydrogen peroxide methods with membrane-bound enzymes. Special care was taken to calibrate the instrument daily with standard solutions in the concentrations typical of
those found in fetal sheep blood in order to maximize the accuracy of the instrument in those ranges. Intra-replicate reproducibility was within 4%.
TABLE II
Calculations and data analysis For the purpose of this report, each animal had a single measurement of blood flow and arteriovenous differences in either the high (n = 13) or the low (n = 10) voltage state. Metabolic fluxes were calculated as the product of arteriovenous difference times the cerebral blood flow. The glucose-oxygen quotient was calculated as 6 times the glucose uptake divided by the oxygen uptake. Perfusion pressure was calculated as the peripheral arterial minus the sagittal sinus pressure. Cerebral vascular resistance was calculated as perfusion pressure divided by cerebral blood flow. Statistical significance was inferred by two-tailed unpaired t-tests.
0 2consumption ~mol/min/100g) Glucose consumption* (~mol/min/100g) Lactate flux** (,umol/min/100g) Glucose:oxygenquotient A-V oxygen content (mmol/l) A-V glucose (mmol/l) A-Vlactate* (mmol/l)
RESULTS
Physiological parameters The high and low voltage states were comparable with respect to arterial pO2, sagittal venous pO2, arterial glucose concentration, and arterial blood pressure as presented in Table I. Cerebral blood flow was significantly greater in the low voltage state and this was due to the difference in cerebral vascular resistance, as perfusion pressure (arterial minus sagittal sinus pressure) was unchanged.
Cerebral metabolic measurements during high and low voltage High voltage (n = 13)
Low voltage (n = 10)
127 + 11
145 + 11
19 _+2
26 __3
3.3 __ 1.4 -3.2 + 2.3 0.91 ___0.10 1.08 + 0.08 1.14__0.05 1.03 _+0.10 3.06 __+0.33 3.32 _+0.38 0.029 __0.010 -0.018 _.+_0.017
*P < 0.05; **P < 0.02.
Cerebral metabolic measurements Oxygen consumption (Table II) was greater during low voltage although the difference did not reach statistical significance. Glucose consumption was significantly greater during low voltage. During high voltage, the cerebral cortex was a net consumer of lactate; conversely, during low voltage, a net effiux of lactate was measured. In the case of glucose and oxygen, the changes in net fluxes were primarily due to the increase in cerebral blood flow, as arteriovenous differences were not changed significantly. On the other hand, the arteriovenous difference for lactate was positive in high voltage and negative in low voltage, and the difference between the means was significant. The glucose:oxygen quotient was 0.91 + 0.10 in high voltage and 1.08 + 0.08 in low voltage.
TABLE 1
Blood oxygen tensions and cerebral hemodynamic parameters during high and low voltage
A r t e r i a l p O 2 (Torr) VenouspO2 (Torr) Arterial glucose (mmol/l) Arterial blood pressure (mm Hg) Cerebral blood flow* (ml/min/100 g) Perfusion pressure (mm Hg) Cerebral vascular resistance* (mm Hg/ml/min/100 g) *P < 0.02.
High voltage (n = 13)
Low voltage (n = 10)
22.5 _+ 0.5 17.6 + 0.6 1.27 + 0.08
21.2 + 1.1 16.3 + 0.8 1.13 + 0.05
46 + 1
44 + 1
112 + 8 40 + 1
145 + 11 40 + 1
0.38 + 0.02
0.29 + 0.02
DISCUSSION
These data demonstrate that the high and low voltage states are metabolically distinct in the fetal sheep. High voltage was characterized by inadequate glucose uptake to account for the oxygen consumption, resulting in a glucose:oxygen quotient less than one. A net influx of lactate was measured in this state. During low voltage, glucose uptake exceeded that which could be aerobically consumed, and resulted in a glucose:oxygen quotient greater than one. The fetal brain was a net producer of lactate during this state. These findings confirm and expand on those of Clapp et al. 2 who found a similar
difference in glucose:oxygen quotients in the two states. Our finding of a higher hemispherical glucose uptake during low voltage versus high voltage activity is consistent with the work of Abrams et al. who reported that fetal brain glucose utilization as determined by the [14C]deoxyglucose method correlated with percentage time in REM sleep. (A close correspondence between the low voltage state and REM sleep has been demonstrated in the fetal sheepS.) Heiss et al. 4, in a small positron emission tomography study of adult humans, also reported an increase in regional cerebral glucose metabolism during dreaming, as opposed to a decrease during non-REM sleep. Our data expand on the findings of Richardson et al.12 who also demonstrated greater cerebral oxidative metabolism during low voltage despite an unchanged oxygen arteriovenous difference. Thus, the increase in metabolism was mediated by an increase in cerebral blood flow, rather than an increase in fractional oxygen extraction. Our experiments exhibited a non-significant decrease in oxygen arteriovenous difference such that while cerebral blood flow was 29% greater in low voltage, oxygen consumption increased by only 14%. Since venous oxygen tension was not different between the states, these data suggest that some factor other than maintenance of tissue oxygen tension may well determine the increment in cerebral blood flow between the states. The magnitude of the increase in cerebral blood flow during low voltage relative to high voltage (29%) more closely resembled the magnitude of the increase in glucose uptake (37%) than that of oxygen uptake (14%). A similar closer correspondence between glucose consumption as opposed to oxygen consumption and cerebral blood flow following functional stimulation has been noted in adult humans. Fox et al. 3, using positron emission tomography techniques, found that following visual stimulation oxygen metabolism increases by only 5%, while glucose metabolism increases by 51% and cerebral blood flow increases by 50%. It may be that cerebral blood flow is more closely coupled to glycolysis than to oxidative metabolism. Paulson and Newman 9 have suggested that cerebral blood flow may be regulated by the release of potassium from
astrocyte endfoot processes, a process that could be more closely linked to functional activation than either glycolysis, oxidative metabolism or maintenance of tissue oxygen partial pressures. One might speculate that the fetus, having a lower arterial pO 2 than the adult, might be closer to some cerebral 'anaerobic threshold'. To our knowledge no experiments have directly addressed this speculation. However, the hypoxic fetal sheep can increase cerebral blood flow to maintain oxygen delivery and consumption even at very low levels of arterial oxygenation 7. Although we have no direct measures of cerebral tissue or mitochondrial pO 2 and therefore cannot exclude the possibility that during low voltage there might be some degree of relative hypoxia which could result in lactate production, the sagittal venous pO 2 does not suggest any difference in oxygenation between the low voltage and high voltage electrocortical states. The finding of a net cerebral lactate efflux and a glucose:oxygen quotient greater than unity during the low voltage electrocortical state indicate anaerobic consumption of a portion of that glucose despite apparently adequate tissue oxygenation. This phenomenon, 'aerobic glycolysis', has been explored at the tissue metabolite level in the adult animal. Ueki et al. 14 reported that following somatosensory stimulation in rats cerebral blood flow increased by 33% and glucose metabolism increased by 57%. Tissue lactate-induced bioluminescence increased by 30% while tissue glucose-induced bioluminescence decreased by 21%. Our study together with those of these authors suggest that anaerobic cerebral glucose consumption may occur during some normal physiologic processes, not just during conditions of severe cerebral hypoxia. Perhaps, as Ueki et al. ~4 have speculated, an increase in ion transport due to the increased transmitter activity during low voltage may result in glycolytic activity to support membrane ion exchange pumps that are dependent on anaerobically produced ATP by virtue of their intracellular location. Although our measurements of lactate flux support the idea that primarily glucose and to a minor extent lactate are the only oxidatively utilized substrates during high voltage and that glucose is the sole substrate for both oxidative and anaerobic energy metabolism during low voltage, it is possible
that o t h e r as yet u n m e a s u r e d energy substrates are utilized for oxidation. If that were the case m o r e of the glucose taken up might actually be utilized for processes of n o r m a l growth and d e v e l o p m e n t rather than for energy production. In conclusion, we have d e m o n s t r a t e d that the c a r b o h y d r a t e metabolism of the fetal brain is dependent on electrocortical state. High voltage is characterized by a glucose:oxygen quotient less than one and a net u p t a k e of lactate. Low voltage exhibits a glucose:oxygen quotient greater than one and a net efflux of lactate. Cerebral blood flow increases in
low voltage and the increase exceeds the m e a s u r e d increase in oxygen consumption. Glucose consumption is also higher in low voltage and a p o r t i o n of that glucose appears to be c o n s u m e d anaerobically.
REFERENCES
8 Makowski, E.L., Schneider, J.M., Tsoulos, N.G., Colwill, J.R., Battaglia, F.C. and Meschia, G., Cerebral blood flow, oxygen consumption, and glucose utilization of fetal lambs in utero, Am. J. Obstet. Gynecol., 114 (1972) 292-303. 9 Paulson, O.B. and Newman, E.A., Does the release of potassium from astrocyte endfeet regulate cerebral blood flow?, Science, 237 (1987) 896-898. 10 Purves, M.J. and James, I.M., Observations on the control of cerebral blood flow in the sheep fetus and newborn lamb, Circ. Res., 25 (1969) 651-667. 11 Rankin, H.G., Landauer, M., Tian, Q. and Phenetton, T.M., Ovine fetal electrocortical activity and regional cerebral blood flow, J. Dev. Physiol., 9 (1987) 537-542. 12 Richardson, B.S., Patrick, J.E. and Abduljabbar, H., Cerebral oxidative metabolism in the fetal lamb: relationship to electrocortical state, Am. J. Obstet. Gynecol., 153 (1985) 426-431. 13 Rudolph, A.M. and Heymann, M.A., The circulation of the fetus in utero: methods for studying distribution of blood flow, cardiac output, and organ blood flow, Circ. Res., 21 (1967) 163-184. 14 Ueki, M., Linn, E and Hossman, K.-A., Functional activation of cerebral blood flow and metabolism before and after global ischemia of rat brain, J. Cereb. Blood Flow Metab., 8 (1988) 486-494.
1 Abrams, R.M., Hutchinson, A.A., Jay, T.M., Sokoloff, L. and Kennedy, C., Local cerebral glucose utilization nonselectively elevated in rapid eye movement sleep of the fetus, Dev. Brain Res., 40 (1988) 65-70. 2 Clapp, J.F., Szeto, H.H., Abrams, R., Larrow, R. and Mann, L.I., Physiologic variability and fetal electrocortical activity, Am. J. Obstet. Gynecol., 136 (1980) 1045-1050. 3 Fox, P.T., Raichle, M.E., Mintun, M.A. and Dence, C., Nonoxidative glucose consumption during focal physiologic neural activity, Science, 241 (1988) 462-464. 4 Heiss, W.D., Pawlik, G., Herholz, K., Wagner, R. and Wienhard, K., Regional cerebral glucose metabolism in man during wakefulness, sleep, and dreaming, Brain Res., 327 (1985) 362-366. 5 Ioffe, S., Jansen, A.H., Russell, B.J. and Chernick, V., Sleep, wakefulness, and the monosynaptic reflex in fetal and newborn lambs, Pfliigers Arch., 388 (1980) 149-157. 6 Jones, M.D., Burd, L.I., Makowski, E.L., Meschia, G. and Battaglia, EC., Cerebral metabolism in sheep: a comparative study of the adult, the lamb, and the fetus, Am. J. Physiol., 229 (1975) 235-239. 7 Jones, M.D., Sheldon, R.E., Peeters, L.L., Meschia, G., Battaglia, F.C. and Makowski, E.L., Fetal cerebral oxygen consumption at different levels of oxygenation, J. Appl. Physiol., 43 (1977) 1080-1084.
ACKNOWLEDGEMENTS This work was s u p p o r t e d by N I H G r a n t s H D 10034-12, H L 7596-02, and the O r e g o n H e a r t Association. The technical assistance of Sharon K n o p p , Neil N o t a r o b e r t o , Gail Willeke and Wally W i c k h a m is gratefully acknowledged.