Pentobarbital Enhances Cyclic Adenosine Monophosphate Production in the Brain by Effects on Neurons but Not Glia

Chemicals used in the current study were obtained from the following sources: isoproterenol, pentobarbital, cytosine-beta-D-arabino-furanoside, 3-isobutyl-1-methylxanthine (IBMX), Hank's balanced salt solution, Dulbecco's modified Eagle's medium with 25 mM hydroxyethylpiperazineethane sulfonic acid (HEPES) (DMEM-H), poly-D-lysine, and trypsin from Sigma (St. Louis, MO); Ham's F-12 and penicillin/streptomycin from JRH Biosciences (Lenexa, KS); and fetal calf serum from Biocell Laboratories (Rancho Dominguez, CA).

Cultures of cerebral cortical neurons and glia were prepared from 17-day gestational Sprague-Dawley rats using protocols approved by the Institutional Animal Care and Use Committee of the University of Pennsylvania, as described previously. [21,22]Whole cerebral hemispheres were collected in Hank's balanced salt solution from fetuses of 17-day gestational Sprague-Dawley rats. They were washed twice in phosphate buffered saline, then digested using trypsin (5 mg *symbol* ml sup -1) in DMEM-H, washed three times in Hank's balanced salt solution, then triturated with a "fire-polished" Pasteur pipette into DMEM-H supplemented with 10% Ham's F12 media, 10% heat inactivated fetal calf serum, 100 micro gram *symbol* ml sup -1 streptomycin, and 100 units *symbol* ml sup -1 penicillin. Cells were plated at a density of 4 x 10 sup 5 cells per 1.2-cm diameter tissue culture well that had been pretreated with 10 micro gram *symbol* ml sup -1 poly-D-lysine. Cultures were grown in a 37 degrees C humidified incubator containing 5% CO sub 2 in air, and were fed approximately every 2 or 3 days with fresh medium. On day 5, neuronal cultures were treated with 10 micro gram *symbol* ml sup -1 cytosine-beta-D-arabino-furanoside for 24 h to stop nonneuronal cell proliferation. Glial cell cultures were prepared by dispersing the mixed cultures by digestion with trypsin at day 5 and replating them at a density of 8 x 10⁴cells per 1.2-cm diameter well. These cultures were also fed every 2 or 3 days but were not treated with cytosine-beta-D-arabino-furanoside. This procedure resulted in preparations that were essentially devoid of neuronal cells. For both types of cultures, feedings after day 5 were with the aforementioned media, with the exception of Ham's F-12, which was removed to exclude glutamate from the feeding. Cultures were used for experiments after 12-16 days in culture.

The brains of female Sprague-Dawley rats (mothers of fetuses used for preparations of primary cultures) were removed after anesthesia with carbon dioxide and death by decapitation. The brains were rinsed in ice-cold DMEM-H that was previously equilibrated with 95% O²/5% CO sub 2 and the cerebral cortex was dissected and prepared as 300 micro meter cross-cut slices using a Mc-Ilwain tissue chopper (Mickle, Gomshall, Surrey, England) as has been described by others. [22]Slices were dispersed in fresh, ice-cold DMEM-H that was previously equilibrated with 95% O²/5% CO².

Cyclic AMP accumulation in cultured cells was determined in cells preloaded with radiolabeled adenine as described by Atkinson and Minneman. [22]Neuronal cultures were incubated with 0.5 micro Ci [2,8-sup 3 Hydrogen]adenine in 0.5 ml growth media for 120 min at 37 degrees C, then washed twice with serum-free DMEM-H. When included, IBMX (0.5 mM) and pentobarbital were added to the cultures in a final reaction volume of 0.5 ml for a 20-min preincubation period. Cyclic AMP accumulation was initiated by the addition of control vehicle or medium containing isoproterenol, clonidine, or forskolin. The reaction was stopped after 10 min by the addition of trichloroacetic acid for a final concentration of 5% in a final volume of 1.0 ml. The contents of the wells were collected into glass tubes, samples were centrifuged at 500g for 10 min after which the [sup 3 Hydrogen]cAMP and [sup 3 Hydrogen]adenosine triphosphate were isolated by sequential chromatography over Dowex and alumina columns as described previously. [17]Cyclic AMP accumulation was calculated as the percentage of [sup 3 Hydrogen]adenosine triphosphate initially present in the cultures that was converted to [sup 3 Hydrogen]cAMP. Cyclic AMP production was linear up to at least 20 min in this model (data not shown).

Measurement of cAMP accumulation in slices was determined in cells preloaded with radiolabeled adenine as described by Atkinson and Minneman. [22]Buffer was decanted from the slices prepared as described earlier and the tissue was incubated in fresh DMEM-H that was previously equilibrated with 95% O²/5% CO²and containing 1.0 micro Ci [2,8-sup 3 Hydrogen]adenine per milliliter for 120 min at 37 degrees C while being bubbled with 95% O²/5% CO², then washed twice with DMEM-H. Aliquots of slices containing approximately 25 mg wet weight of brain tissue were transferred into reaction tubes containing 0.5 mM IBMX and pentobarbital in a final reaction volume of 0.5 ml for a 20-min preincubation period. The reactions were performed and the samples were analyzed as described earlier for cultured cells.

Data from each type of experiment were analyzed using a paired Student's t test, one-sample t test with correction for multiple comparisons, or analysis of variance with post hoc corrections using Dunnett's test, as indicated in the Figure 1legends. A value of P < 0.05 was considered significant.

Before beginning experiments with pentobarbital, the function of receptor-modulated cAMP production by both stimulatory and inhibitory G-protein-linked pathways was tested in cultured neurons. The beta-adrenergic agonist, isoproterenol, significantly stimulated cAMP production in cultures of neurons, with a maximal increase of approximately tenfold over basal at concentrations of 0.1 and 1 micro Meter (Figure 1(A)). The alpha²-adrenergic agonist, clonidine, inhibited forskolin-stimulated cAMP production by approximately 30% at a concentration of 1 micro Meter (Figure 1(B)).

The effect of pentobarbital on cAMP accumulation was tested in these cultures. Isoproterenol (1 micro Meter), as demonstrated earlier, and forskolin (25 micro Meter) significantly stimulated cAMP production compared to basal cAMP production. Pentobarbital enhanced isoproterenol- and forskolin-stimulated cAMP production in these preparations with a biphasic dose-response curve (Figure 2). Cyclic AMP production was enhanced by pentobarbital in a dose-dependent fashion up to a concentration of 250 micro Meter; This concentration increased cAMP production by 40-50% relative to that in controls without pentobarbital. At 500 micro Meter pentobarbital, the magnitude of the enhancement was less. Similar experiments performed in the presence of isoproterenol but in the absence of IBMX showed a similar response but the signal was smaller (data not shown).

To determine whether the effect of pentobarbital required neurons, experiments were performed with cultures of glia. Isoproterenol significantly stimulated cAMP production, compared to basal, at concentrations of 5 nM and 1 micro Meter (data not shown). Cyclic AMP production stimulated by 1 micro Meter isoproterenol was greater in these preparations than in cultures of neurons. Pentobarbital had no effect on cAMP production in the presence of either concentration of isoproterenol (Figure 3).

Experiments were performed with slices prepared from cerebral cortexes of adult Sprague-Dawley rats to determine whether or not pentobarbital enhanced cAMP production in neurons that had developed in vivo as well as those that had developed in culture. Cyclic AMP production in these preparations was significantly stimulated by isoproterenol (Figure 4). Pentobarbital enhanced isoproterenol-stimulated cAMP production by approximately 30% at concentrations of 62.5, 125, and 250 micro Meter, and by nearly 100% at a concentration of 500 micro Meter (Figure 4). Pentobarbital had no effect on cAMP production in the absence of isoproterenol (basal, Figure 4).

Many effects of barbiturates on cellular signaling have been described, including effects on gamma-aminobutyric acid (GABA) and excitatory amino acid signaling as well as effects on other ion channels and enzymes. The reported effects on cAMP production have been variable. Dan'ura et al. [18-20]reported that barbiturates inhibit cAMP production stimulated by forskolin, 5'-guanylylimidodiphosphate (Gpp[NH]p), or NaF- AlCl³in membranes prepared from rat brain. However, we previously reported the enhancement of isoproterenol-stimulated cAMP production in intact S49 mouse lymphoma cells. [17]We sought to determine whether the apparent differences between our findings and those reported by Dan'ura et al. [18-20]were owing to the use of different cell types or the use of preparations of whole cells instead of preparations of membranes. The current report describes experiments performed to determine the effects of pentobarbital on cAMP production in preparations containing intact cells from brain.

To be certain that cultured cells were an appropriate model to study the regulation of receptor-modulated cAMP production, we first demonstrated that both the stimulatory and inhibitory G-protein-linked pathways were functional in these preparations. The cultures responded to both the stimulatory beta-adrenergic receptor agonist, isoproterenol, and the inhibitory alpha²-adrenergic receptor agonist, clonidine, as expected. These results demonstrate that these preparations are a useful model for the study of the regulation of cAMP production.

Pentobarbital enhanced isoproterenol-stimulated cAMP production in cultures of neurons at concentrations up to 500 micro Meter, which spans the range from concentrations that are clinically relevant to those that are toxic. This enhancement reflects an increase in production, not a decrease in breakdown, because the effect was observed when IBMX was included in samples to inhibit phosphodiesterase activity, but IBMX was not required for the effect. Although these cultures predominantly contain neurons, some glia are present in these preparations to support the growth of the neurons. Based only on the data from these preparations, we could not rule out that the effect of pentobarbital is occurring in the glia, and because glia do contribute to the regulation of excitatory neurotransmission, [23]it is important to make this distinction. When cultures are dispersed and replated, essentially only glia grow, neurons do not survive this manipulation, as we [21]and others [22]have described. When cAMP production stimulated by 5 nM or 1 micro Meter isoproterenol was measured in these preparations, pentobarbital had no effect. The production of cAMP by glia stimulated by 1 micro Meter isoproterenol was much greater than that observed in preparations of neurons stimulated by 1 micro Meter isoproterenol but was similar to that in neurons stimulated by 25 micro Meter forskolin. These data suggest that the lack of effect of pentobarbital on glial preparations is not due to the absolute level of activity observed in the absence of pentobarbital, but rather that neurons are required for the effect. The lack of an effect of pentobarbital on cAMP production in unstimulated neurons (basal) is compatible with the hypothesis that pentobarbital influences a component of the G-protein-linked stimulatory pathway and does not stimulate the catalytic moiety directly, similar to what we concluded previously for S49 wild type mouse lymphoma cells. [17].

Additional studies were performed with freshly prepared slices from rat brain to determine whether or not pentobarbital enhanced cAMP production in neurons that developed in vivo, as was observed with cultured neurons that had developed in vitro. The slices were found to contain functionally linked beta-adrenergic receptors and adenylyl cyclase. Isoproterenol (1 micro Meter) significantly stimulated cAMP accumulation in these preparations compared to basal. The magnitude of the stimulation is comparable to that previously reported by others in this preparation. [22]Pentobarbital significantly enhanced cAMP accumulation in these preparations, but only in the presence of isoproterenol. This finding is the same as that observed in neuronal cultures, and further supports the hypothesis that pentobarbital does not directly stimulate adenylyl cyclase. Furthermore, these results strengthen the validity of using cultured neuronal preparations for studies of these pathways and suggest that pentobarbital enhances cAMP production in the brain in situ.

In both cultured neurons and brain slices the maximum effect of pentobarbital was a 30-50% increase in cAMP production. This measurement is an overall average increase in production in the entire sample. Because the effect is not seen in all cells, as demonstrated by the experiments with glial cultures, it is likely that the magnitude of the effect of pentobarbital is much greater in some subpopulation of cells, and therefore, the effect on cellular function would be expected to be quite significant in this group of cells.

Many investigators have demonstrated that cAMP can play a role in the depression of excitatory neurotransmission or enhancement of inhibitory neurotransmission. Activation of beta-adrenergic receptors increases the frequency of inhibitory synaptic currents in rat cerebellar stellate and Purkinje cells, by a mechanism that probably involves intracellular cAMP. [24]Analogs of cAMP inhibit spontaneous firing of Purkinje neurons. [25]gamma-Aminobutyric acid-mediated currents are potentiated by cAMP-dependent protein kinase in cerebellar Purkinje cells. [26,27]gamma-Aminobutyric acid-mediated currents are potentiated by cAMP in cerebrocortical, cerebellar, and hypothalamic neurons. [28]Activation of beta-adrenergic receptors leads to a reduced frequency of spontaneous excitatory spikes and a hyperpolarization of Purkinje cells. [29]Phosphorylation of inhibitory glycine channels expressed in xenopus oocytes by cAMP-dependent protein kinase results in an enhancement of glycine-evoked current. [30,31]Isoproterenol enhanced the inhibitory responses to GABA, mimicking the effect of benzodiazepines in neurons from the lateral hypothalamus. [32]Norepinephrine activation of beta-adrenergic receptors enhanced GABA-activated inhibitory synaptic mechanisms in the cerebellum and cerebral cortex. [33]Enhancement of neurotransmitter-stimulated cAMP production by pentobarbital may, therefore, play a role in the depression of excitatory neurotransmission or enhancement of inhibitory neurotransmission by one or more of these mechanisms.

However, several other investigators have demonstrated an excitatory effect of cAMP on neuronal activity. Cyclic AMP or cAMP-dependent kinase enhanced responses to excitatory amino acids in several experimental models. [29,34-39]Cyclic AMP also has been shown to decrease the inhibitory response to GABA. [40-44]It also appears that in one specific brain region, the locus ceruleus, inhibition of adenylyl cyclase mediates the hypnotic effect of an alpha²-adrenergic receptor agonist. [45]On the basis of these observations, one might speculate that the cAMP response to pentobarbital would be 'anti-anesthetic' and that the enhanced production of cAMP could be a negative feedback response by the neuron to limit the inhibitory influence of the anesthetic.

Cyclic AMP, therefore, has multiple effects on neuronal and brain function, and some of the differences seen in various models could be attributable to different effects on different brain regions or subtypes of neurons. More detailed studies of the effects of pentobarbital on specific regions of the brain or on different subtypes of neurons will be important to understand the full implications of enhanced cAMP production on the function of the brain in intact animals.

The authors thank Steven Irvine and Judith O'Grady for their technical assistance.