Main

Glutamate is the principal excitatory amino acid in the mammalian CNS. It mediates excitatory signaling and synaptic transmission and appears to influence neural growth in the developing brain (1–3). Cell surface receptors through which glutamate exerts its effects may be generally classified into three families:1) NMDA, 2) AMPA/kainate, and 3) metabotropic receptor (4). These receptors tend to increase cerebral glucose utilization (5). Studies suggest that antagonists of both the NMDA and the AMPA receptor may protect against excitotoxic cellular damage, but their specific role remains unclear (6, 7). Although NMDA receptors have been studied to a greater extent, the AMPA receptors are believed to be the major excitatory neurotransmission pathway of the brain. More AMPA receptors are found in the cortex and other forebrain regions than in the hindbrain region (1, 8). It has been shown that AMPA antagonists reduce local glucose utilization in the cerebral cortex, hippocampus, and thalamus of the adult rat (9).

Although a principal excitatory amino acid in the developing brain, the mechanism and conditions under which glutamate and its receptors influence the control of regional cerebral blood flow and oxygen consumption in newborns have not been fully described. Newborns may be more sensitive to the toxic effects of excitatory amino acids such as glutamate (3), and a number of studies have demonstrated elevated glutamate concentrations in association with neuronal damage (10, 11). The relative importance of AMPA receptors may be greater in parts of the newborn brain than of adult brain (2, 12). We previously found that during hypoxia, but not normoxia, cerebral O2 consumption is influenced by an NMDA receptor-mediated mechanism (13). The importance of AMPA receptors in the control of cerebral metabolism in the newborn brain has not been determined. We also wished to determine whether the importance of AMPA receptors in the control of cerebral metabolism changed under different circumstances.

In this study, we hypothesized that non-NMDA glutamate receptors are important controllers of cerebral oxygen supply and consumption balance in newborns and that the relative importance of AMPA receptors would increase with cerebral hypoxia. To evaluate this hypothesis, we subjected newborn piglets to normoxia and hypoxia in the presence and absence of AMPA receptor blockade. We then determined regional cerebral O2 supply/consumption balance. AMPA receptor blockade limited the hypoxia-induced increase in cortical and hypothalamic O2 consumption and blood flow, thereby maintaining cerebral O2 supply and consumption balance in these regions during hypoxic stress.

METHODS

Twenty-seven newborn pigs (2–7 d old) of either sex weighing an average of 1.96 kg were used in this study. All animals were anesthetized initially with ketamine 30 mg/kg and xylazine 10 mg as a single i.m. dose. Additional i.m. doses of 10 mg/kg of ketamine were administered, up to a total of 60 mg/kg, to maintain anesthesia until a femoral venous catheter could be inserted. Anesthesia was then maintained with 40 mg/kg of α-chloralose i.v. in four divided doses as needed. A femoral artery catheter was then inserted for monitoring of heart rate, blood pressure [R411 recorder (Beckman Instruments, Inc., Fullerton, CA) and P23Db transducer (Statham, Cambridge, MA)], and arterial blood gases (ABL 330, Radiometer, Copenhagen, Denmark). Body temperature was maintained at 37°C with a rectal thermistor, heating pad, and heat lamp. Animals were then tracheotomized, and controlled ventilation was instituted by using a Harvard small animal respirator to maintain PaCO2 between 30 and 40 mm Hg. An end-expiratory pressure of 5 cm H2O was used to maintain lung volume during expiration. The animals were then allowed to stabilize for half an hour, at which time control values for heart rate, blood pressure, PaCO2, PaO2, pH, and Hb were obtained. The study was approved by the Robert Wood Johnson Medical School Institutional Animal Care and Use Committee, and animals were handled according to the guidelines established by the National Institutes of Health for care and use of laboratory animals.

The animals were divided into four experimental groups. To block the effects of excitatory amino acids in one-half of the animals, a single 10 mg/kg dose of GYKI 52466 dissolved in distilled water was administered i.v. Normoxia was maintained in one-half of each of these two groups, whereas hypoxia was induced in the other halves of each group by subjecting the animals to a mixture of 8% O2 in N2 through the respirator for 15 min (14). In the group subjected to both hypoxia and GYKI 52466 treatments, GYKI 52466 was administered 5 min before inducing hypoxia, and blood flow measurements were completed between 15 and 20 min after GYKI 52466 administration. Values for heart rate, blood pressure, PaCO2, PaO2, pH, and Hb were repeated 10–15 min after treatment with GYKI 52466, hypoxic gas, or both. The normoxic group not receiving GYKI 52466 formed the control group. Regional cerebral blood flow, cerebral oxygen extraction, and consumption were then determined in all four groups.

Measurement of cerebral blood flow.

For the measurement of cerebral blood flow, 50 μCi of [14C]iodoantipyrine (Amersham, Arlington Heights, IL) were infused i.v. with an infusion pump (Sage Instruments, Cambridge, MA). At the time of entry of the isotope into the venous circulation, the arterial catheter was cut to a length of 15–20 mm to minimize smearing in the sampling catheter. Twenty timed samples were withdrawn from the arterial catheter and collected approximately every 3 s in capillary tubes over a period of 60 s. The animal was decapitated at the moment the last sample was obtained. The head was quickly cut in the midsagittal plane and frozen in liquid nitrogen for later analysis. While frozen, samples of brain tissue were obtained from three regions: cortex, hypothalamus, and pons. Blood and tissue samples were then placed in a tissue solubilizer (Soluene, Packard, Downers Grove, IL) and 24 h later, were placed in a counting fluid (Dimiscint, National Diagnostic, Manville, NJ) and agitated. These samples were counted on a Beckman LS-230 liquid scintillation counter. Carbon tetrachloride was used to prepare quench curves, and the isotope counts were adjusted for color and quench correction.

Regional blood flow determinations were made with a computer program based on the following equation:EQUATION 1

where Ci(T) equals the concentration of the [14C]iodoantipyrine at the time of decapitation, λ equals the tissue:blood partition coefficient, CA is the arterial concentration of the tracer, and t equals time. K is the constant that represents the rate of blood flow in the tissue and is defined as follows: K = mF/λW, where m is the constant related to diffusion equilibrium between blood and tissue during passage from the arterial to the venous end of the capillary, and F/W equals the blood flow per unit mass of tissue. The λ value of 0.8 calculated by Sakurada et al. (15) was used. Blood flow is expressed in mL/min/100 g.

Measurement of oxygen consumption.

To permit determination of regional arterial and venous O2 saturation, the head was frozen and stored in liquid nitrogen when the animal was decapitated. The same brain regions used for blood flow were isolated and examined. Preparation of samples was performed at −20°C. The frozen brain was cut into wafers with a bandsaw. The details and validation of microspectrophotometric analysis have been described previously (16, 17). Briefly, 20-μm-thick frozen tissue sections were cut on a rotary microtome in a −30°C cold box that was flushed with nitrogen. Each section was then transferred to a precooled glass slide, covered with degassed silicone oil and a coverslip, and rapidly transferred to the nitrogen-flushed cold stage of the microspectrophotometer. Small arteries and veins, identified by the presence or absence of the muscular media (20–75 μm in diameter), were examined, and the O2 saturation of blood contained within the vessels was determined with measurements of optical densities at 560, 568, and 523 nm. This three-wavelength method corrects for the light scattering in the frozen blood. Microspectrophotometric measurements were obtained from a total of eight veins and five arteries per region. The size of the measuring spot was 8 μm. The volume of tissue examined in each region was 0.5 cm3. Only vessels in the transverse section were studied so that the path of light traversed only the blood. Regional O2 extraction in the various areas of the brain was determined by multiplying the regional average arterial-venous O2 difference by the arterial Hb concentration times 1.36 mL O2/g, the maximal binding capacity of Hb for O2/g. Using the Fick principle, we calculated the O2 consumption for each region as the product of average regional blood flow and O2 extraction (18).

Protocol.

Group 1 (NORM) animals (n= 7) served as the control group. They were catheterized, ventilated, and had baseline hemodynamic and blood gas measurements performed. Neither hypoxia nor GYKI 52466 treatment was instituted before the measurement of cerebral blood flow. Group 2 (GYKI) animals (n= 7) were treated with GYKI 52466 under normoxic conditions to assess the effect of AMPA receptor blockade at rest. Group 3 (HYP) animals (n= 6) were subjected to hypoxia to assess the changes induced by hypoxia alone. Group 4 (HYP+GYKI) was pretreated with GYKI 52466 before the induction of hypoxia (n= 7) to assess the effect of AMPA receptor blockade on any changes due to hypoxia.

Data analysis.

A repeated measure analysis of variance was used to compare differences in cerebral blood flow, oxygen consumption, oxygen extraction, and venous oxygen saturation between regions. The statistical significance of differences was determined by Duncan's procedure. A p value of <0.05 was accepted as significant. Data are expressed as means ± SEM.

RESULTS

Hemodynamic and blood gas data for all four groups are shown in Table 1. There were no statistically significant differences in heart rate, PaCO2, pH, or Hb among the normoxia (NORM), GYKI 52466 (GYKI), hypoxia (HYP) or both GYKI 52466 plus hypoxia (HYP+GYKI) groups, before treatment. Systolic blood pressure in the GYKI group decreased from 88 ± 4 mm Hg before treatment to 75 ± 3 mm Hg after treatment. There were also significant decrements in systolic, diastolic. and mean arterial blood pressure after treatment in the HYP+GYKI group compared with the pretreatment values. There were no differences in PaO2, PaCO2, or pH among the groups before treatment. In the groups in which hypoxia was induced, the PaO2 was significantly lowered, 89 ± 3 to 32 ± 2 mm Hg in the HYP group and 92 ± 4 to 31 ± 4 mm Hg in the HYP+GYKI group. Hypoxia did not alter either PaCO2 or arterial pH.

Table 1 Hemodynamic and blood gas data. * Significantly different from pretreatment values. † Significantly different from both nonhypoxic groups. ‡ Significantly different from both GYKI-treated groups.
Full size table

rCBF values from the cortex, hypothalamus, and pons for all four groups are presented in Figure 1. There were no differences in rCBF among the three regions in the control group; neither were there significant differences between the GYKI and the control groups. In the GYKI group, rCBF was higher in the pons (53 ± 8 mL/min/100 g) than in the cortex (34 ± 6). The HYP group showed significant increases in rCBF in the cortex and hypothalamus of 97 ± 16 and 99 ± 17, respectively, compared with the same regions in the control group: 50 ± 10 (cortex) and 41 ± 9 (hypothalamus). The increase in rCBF that occurred in the cortex and hypothalamus in the HYP group was not observed in the HYP+GYKI group, but rather, blood flow was similar to that of controls, with values of 68 ± 14 and 73 ± 12, respectively. The HYP+GYKI group also had a higher rCBF in the pons (92 ± 10) than in the cortex (68 ± 14) of that group. There was no statistically significant difference in pontine rCBF in any of the three treatment groups, compared with that of the control group.

Figure 1
figure 1

Effects of GYKI 52466 (GYKI) on regional cerebral blood flow under normoxic (NORM) and hypoxic (HYP) conditions. Values are presented for the cortex (CORT), hypothalamus (HYPO) and pons (PONS). *Significantly different from the same NORM region. +Significantly different from the CORT in the same group.

Full size image

Figure 2 shows the results for cerebral oxygen extraction. Although a slight decrease in oxygen extraction was observed in all regions of both hypoxic groups, compared with control, this difference was not statistically significant. There was, however, a difference in oxygen extraction in all regions of the HYP group compared with the GYKI group. Similarly, there was a significant decrease in regional oxygen extraction in the HYP+GYKI group compared with the GYKI group, except that values in the pons did not reach statistical significance, whereas those in the cortex and hypothalamus were significantly different.

Figure 2
figure 2

Effects of GYKI 52466 (GYKI) on regional cerebral oxygen extraction under normoxic (NORM) and hypoxic (HYP) conditions. Values are presented for the cortex (CORT), hypothalamus (HYPO), and pons (PONS). #Significantly different from the same GYKI region.

Full size image

Figure 3 represents the regional cerebral oxygen consumption in the three regions for each of the four experimental groups. There were no regional differences in cerebral oxygen consumption in the NORM group. During normoxia (in the GYKI group), GYKI 52466 had no significant effect on cerebral oxygen consumption in any region compared with the controls. However, we observed a difference in pontine compared with cortical O2 consumption in the GYKI group: 1.8 ± 0.2 mL O2/min/100 g for cortex and 2.8 ± 0.5 for pons. There was a significant increase from control in cerebral oxygen consumption in the cortex and hypothalamus in the HYP group (2.4 ± 0.5 to 3.7 ± 0.4 mL O2/min/100 g in the cortex and 2.5 ± 1 to 3.8 ± 0.5 mL O2/min/100 g in the hypothalamus), but not in the pons (3.0 ± 1 to 3.7 ± 0.7 mL O2/min/100 g). When GYKI 52466 was administered before hypoxia (HYP+GYKI), regional cerebral oxygen consumption remained similar to control: specifically, we observed no increases in O2 consumption in the cortex and hypothalamus, as occurred in the HYP group. Oxygen consumption was significantly higher in the pons compared with the cortex and hypothalamus in the HYP+GYKI group, with values of 2.5 ± 0.4 for the cortex, 3.0 ± 0.5 for the hypothalamus, and 3.9 ± 0.4 for the pons.

Figure 3
figure 3

Effects of GYKI 52466 (GYKI) on regional cerebral oxygen consumption under normoxic (NORM) and hypoxic (HYP) conditions. Values are presented for the cortex (CORT), hypothalamus (HYPO) , and pons (PONS). *Significantly different from the same NORM region. +Significantly different from the CORT in the same group. • Significantly different from the HYPO in the same group.

Full size image

DISCUSSION

The main findings in this study suggest that 1) AMPA receptors are important in the control of cerebral oxygen consumption and blood flow in the cortex and hypothalamus during hypoxia, and 2) AMPA receptors appear to exert little effect on flow and metabolism in the unstressed newborn animal. When the animals were subjected to hypoxia, cerebral O2 consumption and blood flow were increased in the cortex and hypothalamus. When pretreated with the AMPA receptor antagonist GYKI 52466 before hypoxia, the hypoxia-induced increases in O2 consumption and blood flow were abolished in those same regions. This suggests that much of the hypoxia-induced cerebral blood flow increase was driven by an increase in oxygen consumption from AMPA receptor stimulation.

AMPA receptor antagonism alone had no effect on cerebral blood flow, oxygen extraction, or oxygen consumption in any of three brain regions, compared with controls during normoxia. Blockade of NMDA receptors also appears to have little effect on cerebral blood flow or metabolism in the unstressed newborn pig (13). It may be that not all AMPA receptors are not fully functional in the newborn (2). However, in the adult rat, NMDA receptors may have effects on basal cerebral metabolism (9, 19). This could be the cause of the cerebral vasodilation seen with glutamate receptor activation (20). Glutamate can affect cerebral blood flow in newborn at least by NMDA activation (21). The lack of change in cerebral blood flow after AMPA receptor blockade could be related to the lack of glutamate receptors on cerebral blood vessels (22), the lack of importance of AMPA receptors at this developmental stage, or the lack of metabolic effects. The data suggest that in control normoxic newborn pigs, AMPA does not control cerebral vascular tone or metabolism at rest. Another explanation for this could be related to depression of cerebral metabolism by the anesthetic effects of α-chloralose.

It has been shown that cerebral blood flow increases during hypoxia in both adults and newborns (14, 23). In adult animals, this vasodilation is not usually accompanied by major changes in cerebral metabolism (14, 24). Previous work in our laboratory has demonstrated a role for glutamate via NMDA receptors in the control of cerebral O2 supply and consumption during hypoxia in newborns (13). This study also demonstrated that the increase in cerebral blood flow during hypoxia was accompanied by an increase in cerebral oxygen consumption in the newborn. The specific influence of non-NMDA receptors on cerebral metabolism and, particularly, on O2 supply/consumption balance during hypoxia had not been characterized.

Glutamate released into the extracellular space during hypoxic-ischemic stress is thought to play an integral role in cell death by inducing the intracellular influx of ions such as Ca2+ and Na+ causing neuronal depolarization (10, 11). Glutamate interacts with a number of distinct receptor subtypes, including the NMDA, AMPA, and kainate receptors, and the possibility has been suggested that antagonists of these receptors are cytoprotective (6, 7). The importance of AMPA receptors in cerebroprotection during transient ischemia has been described previously. Le Peillet et al. (25) demonstrated protection against cortical cell loss in a transient global ischemia rat model using GYKI 52466, whereas Buchan et al. (26) showed similar results in focal ischemia. In both instances, the mechanisms of action were unclear.

AMPA receptors may be important in the newborn in terms of neuronal control of respiration (27). These effects on respiration may be enhanced during hypoxia. Hollmann et al. (28) described the four polypeptide subunits, i.e., GluR1, GluR2, GluR3, and GluR4, that constitute the family of subunits with a high affinity for AMPA, and they suggested that the degree of Ca2+ permeability depends on subunit composition of the glutamate receptor channels. These receptors develop at different rates in the newborn (2, 12). Pellegrini-Giampetro et al. (29) suggested that a switch in glutamate receptor subunit expression in certain brain regions after global ischemia could be a potential mechanism for postischemic increase in Ca2+ permeability through channels linked to AMPA receptors. This could have important metabolic consequences during hypoxia. Our data suggest a possible mechanism for cerebroprotection, in that the hypoxia-induced increase in cerebral blood flow appears to be driven by the increase in cerebral oxygen consumption via the effects of glutamate on AMPA receptors. AMPA receptors may be important regulators of cerebral metabolism and activity under some circumstances (27, 30).

Our data demonstrated an increase in cortical and hypothalamic O2 consumption and regional blood flow induced by hypoxia which was abolished when GYKI 52466 was administered before the institution of hypoxia. This suggests that part of the response to hypoxia in these regions was due to glutamate release acting via AMPA receptors. Part of this metabolic effect may be related to Ca2+ permeability changes (28). Previously, NMDA receptors were also shown to be important in this hypoxia-induced increase in cerebral oxygen consumption (13). These data support previous findings in our laboratory and by others that there is tight coupling between O2 consumption and blood flow. We did not observe the same response in the pons, and this may reflect differences in regional AMPA receptor distribution and may be related to the lower number of AMPA receptors in the hindbrain compared with the forebrain (1, 8).

In summary, our study demonstrated that when newborn piglets were subjected to hypoxia, cerebral blood flow and oxygen consumption in the cortex and hypothalamus increased. These effects were abolished with the administration of the AMPA receptor antagonist GYKI 52466. This suggests that, at least in the newborn piglet, excitatory amino acids are important in cerebral metabolism, where under hypoxic stress they appear to help maintain the balance between cerebral O2 supply and consumption in certain brain regions. Specifically, this study suggests that AMPA receptors are important in the cortex and hypothalamus during hypoxia. The differences in regional AMPA receptor distribution, activity, developmental stage, subtypes, and regulation in the newborn are not fully understood. However, AMPA receptors are important in the increased cortical and hypothalamic O2 consumption caused by hypoxia.