INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

CEREBRAL PROTEIN SYNTHESIS has been studied during brain growth and development, both pre- and postnatally in sheep (1, 24) and postnatally in the rat (6, 26), because proteins are integral components of structural elements in brain tissue and with the metabolism of brain protein directly related to maturational events (26). Initial study in the near-term ovine fetus by Schaefer and Krishnamurti (24) using a tyrosine isotopic-dilution technique showed high rates of cerebral protein synthesis with a fractional protein synthetic rate between 14 and 37%/day and an absolute rate of synthesis of ~1 g/day. A subsequent study in fetal sheep by Abrams and colleagues (1) using [¹⁴C]leucine autoradiography reported a rate of leucine incorporation into brain protein of ~5 nmol · g¹ · min¹, with an overall increase through the latter part of gestation and into the early postnatal period likely reflecting cerebral myelination. Postnatal studies in rats have similarly shown high rates of brain protein synthesis during early development, with peak values occurring shortly after birth and gradually decreasing thereafter (6, 26). While high rates of cerebral protein synthesis are thus evident during early life in both sheep and rats in support of the brain's growth and development, these rates are considerably greater than anticipated protein accretion, indicating that protein degradation must also play an active role in the development of the brain (6).

The increase in brain weight for the ovine species during early development occurs in two phases, one up to 90 days postconception followed by a more rapid and larger increase thereafter that continues to birth at ~145 days, leading to the classification of sheep as a prenatal brain developer (14). These two phases appear to reflect an increase in neuroblast multiplication, followed by neuroglial multiplication and myelination, and with growth and development of the cerebral hemispheres in advance of that in the cerebellum. For the ovine species, well-differentiated electrocortical (ECOG) patterns as a measure of neurobehavioral development are evident from 120 days gestation onward, with initially a high proportion of time in the low-voltage (LV) ECOG state in the presence of rapid eye movements (REM), and with a progressive decrease in this activity thereafter (27). Observations on the maturation of electrocortical patterns in utero or of behavioral activity from birth indicate a similar trend in the development of sleep-wakefulness patterns in humans and other mammals (22, 27), with the rate of such development in relation to birth well correlated with the neuroanatomic maturity of the brain (20).

A behavioral state effect on cerebral metabolic rate is evident for the ovine fetus near term, with a significant increase in cerebral blood flow and in oxygen and glucose uptake during the LV/REM state compared with that of the high-voltage (HV) ECOG state with no rapid eye movements [non-rapid eye movement (NREM)] (21), presumably reflecting an increase in neuronal functional activity and/or synthetic processes within the brain. The relationship of behavioral state activity to biosynthetic processes within the brain during development remains unclear, although recent studies in adult animals have shown higher rates of cerebral protein synthesis to be positively correlated with the occurrence of NREM sleep rather than REM sleep (16, 19). We have therefore determined the cerebral uptake of leucine and of [¹⁴C]leucine during continuous tracer infusion as measures of leucine metabolism in relation to behavioral state activity in the ovine fetus near term. Regional fractional protein synthetic rates and flux of leucine into protein within the brain were also determined from the specific activity of protein-bound [¹⁴C]leucine within brain tissues as additional measures of cerebral protein synthesis.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Surgical procedure. Nine fetal sheep were surgically prepared at ~129 days gestation (term = 145 days). Anesthesia was induced with an intravenous injection of thiopental sodium (1 g in 40 ml solution; Abbott Laboratories, Montreal, Canada) and was maintained with 1-1.5% halothane in oxygen (Halocarbon Laboratories, Hackensack, NJ). Before fetal surgery, a polyvinyl catheter (V11, Bolab, Lake Havasu City, AZ) was placed in the maternal femoral vein. Using sterile technique, a midline incision was made in the ewe's lower abdominal wall, and the uterus was palpated to determine fetal position. An incision was then made in the uterus, and the fetal head, neck, and upper body were exteriorized. Polyvinyl catheters (V4, Bolab) were placed in the axillary artery for blood sampling, in the cephalic vein for administration of antibiotics and radiolabeled leucine, in the trachea for monitoring of fetal breathing movements, and in the amniotic cavity for pressure recording. Electrodes of teflon-coated stainless steel wire (Cooner Wire, Chatsworth, CA) were placed biparietally on the dura for ECOG recording, through the lateral orbital ridge of the zygomatic bone of each eye for electroocular (EOG) recording, and on the nuchal muscle for electromyographic (EMG) recording. A midline window of bone (~2 cm²) was removed from the fetal skull, immediately anterior to the coronal suture; the dura on either side of the superior sagittal sinus was incised, and a transit time flow probe (3 mm, S-series, Transonic Systems, Ithaca, NY) was secured around the vessel. The cavity surrounding the probe was filled with bone wax to prevent leakage of cerebral spinal fluid. Another small midline window of bone (~0.5 cm²) was removed between the coronal and lambdoid sutures and rostral to the bifurcation of the superior sagittal sinus. A nonocclusive polyvinyl catheter (V4, Bolab) was inserted into the vessel for blood sampling and was directed caudally by ~1 cm. Sagittal sinus blood flow (Q_ss) was briefly monitored intraoperatively to ensure that the flow probe was functioning properly and to observe the effects of subsequent catheter placement in the sagittal sinus on Q_ss. After placement of the probe and catheter, the scalp was sewn over. The uterus and abdomen were closed in layers, and all catheters, electrodes, and the Transonic flow probe lead were exteriorized to the flank of the ewe and secured to the ewe's back in a plastic pouch. The ewes received a long-acting, broad-spectrum antibiotic immediately before surgery (1.2 g oxytetracycline im; LA-200, Rogar, STB, London, Canada) and were given intravenous fluids throughout the surgery (1,000 ml 0.9% saline solution).

Postoperative care. After surgery, ewes were placed in metabolic cages suitable for continuous monitoring and were given an analgesic (1.5 ml Banamine). Antibiotics were administered for 3 days postoperatively to the fetus (1,000,000 IU sodium penicillin G iv; Ayerst, Montreal, Canada) and to the amniotic cavity (1,000,000 IU sodium penicillin G). The ewes were allowed 4 days of postoperative recovery, during which maternal and fetal catheters were flushed daily with heparinized saline to maintain patency, and fetal arterial blood was collected for blood gas analysis. Animals were allowed food and water ad libitum, and all surgical, postoperative, and experimental procedures followed the guidelines provided by the Canadian Council on Animal Care and the University of Western Ontario Council on Animal Care.

Physiological measurements. On the day of experimental study, strain-gauge transducers (Statham model P-2310, Gould, Oxnard, CA) and a physiograph recorder (model 78D, Grass Instrument, Quincy, MA) were used to record amniotic and tracheal pressures. Fetal ECOG, EOG, and EMG potentials were displayed directly on the chart recorder after passing through a passive band-pass filter, 0.3-30 Hz on the preamplifier for ECOG and EOG recordings, and 10-90 Hz on the preamplifier for EMG recordings. The amplified, filtered ECOG signal was further processed by means of a frequency integrator with the ECOG frequency shift displayed separately on the chart recorder. The mean volume blood flow measurement derived from the flow probe was processed using a Transonic flowmeter (model T208, Transonic Systems) and was displayed on the chart recorder with the normal scale calibration set at 0-50 ml/min. Animals were studied over a 6-h period, with a continuous infusion of L-[1-¹⁴C]leucine in sterile normal saline into the fetal cephalic vein at 0.15 µCi/min (in 30 ml normal saline at a rate of 5 ml/h). Arterial blood samples (1 ml) were drawn at 15, 30, 60, and 90 min of infusion to determine that the specific activity of leucine had reached steady state. An arterial blood sample (1 ml) was additionally drawn before initiation of the infusion and was analyzed for blood gases and pH as a measure of fetal well being. After the first 2 h of infusion, spontaneous changes in behavioral state were scored and four LV/REM and four HV/NREM epochs were studied. Cerebral arterial-venous (A-V) difference samples were collected from the axillary artery (1.5 ml) and the sagittal sinus (1.5 ml) at least 5 min into each state epoch, with LV measurements made only if EOG activity was present and in the absence of nuchal muscle EMG activity, and HV measurements made only if EOG activity was absent and in the presence of nuchal muscle EMG activity. The remaining whole blood was then spun (4°C, 10 min at 9,000 g), and the plasma was titrated off and divided into two equal aliquots for duplicate measurements. All plasma aliquots were frozen at 80°C for subsequent measurement of leucine concentration and specific activity and -ketoisocaproic acid (-KIC) concentration.

Chemical analysis. Blood gases and pH were measured using an ABL-500 blood gas analyzer with temperature corrected to 39.5°C (Radiometer, Copenhagen, Denmark).

Data analysis and mathematical calculations. Q_ss as a relative measure of cerebral blood flow was determined every 10 s and averaged over the duration of each A-V difference blood sample draw (~60 s) using a computerized data-acquisition system (CADA, Hartronix Computer Solutions, Toronto, Canada).

1

1

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Fetal arterial blood gases and pH measured within the normal range, with an average PO₂ of 23.3 ± 1.2 mmHg, PCO₂ of 50.3 ± 1.4 mmHg, and pH of 7.37 ± 0.01, indicating that all animals were in good health at the time of experimental study. SA_P leucine reached steady-state values by 120 min of infusion. Figure 1 illustrates mean SA_P leucine at time 0 and at 15, 30, 60, and 90 min of infusion and throughout the blood sampling period, demonstrating the attainment of plateau values, which averaged 6.7 ± 0.7 dpm/nmol. SA_IF leucine values measured in the fetal brain tissue at the end of the 6-h infusion period were lower than SA_P leucine plateau values, averaging 5.3 ± 0.4 dpm/nmol, while SA_PB values were considerably lower, averaging 0.27 ± 0.02 dpm/nmol.

View larger version (10K): [in this window] [in a new window]   Fig. 1.   Arterial plasma leucine specific activity as measured throughout the infusion period with attainment of plateau specific activity values by 120 min of infusion.

Cerebral leucine and [¹⁴C]leucine uptake. Blood flow in the superior sagittal sinus was affected by behavioral state activity, as studied at the time of cerebral A-V difference leucine measurements, and was increased in each of the animals during the LV/REM state compared with the HV/NREM state, 19.8 ± 2.3 vs. 15.0 ± 1.7 ml/min (P < 0.001) (Table 1). Accordingly, cerebral blood flow per unit mass of brain tissue was similarly increased, 206 ± 27 vs. 156 ± 20 ml · 100 g¹ · min¹ (P < 0.002) (Table 1). Conversely, the A-V difference for leucine and leucine fractional extraction were increased and to a greater extent during the HV/NREM state, 4.7 ± 1.0 vs. 2.0 ± 0.8 µmol/l (P < 0.001) and 2.1 ± 0.4 vs. 0.8 ± 0.5% (P < 0.02), respectively, but with no measurable A-V difference for -KIC during either behavioral state (Table 1). As such, cortical leucine uptake and thus cerebral leucine uptake were both increased during the HV/NREM state compared with the LV/REM state, 77 ± 25 vs. 44 ± 30 nmol/min and 796 ± 266 vs. 467 ± 315 nmol · 100 g¹ · min¹, respectively (both P < 0.02) (Table 1). The A-V difference for [¹⁴C]leucine and [¹⁴C]leucine fractional extraction were also increased during the HV/NREM state and more so than their unlabeled counterparts, 62 ± 9 vs. 20 ± 10 dpm/ml and 4.4 ± 0.5 vs. 1.6 ± 0.8%, respectively (both P < 0.01), but again with no measurable A-V difference for -[¹⁴C]KIC during either behavioral state (Table 1). As such, cortical [¹⁴C]leucine uptake and thus cerebral [¹⁴C]leucine uptake were also increased during the HV/NREM state and more so than their unlabeled counterparts, 971 ± 199 vs. 341 ± 218 dpm/min and 10,149 ± 2,266 vs. 3,460 ± 2,303 dpm · 100 g¹ · min¹, respectively (both P = 0.05) (Table 1).

Regional cerebral protein synthesis. Cerebral fractional synthetic rates for protein averaged ~20%/day and were similar as measured for the cerebral cortex, cerebellum, brain stem, and spinal cord, but significantly higher as measured for the pituitary gland at 29 ± 3%/day (P < 0.05) (Fig. 2). The unidirectional flux of leucine into cerebral protein averaged ~0.7 µmol · 100 g¹ · min¹ and was again similar as measured for the cerebral cortex, cerebellum, and brain stem, but significantly lower as measured for the spinal cord at 0.41 ± 0.03 µmol · 100 g¹ · min¹, and significantly higher as measured for the pituitary gland at 1.40 ± 0.12 µmol · 100 g¹ · min¹ (both P < 0.05) (Fig. 2). The estimated absolute synthetic rate for the brain as a whole was ~1 g/day, constituting mainly that from the cerebral cortex at 1.0 ± 0.1 g/day, and with considerably smaller contributions from the cerebellum and brain stem at 0.12 ± 0.01 and 0.06 ± 0.01 g/day, respectively.

View larger version (20K): [in this window] [in a new window]   Fig. 2.   Fractional synthetic rate (A) and unidirectional flux of leucine into cerebral protein (B). Different letters indicate significance at P < 0.05.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The present study utilized a continuous intravenous infusion of L-[1-¹⁴C]leucine to investigate leucine metabolism in the near-term ovine fetal brain. This was accomplished with the measurement of tissue protein leucine specific activity, which reflects the unidirectional flux of leucine into cerebral protein, i.e., protein synthesis, and using Fick methodology to measure the uptake of labeled/unlabeled leucine by the brain, which reflects that utilized for protein synthesis/accretion, plus any leucine that is oxidized. A-V differences were collected in both the LV/REM and HV/NREM behavioral states to determine the relationship between cerebral leucine metabolism and behavioral state activity during brain development. The label on L-[1-¹⁴C]leucine at the carboxyl position can be measured in leucine that has been incorporated into protein, or it can be followed through the catabolic pathway, which involves the reversible transamination with -ketoglutarate to form -KIC, followed by oxidative decarboxylation to form ¹⁴CO₂, which diffuses out of the tissue (8). The measurement of tissue SA_PB leucine along with that of the precursor pool at a defined time from the start of the tracer infusion provides a means of estimating the fractional synthetic rate of brain proteins, i.e., the percentage of brain proteins newly synthesized per unit time. The accretion of proteins within the brain in turn represents the difference between the synthesis of protein and the degradation, or breakdown, of protein. The measurement of unlabeled leucine uptake by the brain here studied using Fick methodology will reflect that leucine directed toward protein accretion and any leucine that is oxidized. The infused labeled leucine will also be taken up by the brain and will equilibrate with the intracellular free leucine pool (25). However, depending on the half-life of proteins, which approximates 4 days for the adult brain (12), relative to the infusion time (6 h), there should be limited recycling of labeled leucine out of brain proteins due to degradation over the course of the study. Thus the measurement of labeled leucine uptake by the brain should mainly reflect that leucine directed toward protein synthesis as well as any that is oxidized.

Measures of cerebral protein synthesis in the present study were determined using tissue SA_IF leucine for the precursor pool and, although likely to be more representative than SA_P, may in fact underestimate the actual specific activity values of the immediate precursor amino acyl-tRNA. This could occur due to recycling of unlabeled leucine into the intracellular free leucine pool from protein degradation and tRNA species that are preferentially charged from the plasma extracellular rather than intracellular free pool (5). As such, measures of cerebral protein synthesis may be overestimated; however, this should be minimal, since the SA_IF leucine values were only slightly lower than the SA_P leucine plateau values, averaging 5.3 vs. 6.7 dpm/nmol, i.e., an ~20% difference between the maximum and minimum measures of cerebral protein synthesis when using the SA_IF and SA_P leucine values for the precursor pool, respectively. The cerebral fractional synthetic rate for protein as measured averaged ~20%/day for the brain as a whole, which is within the range of values previously reported for the ovine fetus at a slightly younger gestational age by Schaefer and Krishnamurti (24) using tyrosine rather than leucine as the tracer amino acid. The unidirectional flux of leucine into cerebral protein, as measured using the fractional synthetic rate and the concentration of leucine in protein, averaged ~0.7 µmol · 100 g¹ · min¹, which is similar to that reported by Abrams and colleagues (1) using leucine autoradiography at ~5 nmol · g¹ · min¹ for animals at 130-135 days gestation. The estimated absolute protein synthetic rate here measured for the brain as a whole using the fractional synthetic rate and the dry weight of the brain as an estimate of protein content was ~1 g/day, which is the same as that reported by Schaefer and Krishnamurti (24). This value is much higher than anticipated protein accretion at ~36 mg/day as calculated for the cerebral hemispheres, brain stem, and cerebellum of the ovine fetus between 121 and 150 days gestation from the data reported by McIntosh and colleagues (14) relating to protein content within the brain at these developmental time points. As such, the high rate of protein synthesis noted for the ovine fetal brain near term must be closely linked to a rate of protein degradation, which is almost as high as that for synthesis. Study of protein metabolism in the rat brain at 10-20 days postnatal and thus at a developmental stage comparable to that of the near-term ovine fetal brain (14) likewise reveals that protein accretion is the result of a relatively small difference between a high rate of degradation and an even higher synthesis rate, but with these rates now slowing and converging from that evident during the first week postnatal, with a resultant slowing of protein accretion (6).

In the present study, there were no regional differences in the measures of protein synthesis from the cerebral cortex, cerebellum, or brain stem, which one might expect given the heterogeneous makeup of each of these structures and random tissue sampling. However, the findings of Abrams and colleagues (1) with the detailed study of regional protein synthesis within the brain using leucine autoradiography would also indicate that the composite rates for these brain regions will be similar albeit with subregional differences with rates higher in the pineal body and hypothalamic nuclei, and lower in the white matter. Of interest, measures of protein synthesis here studied were significantly higher in the pituitary gland, with the fractional synthetic rate at ~30%/day and the unidirectional flux of leucine into protein at ~1.4 µmol · 100 g¹ · min¹, which may be attributed to the active synthesis and export of peptide hormones from this region (25). Conversely, the unidirectional flux of leucine into protein was significantly lower in the spinal cord at ~0.41 µmol · 100 g¹ · min¹, which is similar to that reported by Abrams and colleagues (1) and may reflect slower turnover of structural components within the nervous system compared with those involved in functional attributes.

Cerebral leucine uptake in the present study was determined using A-V differences for plasma leucine because the plasma transport of this amino acid across tissues per unit volume is reported to be at the same rate as blood cellular transport in sheep (9). It is also recognized that -KIC is interconvertible with leucine within the cell and can move freely between the intra- and extracellular pools. As such, the uptake of -KIC could potentially contribute to the intracellular leucine precursor pool (15). However, there was no measurable uptake of -KIC by the ovine fetal brain, indicating that the contribution of -KIC to the intracellular leucine pool was minimal. Cerebral leucine uptake as measured for those cortical areas drained by the superior sagittal sinus averaged ~60 nmol/min and, assuming that this blood flow represents ~20% of total brain flow (7), results in a cerebral leucine uptake of ~630 nmol · 100 g¹ · min¹. This value is threefold higher than that previously reported for the adult sheep brain (18) and is in keeping with a greater amino acid requirement during brain development due to protein accretion and possibly to some degree of oxidative metabolism, although glucose uptake as a metabolic fuel is sufficient to account for the oxidative needs of the ovine fetal brain (4, 11). Of interest, cerebral [¹⁴C]leucine uptake averaged ~6,800 dpm · 100 g¹ · min¹, which is equivalent to 1,000 nmol · 100 g¹ · min¹ as calculated using the plateau SA_p leucine value of 6.7 dpm/nmol. This uptake of labeled leucine is ~2-fold greater than the unidirectional flux of leucine into cerebral protein as measured here and by Abrams and colleagues (1). To the extent that this uptake measurement reflects leucine directed toward protein synthesis as well as that which is oxidized, a degree of oxidative metabolism is thereby supported.

Cerebral leucine uptake as studied was effected by behavioral state activity, with uptake measurements for both leucine and [¹⁴C]leucine significantly increased during the HV/NREM state by 1.7-fold and 2.8-fold, respectively, compared with that of the LV/REM state. Cerebral uptake values were in fact not significantly different from zero during the LV/REM state. Conversely, these values during the HV/NREM state were all significantly different from zero with [¹⁴C]leucine uptake at 10,149 dpm · 100 g¹ · min¹ or 1,514 nmol · 100 g¹ · min¹ as calculated using the plateau SA_P leucine value, ~2-fold higher than unlabeled leucine uptake at 796 nmol · 100 g¹ · min¹. It is unlikely that the increased uptake of leucine during HV/NREM is attributable to changes in the transport of leucine via the blood-brain barrier (BBB) neutral amino acid transporter, as leucine has a high BBB permeability (17) and as there is no change in the arterial plasma concentration of leucine (as measured here) or the other neutral amino acids (23) between the two behavioral states. Assuming that [¹⁴C]leucine uptake reflects mainly protein synthesis whereas unlabeled leucine uptake reflects protein synthesis plus degradation, i.e., accretion as previously discussed, then both synthesis and degradation must also be increased during the HV/NREM state compared with the LV/REM state. However, the increase in synthesis will also be dependent on the extent of any leucine oxidation by the brain. For example, if 50% of the cerebral leucine uptake were to be oxidatively metabolized, i.e., ~400 nmol · 100 g¹ · min¹, with the remainder directed toward protein accretion, then that attributed to synthesis vs. degradation would measure ~1,115 and 715 nmol · 100 g¹ · min¹, respectively. While this degree of oxidative metabolism would account for <2 µmol of oxygen · 100 g¹ · min¹, it is of interest that Chao and colleagues (4) found measured oxygen uptake by the ovine fetal brain during the HV/NREM state to be somewhat higher than that accounted for by glucose oxidation alone, 127 vs. 114 µmol · 100 g¹ · min¹, with the probability of additional metabolic substrates including that of lactate at this time. These findings now add to those in adult animals where higher rates of cerebral protein synthesis have been shown to be positively correlated with the occurrence of NREM sleep (16, 19) and further support a role for behavioral state activity in the brain's growth and development. Together, these studies would also indicate that the decrease in the brain's metabolic demand during NREM sleep as seen in the ovine fetus (21) and in other species postnatally, including humans (13), does not result from a decrease in biosynthetic activity and may, in fact, favor the synthesis of new proteins. This would support the restorative theory of sleep (2) whereby energy conservation during NREM sleep favors the anabolic restoration of tissue components.

Perspectives