Life Sciences, Printed
Vol.
50, pp.
1701-1709
Pergamon
Press
in t h e U S A
EFFECTS OF PENTOBARBITAL TOLERANCE AND DEPENDENCE ON CONVULSANT AND GABA A RECEPTOR ANTAGONIST BINDING
Paul A. Saunders 1, Toshiyuki Kimura, Teiji Miyaoka, I.K. Ho Department of Pharmacology and Toxicology University of Mississippi Medical Center Jackson, Mississippi 39216-4505 (Received
in final
form March 24,
1992)
Summarv
Experiments were performed which examined the effects of pentobarbital tolerance and dependence on GABAA receptor antagonist binding. In rats implanted with pentobarbital pellets for 7 days, followed by 24 hours of withdrawal, there was a significant decrease in the latency of TBPS-induced seizures and an increase in [35S]TBPS binding in the frontal cortex. The pent(2barbital tolerant rats had a significant increase in the low affinity KD of [°H]SR95531 binding. Removal of the pellets for 24 hours caused a reversal of the effect on the low affinity K D and caused a decrease in the number of low affinity binding sites. In vitro addition of pentobarbital to bin~ling assays produced a decrease in the number of high affinity [oH]SR95531 binding sites without changing low affinity binding. In the cerebellum, the binding in none of the treatment groups was significantly different from placebo. These observations suggest that pentobarbital tolerance and withdrawal cause changes in the properties of the GABAA receptor antagonist binding site which are different from those caused by in vitro exposure to the drug. Although barbiturates have been used as sedative-hypnotics and anticonvulsants for almost a century, relatively little is known about the mechanisms underlying the development of barbiturate tolerance and withdrawal in the central nervous system. Despite studies which show that barbiturates have effects on the GABAA receptor chloride channel complex in vivo (1) and in vitro (2), the influence of barbiturate tolerance and dependence on the properties of the GABAA receptor complex has not been consistently reported. Increases (3), decreases (4), or no changes in GABA receptor binding have been reported (5). Decreases in both benzodiazepine number (6) and affinity (7) have also been seen. The discrepancies in the literature may at least in part be due to differences in animal treatment protocols or binding assay procedures. Our laboratory has developed a pellet implantation model of barbiturate tolerance which produces both CNS tolerance to pentobarbital and increased sensitivity to convulsants in barbiturate withdrawal (8). Using this model, we have shown that pentobarbital tolerance causes an increase in the K D of the low affinity GABA receptor site and a decrease in benzodiazepine binding (9). 1present address: Laboratory of Neurophysiology, NINDS, National Institutes of Health, Building 36, Room 2C02, Bethesda, Maryland 20892 Copyright
0024-3205/92 $5.00 + .00 © 1992 Pergamon Press Ltd All rights
reserved.
1702
Barbiturates and [3H]SR95531 Binding
Vol. 50, No. 22, 1992
thdrawal was shown to cause an increase in the number of binding sites for S]TBPS (10), a convulsant which blocks the GABA-activated chloride channel (11). nce the binding of some receptor antagonists has been observed to be thermodynamically different from that of agonists (12), binding studies using antagonists may reveal properties of receptors not detected by other methods. Recently, a new GABA receptor antagonist, SR95531, became commercially available. This compound is potentially an improvement over bicuculline, for the binding of ~H]bicuculline methiodide is best fit to a one-site model (13), while the binding of [°H]SR95531 is best fit to a two-site model (14). A two-site model for a GABAA antagonist binding is intriguing, for the allosteric interactions of the GABA receptor agonist conformation are dependent on the functional integrity of the low affinity site (15). The following experiments were performed to examine the effects of pentobarbital tolerance, dependence, and in vitro pentobarbital on [°H]SR95531 binding. Methods Animals and materials Male Sprague Dawley rats weighing 200 to 225 grams at the beginning of the experiment were obtained from Harlan Sprague Dawley (Prattville, AL) and housed in the animal quarters on a 14 hour light, 10 hour dark photoperiod with food and water available ad libitum for a week before use. Pentobarbital pellets for implantation were fgrmulated by the University of California at San Francisco. Radiolabeled [°H]SR95531 and [35S]TBPS were purchased from New England Nuclear (Wilmington, DE). Other chemicals were purchased from either Research Biochemicals Incorporated (Natick, MA) or Sigma Chemical Company (St. Louis, MO). Induction of Dentobarbital tolerance and withdrawal Rats were made tolerant to and dependent on pentobarbital by methods previously described (9). The animals were implanted with pellets containing either 75 mg of pentobarbital free acid or the placebo excipient. Two pellets were implanted subcutaneously every other day for 7 days. A total of four pairs of pellets were implanted. The incisions for implanting the pellets were placed on alternating flanks of the rat and no two pairs of implanted pellets were in contact with each other. The incisions for implanting pellets were closed with the cyanoacrylate adhesive, Krazy Glue (B. Jadow & Sons, Inc., NY). To initiate withdrawal, the pellets were removed from the rats. The pellets were wrapped in a piece of nylon stocking before implantation to facilitate their removal. The pellets were located by palpation and the area around each pellet was injected subcutaneously with about 200 #1 of 1% lidocaine hydrochloride. No more than 1 ml of lidocaine local anesthetic was used on any rat. The skin overlying the pellets was opened with a pair of scissors and the pellets pulled out using a pair of forceps. The wound in the skin was then sealed with cyanoacrylate adhesive (Krazy Glue) and a small amount of povidone iodine was painted on the wound as a disinfectant. After removal of the pellets, the placebo- and pentobarbital-implanted rats were designated as the placebo + sham and dependent groups, respectively.
Vol. 50, No. 22, 1992
Barbiturates
and [3H]SR95531 Binding
1703
Since no spontaneous seizures were observed, sensitivity of the rats to chemically-induced convulsions was examined. To test for physical dependence, the rats were injected subcutaneously with 100 #g/kg of t-butylbicyclophosphorothionate (TBPS) 24 hours after removal of the pellets. Approximately 2 mg of TBPS were dissolved in 100 #1 of dimethylsulfoxide (DMSO) and diluted to 1 mg/ml with corn oil. A suspension was made by vortex mixing the corn oil and DMSO. The suspension was diluted 1:10 with corn oil to form a 100 #g/ml solution for injection. The rats were observed for 1 hour for the onset of convulsions. The results were analyzed by grouped t-test
Membrane preparation Rats in each of the different treatment groups were sacrificed by decapitation, and the brains were removed. The frontal cortex and cerebellum were dissected as described by Glowinski and Iversen (16). The brain tissues were homogenized in 15 volumes of 0.32 M sucrose for 1 minute and centrifuged at 1,000 x g for 10 minutes. The supernatant was decanted and centrifuged at 20,000 x g for 20 minutes. The supernatant was discarded and the remaining pellet was homogenized for 15 seconds in 40 volumes of ice-chilled distilled water and centrifuged at 8,000 x g for 20 minutes. The supernatant and the soft upper buffy layer of the pellet were combined and centrifuged at 48,000 x g for 20 minutes. The supernatant was discarded and the pellet was homogenized for 15 seconds in 40 volumes of 50 mM Tris-citrate buffer, pH 7.1, and was centrifuged at 48,000 x g for 20 minutes. The pellet obtained was then homogenized for 15 seconds in 10 volumes of Tris-citrate buffer and frozen at -80°C for at least 48 hours before use in binding assays. On the day a binding assay was to be performed, the membranes were thawed, diluted to 40 volumes with 50 mM Tris-citrate buffer, pH 7.1, homogenized for 15 seconds, and centrifuged at 48,000 x g for 20 min. The pellet was homogenized for 15 seconds in 40 volumes of Tris-citrate buffer and incubated for 30 min at 25°C. The incubated membranes were then centrifuged at 48,000 x g for 20 min, and the pellets were homogenized for 15 seconds in enough Tris-citrate buffer to give a protein concentration of about 1 mg/ml. This final membrane preparation was added to the binding assays described below. Before each assay was initiated, an aliquot of 100 #1 was taken for determination of protein content by the method of Lowry et al. (17).
¢onvulsant bindinq - [3sS]TBP$ The [35S]TBPS binding assay was carried out as previously described (10). The membranes were prepared as described above, except that the 50 mM Triscitrate buffer was adjusted to a pH of 7.4 at room temperature. Two hundred microliters of the membrane preparation were added to glass test tubes containing [35S]TBPS, 200 mM KCI, and 5gmM Tris-citrate buffer, pH 7.4. The final volume was 0.5 ml. The concentration of [°°S]TBPS added was fixed at 5 nM. Scatchard plots were generated by diluting the specific activity of the radioligand with non-radioactive TBPS. The final concentrations of [ 35S]TBPS were 5 to 205 nM. The contents of the test tubes were vortex mixed and the test tubes were placed in a shaking bath at 24°C for 100 minutes. The incubations were terminated by aspirating the reaction mixtures through GF/B filters using a Brandel cell harvester model M24R (Gaithersburg, MD). The filters were washed.twice with 3 ml of ice-cold buffer which contained 200 mM KCI. Specifically bound [°°S]TBPS was defined as the amount of radioactivity displaceable by 100 #M picrotoxinin. Specific binding was greater than 90%. The number of binding sites was expressed as pmol/mg protein. The KD and Bmax values were calculated by linear regression, tabulated, and compared by grouped t-tests.
1704
Barbiturates
and [3H]SR95531 Binding
Vol. 50, No. 22, 1992
GABA antagonist binding - [3H]SR95531 The binding of [3H]SR95531 was performed by a modification of the methods described by Heaulme et al. (14). Two hundred #l of the membrane preparation were added to ice-chilled glass test tubes containing [oH]SR95531 in 0.8 ml of 50 mM Tris-citrate buffer, pH 7.1. The final volume was 1 ml. The concentration of [3H]SR95531 added was fixed at 6 nM. Scatchard plots were generated by diluting the specific activity of the radioligand with non-radioactive SR95531. The final concentrations of [~H]SR95531 were from 6 to 406 nM. Sodium pentobarbital was added to some binding assays at concentrations between 0.2 mM to 5 mM. The contents of the test tubes were vortex mixed and the test tubes were placed in an ice bath for 45 minutes. The incubations were terminated by aspirating the reaction mixtures through GF/B filters using a Brandel cell harvester model M24R (Gaithersburg, MD). The filters were washed twice with 5 ml of buffer. Nonspecific binding was defined as binding in the presence of 100 #M SR95531~ Specific binding was 87% of total binding in the frontal cortex. Scatchard plots of ['~H]SR95531 were curvilinear, and the binding data were fitted to a two binding site model using LIGAND, an iterative curve fitting program written for IBM PC compatible computers (18). The number of binding sites was expressed as pmol/mg protein. The results were tabulated and the binding properties of the different treatment groups were compared by grouped t-tests. The IC5o for the inhibition of [3H]SR95531 binding was determined by probit analysis.
Results Twenty-four hours after the pellets were removed from placebo- and pentobarbital-implanted rats, the dependent rats had a significant decrease (18%) in the latency of TBPS-induced seizures and a significant increase (19%) in cortical [35S]TBPS binding (Table 1). After 7 days of pentobarbital exposure, the KD of low affinity [3H]SR95531 binding was increased in the frontal cortex (Table 2). Removal of the pentobarbital pellets for 24 hours to induce withdrawal reversed the effect of pentobarbital exposure on the KD of the low affinity binding site, and decreased the Bmax of the low affinity site relative to the placebo + sham groujo. In contrast, pentobarbital exposure and withdrawal had no effect on cerebellar [°H]SR95531 binding. Figure 1 shows the concentration-response curve for the inhibition of cortical [3H]SR95531 binding by sodium pentobarbital. The IC50 for inhibition of [3H]SR95531 binding by pentobarbital was 6.5 mM +__0.56. The Ki for inhibition of binding calculated (19) was 3.40 mM. The inhibition was chloride ion independent in the cortex and cerebellum (data not shown). When Scatchard analyses were performed on cortical membranes in the presence or absence of 5 mM sodium pentobarbital, the Bma x of the high affinity binding site was decreased without a change in K D (Figure 2 and Table 3). Neither the K D nor the Bma x of the low affinity site was changed by in vitro pentobarbital.
Vol. 50, No. 22, 1992
Barbiturates
and [3H]SR95531 Binding
1705
Table 1. Effects of pentobarbital withdrawal on TBPS-induced seizures and cortical [35S]TBPS binding Time to first convulsion (min)
(N)
Placebo + Sham (12) 36.41 + 2 . 2 8
(4)
51.8+3.8
2.634-0.07
Dependent
(5)
56.5 4- 1.7
3.12 + 0.08 *
(N)
(15) 30.00 4- 1.32"
[35S]TBPS binding KD Bmax (nM) (pmol/mg)
• P < 0.05, t-test
Table 2. Effect of pentobarbital pellet implantation on [3H]SR95531 binding
(N)
KDH (nM)
BmaxH (pmol/mg)
KDL (nM)
BmaxL (pmol/mg)
Placebo
(9)
6.91 4- 1.11
0.44 4- 0.05
204 4. 29
2.80 4- 0.43
Tolerant
(12)
7.41 +
0.76
0.54 4. 0.04
338 +__46*
3.17 4-0.41
6.94 +
1.04
0.47 4- 0.05
277 4- 33
3.67 +__0.48
Dependent (10) 6.86 4- 0.85
0.55 4- 0.06
211 + 3 1
1.95 4- 0 . 3 2 #
Cortex
PI. +Sham(8)
Cerebellum Placebo
(7) 10.64 +
1.53
0.69 +
0.08
133 + 2 0
1.95 + 0 . 3 5
Tolerant
(7) 10.99 +
0.60
0.85 +
0.07
149 + 2 9
1.90 + 0 . 3 9
PI. + Sham (5) 12.31 +
1.35
1.10 +
0.11
158 + 3 3
1.72 + 0 . 2 6
Dependent (4) 11.44 +
1.77
0.85 +
0.14
151 + 4 3
1.90 + 0 . 3 2
*P < 0.05 vs placebo
#P < 0.01 vs sham
Discussion Our laboratory has previously shown that the pentobarbital withdrawal model used for these experiments causes a decreased latency of pentylenetetrazol-induced seizures which correlated with changes in [35S]TBPS binding (10). To extend this observation and validate the experimental model of pentobarbital withdrawal being
1706
Barbiturates and [3H]SR95531 Binding
Vol. 50, No. 22, 1992
100
._=
eo
t-~
60
"0 t-
0 ¢.,,
0 C)
40
2O o
104
10-3
10 "2 M
[Pentoborbitol]
Figure 1. Concentration-response curve for the inhibition of [3H]SR95531 binding by pentobarbital Specific binding of 6 nM [3H]SR95531 was measured in the absence of pentobarbital or the presence of 0.1 to 5 mM sodium pentobarbital. The binding in the presence of pentobarbital was divided by the binding in the absence of pentobarbital and expressed as a percentage of control. The IC50 for inhibition of [3H]SR95531 binding by pentobarbital was 6.5 mM + 0.56 (N = 8). The calculated Ki for inhibition of binding (19) was 3.40 mM. 0-025
m
A
0.025
0.020
0.020
0.015
0.015
0.010
0.010
B
0.005
1
1
2
Bound ( ~ / . ~
2
3
Bound C,~/,,~)
Figure 2. Scatchard plots of cortical [3H]SR95531 binding in the presence and absence of 5 mM sodium pentobarbital Representative Scatchard plots of [3H]SR95531 binding to cortical membranes in the presence and absence of 5 mM p entobarbital were generated using concentrations of 6 to 400 nM of ['~H]SR95531. A: binding in the absence of pentobarbital. B: binding in the presence of 5 mM sodium pentobarbital.
Vol. 50, NO. 22, 1992
Barbiturates and [3H]SR95531 Binding
1707
Table 3. Effect of in vitro pentobarbital on [3H]SR95531 binding
(N) Control
KDH
Bmax H
KDL
BmaxL
(nM)
(pmol/mg)
(nM)
(pmol/mg)
(6)
5.16 +
0.75
0.59 +
0.07
285 + 4 6
3.32 +0.43
5 mM (6) pentobarbital
7.07 +
1.26
0.27 +
0.05 *
228 + 4 6
3.22 +0.54
*
P< 0.01 vs control, t-test
used, sensitivity to TBPS-induced seizures was examined. The degree of increased seizure sensitivity in the pentobarbital withdrawn rats (18%) was similar to the increase in the number of [35S]TBPS binding sites observed in the frontal cortex (19%). Since TBPS blocks GABA-activated but not glycine-activated chloride channels (20), the changes in binding and convulsant sensitivity are likely to be due to changes in the properties of the GABAA receptor complex. Pentobarbital tolerance increased the K D of the low affinity [3H]SR95531 binding in the frontal cortex (Table 2). Removal of the pellets for 24 hours to induce withdrawal in the dependent group reversed the effect on the low affinity site KD and caused a decrease in the number of low affinity binding sites. The fact that pentobarbital tQlerance and withdrawal produced different changes in the properties of cortical [oH]SR95531 binding suggests that there are multiple ways of regulating GABAA receptor activity. Two mechanisms of regulating the properties of the GABA receptor have been proposed. First, it is well documented that isotypes of the subunits of the GABAA receptor affect the properties of the complex. Changes in subunit composition would therefore have profound effect on receptor properties. The lack of an effect of pentobarbital tolerance and withdrawal on the binding of [3H]SR95531 in the cerebellum may be due to a preponderance of receptors with a subunit composition which makes them refractory to modulation by barbiturates. Second, many of the GABAA receptor subunits multiple consensus sequences for phosphorylation by protein kinases (21). Manipulation of protein phosphoryiation has been observed to alter GABAA receptor activity in cultured cells (22). Similar receptor regulation may take place ~n vivo. Either of these mechanism, or both, may be involved in causing the multiple changes in receptor properties observed in our experiments. Although the IC~o of the in vitro inhibition of [3H]SR95531 binding by pentobarbital seems high, it is important for three reasons. First, it shows that within the therapeutic concentration range the binding of this antagonist is insensitive to modulation by barbiturates. This suggests that the changes in binding properties observed were due to fundamental changes in the GABAA receptor, not residual pentobarbital in the membranes. Second, while the in vitro effects of pentobarbital are on the HIGH affinity site, the changes in receptor properties caused by pentobarbital tolerance and dependence were in the LOW affinity binding site, suggesting that the changes were independent of allosteric modulation. Third, [3H]SR95531 is a relatively new compound and the interactions of this ligand with modulators of the GABAA receptor complex have not been fully characterized. Our data show that when modulation by barbiturates can be produced, the number of high affinity binding sites is affected and the low affinity sites are unchanged. Since it had been previously shown that barbiturates decrease the affinity of [3H]bicuculline
1708
Barbiturates and [3H]SR95531 Binding
Vol. 50, No. 22, 1992
methchloride binding (23), this suggests that there are differences in the binding properties of these two antagonists which merit further study. In summary, withdrawal from pentobarbital tolerance causes an increase in TBPS-induced seizure sensitivity which correlates with an increase in cortical [35S]TBPS binding. Pentobarbital tolerance and dependence caused brain regionspecific changes in the binding of the GABA receptor antagonist, [3H]SR95531. The cortex showed an increase in KD of the low affinity binding site with pentobarbital tolerance. The change in K D was reversed by withdrawal, and caused a decrease in the number of binding sites. While the binding in the frontal cortex was affected by pentobarbital tolerance and dependence, the cerebellum was unaffected by pentobarbital treatment. While the low affinity antagonist binding site was affected by ~nvivo treatment with pentobarbital, the high affinity site was affected by pentobarbital in vitro. These observations suggest that pentobarbital tolerance and withdrawal cause changes in the properties of the GABAA receptor antagonist binding site which are different from those caused by in vitro exposure to the drug. The similarities of the changes observed to those seen in GABA agonist binding (9) are even more intriguing in light of the differences in the allosteric effects of pentobarbital on [3H]SR95531 binding in vitro. Binding studies using the ligand SR95531 may therefore give new insights into the properties of the GABAA receptor complex. This research was supported by grant DA-04480 from NIDA. References
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21.
S. TZENG and I.K. HO. Biochem. Pharmacol. 26699-704 (1977). R.W. OLSEN. J. Neurochem. 371-13 (1981). S.P. SlVAM, T. NABESHIMA and I.K. HO. J. Neurosci. Res. 737-47 (1982). H. MOHLER, T. OKADA and S.J. ENNA. Brain Res. 1,56391-395 (1978). Y. TATSUOKA, Y. KATO, K. YOSHIDA and H. IMURA. Neurosci. Lett. 46255260 (1984). S. LILJEQUIST and B. TABAKOFF. Alcohol 2215-220 (1985). B.R. SONAWANE, S.J. YAFFE and B.H. SHAPIRO. Life Sci. 2"71335-1338 (1980). B.A. FLINT and I.K. HO. Eur. J. Pharmaco165 355-363 (1980). P.A. SAUNDERS, Y. ITO, M.L. BAKER, A.S. HUME and I.K. HO. Pharm. Biochem. Behav. 37343-348 (1990). Y. ITO, P.A. SAUNDERS, D.K. LIM and I.K. HO. J. Neurochem. 521093-1098 (1989). C. VAN RENTERGHEM, G. BILBE, S. MOSS, T.G. SMART, A. CONSTANTI, D.A. BROWN and E.A. BARNARD. Mol. Brain Res. 221-31 (1987). R.B. RAFFA and F. PORRECA. Life Sci. 44245-258 (1989). H. MOHLER and T. OKADA. Mol. Pharmacol. 1_44256-265 (1977). M. HEAULME, J.P. CHAMBON, R. LEYRIS, C.G. WERMUTH and K. BlZlERE. J. Neurochem. 4__881677-1686(1987). T.P. BURCH, R. THYAGARAJAN and M.K. TICKU. Mol. Pharmacol. 2352-59 (1983). J. GLOWlNSKI and L.I. IVERSEN. J. Neurochem. 13655-699 (1966). O.H. LOWRY, N.J. ROSEBROUGH, A.L. FARR and R.J. RANDALL. J. Biol. Chem. 193265-275 (1951). G.A. MCPHERSON. Compu. Prog. in Biomed. 1_7107-114 (1983). Y.-C. CHENG and W.H. PRUSOFF. Biochem. Pharmacol. 2__223099-3108 (1973). A. RIENITZ, C.-M. BECKER, H. BETZ and B. SCHMI'IH. Neurosci. Lett. 769195 (1987). R.W. OLSEN and A.J. TOBIN. FASEB J. 41469-1480 (1990).
Vol. 50, No. 22, 1992
22. 23.
Barbiturates
and [3H]SR95531 Binding
1709
Q.X. CHEN, A. STELZER, A.R. KAY and R.K.S. WONG. J. Physiol 420 207-221 (1990). E.H.F. WONG, A.M. SNOWMAN, L.M.F. LEEB-LUNDBERG, and R.W. OLSEN Europ. J. Pharmacol. 102 205-212 (1984).
Vol.
50, pp.
1701-1709
Pergamon
Press
in t h e U S A
EFFECTS OF PENTOBARBITAL TOLERANCE AND DEPENDENCE ON CONVULSANT AND GABA A RECEPTOR ANTAGONIST BINDING
Paul A. Saunders 1, Toshiyuki Kimura, Teiji Miyaoka, I.K. Ho Department of Pharmacology and Toxicology University of Mississippi Medical Center Jackson, Mississippi 39216-4505 (Received
in final
form March 24,
1992)
Summarv
Experiments were performed which examined the effects of pentobarbital tolerance and dependence on GABAA receptor antagonist binding. In rats implanted with pentobarbital pellets for 7 days, followed by 24 hours of withdrawal, there was a significant decrease in the latency of TBPS-induced seizures and an increase in [35S]TBPS binding in the frontal cortex. The pent(2barbital tolerant rats had a significant increase in the low affinity KD of [°H]SR95531 binding. Removal of the pellets for 24 hours caused a reversal of the effect on the low affinity K D and caused a decrease in the number of low affinity binding sites. In vitro addition of pentobarbital to bin~ling assays produced a decrease in the number of high affinity [oH]SR95531 binding sites without changing low affinity binding. In the cerebellum, the binding in none of the treatment groups was significantly different from placebo. These observations suggest that pentobarbital tolerance and withdrawal cause changes in the properties of the GABAA receptor antagonist binding site which are different from those caused by in vitro exposure to the drug. Although barbiturates have been used as sedative-hypnotics and anticonvulsants for almost a century, relatively little is known about the mechanisms underlying the development of barbiturate tolerance and withdrawal in the central nervous system. Despite studies which show that barbiturates have effects on the GABAA receptor chloride channel complex in vivo (1) and in vitro (2), the influence of barbiturate tolerance and dependence on the properties of the GABAA receptor complex has not been consistently reported. Increases (3), decreases (4), or no changes in GABA receptor binding have been reported (5). Decreases in both benzodiazepine number (6) and affinity (7) have also been seen. The discrepancies in the literature may at least in part be due to differences in animal treatment protocols or binding assay procedures. Our laboratory has developed a pellet implantation model of barbiturate tolerance which produces both CNS tolerance to pentobarbital and increased sensitivity to convulsants in barbiturate withdrawal (8). Using this model, we have shown that pentobarbital tolerance causes an increase in the K D of the low affinity GABA receptor site and a decrease in benzodiazepine binding (9). 1present address: Laboratory of Neurophysiology, NINDS, National Institutes of Health, Building 36, Room 2C02, Bethesda, Maryland 20892 Copyright
0024-3205/92 $5.00 + .00 © 1992 Pergamon Press Ltd All rights
reserved.
1702
Barbiturates and [3H]SR95531 Binding
Vol. 50, No. 22, 1992
thdrawal was shown to cause an increase in the number of binding sites for S]TBPS (10), a convulsant which blocks the GABA-activated chloride channel (11). nce the binding of some receptor antagonists has been observed to be thermodynamically different from that of agonists (12), binding studies using antagonists may reveal properties of receptors not detected by other methods. Recently, a new GABA receptor antagonist, SR95531, became commercially available. This compound is potentially an improvement over bicuculline, for the binding of ~H]bicuculline methiodide is best fit to a one-site model (13), while the binding of [°H]SR95531 is best fit to a two-site model (14). A two-site model for a GABAA antagonist binding is intriguing, for the allosteric interactions of the GABA receptor agonist conformation are dependent on the functional integrity of the low affinity site (15). The following experiments were performed to examine the effects of pentobarbital tolerance, dependence, and in vitro pentobarbital on [°H]SR95531 binding. Methods Animals and materials Male Sprague Dawley rats weighing 200 to 225 grams at the beginning of the experiment were obtained from Harlan Sprague Dawley (Prattville, AL) and housed in the animal quarters on a 14 hour light, 10 hour dark photoperiod with food and water available ad libitum for a week before use. Pentobarbital pellets for implantation were fgrmulated by the University of California at San Francisco. Radiolabeled [°H]SR95531 and [35S]TBPS were purchased from New England Nuclear (Wilmington, DE). Other chemicals were purchased from either Research Biochemicals Incorporated (Natick, MA) or Sigma Chemical Company (St. Louis, MO). Induction of Dentobarbital tolerance and withdrawal Rats were made tolerant to and dependent on pentobarbital by methods previously described (9). The animals were implanted with pellets containing either 75 mg of pentobarbital free acid or the placebo excipient. Two pellets were implanted subcutaneously every other day for 7 days. A total of four pairs of pellets were implanted. The incisions for implanting the pellets were placed on alternating flanks of the rat and no two pairs of implanted pellets were in contact with each other. The incisions for implanting pellets were closed with the cyanoacrylate adhesive, Krazy Glue (B. Jadow & Sons, Inc., NY). To initiate withdrawal, the pellets were removed from the rats. The pellets were wrapped in a piece of nylon stocking before implantation to facilitate their removal. The pellets were located by palpation and the area around each pellet was injected subcutaneously with about 200 #1 of 1% lidocaine hydrochloride. No more than 1 ml of lidocaine local anesthetic was used on any rat. The skin overlying the pellets was opened with a pair of scissors and the pellets pulled out using a pair of forceps. The wound in the skin was then sealed with cyanoacrylate adhesive (Krazy Glue) and a small amount of povidone iodine was painted on the wound as a disinfectant. After removal of the pellets, the placebo- and pentobarbital-implanted rats were designated as the placebo + sham and dependent groups, respectively.
Vol. 50, No. 22, 1992
Barbiturates
and [3H]SR95531 Binding
1703
Since no spontaneous seizures were observed, sensitivity of the rats to chemically-induced convulsions was examined. To test for physical dependence, the rats were injected subcutaneously with 100 #g/kg of t-butylbicyclophosphorothionate (TBPS) 24 hours after removal of the pellets. Approximately 2 mg of TBPS were dissolved in 100 #1 of dimethylsulfoxide (DMSO) and diluted to 1 mg/ml with corn oil. A suspension was made by vortex mixing the corn oil and DMSO. The suspension was diluted 1:10 with corn oil to form a 100 #g/ml solution for injection. The rats were observed for 1 hour for the onset of convulsions. The results were analyzed by grouped t-test
Membrane preparation Rats in each of the different treatment groups were sacrificed by decapitation, and the brains were removed. The frontal cortex and cerebellum were dissected as described by Glowinski and Iversen (16). The brain tissues were homogenized in 15 volumes of 0.32 M sucrose for 1 minute and centrifuged at 1,000 x g for 10 minutes. The supernatant was decanted and centrifuged at 20,000 x g for 20 minutes. The supernatant was discarded and the remaining pellet was homogenized for 15 seconds in 40 volumes of ice-chilled distilled water and centrifuged at 8,000 x g for 20 minutes. The supernatant and the soft upper buffy layer of the pellet were combined and centrifuged at 48,000 x g for 20 minutes. The supernatant was discarded and the pellet was homogenized for 15 seconds in 40 volumes of 50 mM Tris-citrate buffer, pH 7.1, and was centrifuged at 48,000 x g for 20 minutes. The pellet obtained was then homogenized for 15 seconds in 10 volumes of Tris-citrate buffer and frozen at -80°C for at least 48 hours before use in binding assays. On the day a binding assay was to be performed, the membranes were thawed, diluted to 40 volumes with 50 mM Tris-citrate buffer, pH 7.1, homogenized for 15 seconds, and centrifuged at 48,000 x g for 20 min. The pellet was homogenized for 15 seconds in 40 volumes of Tris-citrate buffer and incubated for 30 min at 25°C. The incubated membranes were then centrifuged at 48,000 x g for 20 min, and the pellets were homogenized for 15 seconds in enough Tris-citrate buffer to give a protein concentration of about 1 mg/ml. This final membrane preparation was added to the binding assays described below. Before each assay was initiated, an aliquot of 100 #1 was taken for determination of protein content by the method of Lowry et al. (17).
¢onvulsant bindinq - [3sS]TBP$ The [35S]TBPS binding assay was carried out as previously described (10). The membranes were prepared as described above, except that the 50 mM Triscitrate buffer was adjusted to a pH of 7.4 at room temperature. Two hundred microliters of the membrane preparation were added to glass test tubes containing [35S]TBPS, 200 mM KCI, and 5gmM Tris-citrate buffer, pH 7.4. The final volume was 0.5 ml. The concentration of [°°S]TBPS added was fixed at 5 nM. Scatchard plots were generated by diluting the specific activity of the radioligand with non-radioactive TBPS. The final concentrations of [ 35S]TBPS were 5 to 205 nM. The contents of the test tubes were vortex mixed and the test tubes were placed in a shaking bath at 24°C for 100 minutes. The incubations were terminated by aspirating the reaction mixtures through GF/B filters using a Brandel cell harvester model M24R (Gaithersburg, MD). The filters were washed.twice with 3 ml of ice-cold buffer which contained 200 mM KCI. Specifically bound [°°S]TBPS was defined as the amount of radioactivity displaceable by 100 #M picrotoxinin. Specific binding was greater than 90%. The number of binding sites was expressed as pmol/mg protein. The KD and Bmax values were calculated by linear regression, tabulated, and compared by grouped t-tests.
1704
Barbiturates
and [3H]SR95531 Binding
Vol. 50, No. 22, 1992
GABA antagonist binding - [3H]SR95531 The binding of [3H]SR95531 was performed by a modification of the methods described by Heaulme et al. (14). Two hundred #l of the membrane preparation were added to ice-chilled glass test tubes containing [oH]SR95531 in 0.8 ml of 50 mM Tris-citrate buffer, pH 7.1. The final volume was 1 ml. The concentration of [3H]SR95531 added was fixed at 6 nM. Scatchard plots were generated by diluting the specific activity of the radioligand with non-radioactive SR95531. The final concentrations of [~H]SR95531 were from 6 to 406 nM. Sodium pentobarbital was added to some binding assays at concentrations between 0.2 mM to 5 mM. The contents of the test tubes were vortex mixed and the test tubes were placed in an ice bath for 45 minutes. The incubations were terminated by aspirating the reaction mixtures through GF/B filters using a Brandel cell harvester model M24R (Gaithersburg, MD). The filters were washed twice with 5 ml of buffer. Nonspecific binding was defined as binding in the presence of 100 #M SR95531~ Specific binding was 87% of total binding in the frontal cortex. Scatchard plots of ['~H]SR95531 were curvilinear, and the binding data were fitted to a two binding site model using LIGAND, an iterative curve fitting program written for IBM PC compatible computers (18). The number of binding sites was expressed as pmol/mg protein. The results were tabulated and the binding properties of the different treatment groups were compared by grouped t-tests. The IC5o for the inhibition of [3H]SR95531 binding was determined by probit analysis.
Results Twenty-four hours after the pellets were removed from placebo- and pentobarbital-implanted rats, the dependent rats had a significant decrease (18%) in the latency of TBPS-induced seizures and a significant increase (19%) in cortical [35S]TBPS binding (Table 1). After 7 days of pentobarbital exposure, the KD of low affinity [3H]SR95531 binding was increased in the frontal cortex (Table 2). Removal of the pentobarbital pellets for 24 hours to induce withdrawal reversed the effect of pentobarbital exposure on the KD of the low affinity binding site, and decreased the Bmax of the low affinity site relative to the placebo + sham groujo. In contrast, pentobarbital exposure and withdrawal had no effect on cerebellar [°H]SR95531 binding. Figure 1 shows the concentration-response curve for the inhibition of cortical [3H]SR95531 binding by sodium pentobarbital. The IC50 for inhibition of [3H]SR95531 binding by pentobarbital was 6.5 mM +__0.56. The Ki for inhibition of binding calculated (19) was 3.40 mM. The inhibition was chloride ion independent in the cortex and cerebellum (data not shown). When Scatchard analyses were performed on cortical membranes in the presence or absence of 5 mM sodium pentobarbital, the Bma x of the high affinity binding site was decreased without a change in K D (Figure 2 and Table 3). Neither the K D nor the Bma x of the low affinity site was changed by in vitro pentobarbital.
Vol. 50, No. 22, 1992
Barbiturates
and [3H]SR95531 Binding
1705
Table 1. Effects of pentobarbital withdrawal on TBPS-induced seizures and cortical [35S]TBPS binding Time to first convulsion (min)
(N)
Placebo + Sham (12) 36.41 + 2 . 2 8
(4)
51.8+3.8
2.634-0.07
Dependent
(5)
56.5 4- 1.7
3.12 + 0.08 *
(N)
(15) 30.00 4- 1.32"
[35S]TBPS binding KD Bmax (nM) (pmol/mg)
• P < 0.05, t-test
Table 2. Effect of pentobarbital pellet implantation on [3H]SR95531 binding
(N)
KDH (nM)
BmaxH (pmol/mg)
KDL (nM)
BmaxL (pmol/mg)
Placebo
(9)
6.91 4- 1.11
0.44 4- 0.05
204 4. 29
2.80 4- 0.43
Tolerant
(12)
7.41 +
0.76
0.54 4. 0.04
338 +__46*
3.17 4-0.41
6.94 +
1.04
0.47 4- 0.05
277 4- 33
3.67 +__0.48
Dependent (10) 6.86 4- 0.85
0.55 4- 0.06
211 + 3 1
1.95 4- 0 . 3 2 #
Cortex
PI. +Sham(8)
Cerebellum Placebo
(7) 10.64 +
1.53
0.69 +
0.08
133 + 2 0
1.95 + 0 . 3 5
Tolerant
(7) 10.99 +
0.60
0.85 +
0.07
149 + 2 9
1.90 + 0 . 3 9
PI. + Sham (5) 12.31 +
1.35
1.10 +
0.11
158 + 3 3
1.72 + 0 . 2 6
Dependent (4) 11.44 +
1.77
0.85 +
0.14
151 + 4 3
1.90 + 0 . 3 2
*P < 0.05 vs placebo
#P < 0.01 vs sham
Discussion Our laboratory has previously shown that the pentobarbital withdrawal model used for these experiments causes a decreased latency of pentylenetetrazol-induced seizures which correlated with changes in [35S]TBPS binding (10). To extend this observation and validate the experimental model of pentobarbital withdrawal being
1706
Barbiturates and [3H]SR95531 Binding
Vol. 50, No. 22, 1992
100
._=
eo
t-~
60
"0 t-
0 ¢.,,
0 C)
40
2O o
104
10-3
10 "2 M
[Pentoborbitol]
Figure 1. Concentration-response curve for the inhibition of [3H]SR95531 binding by pentobarbital Specific binding of 6 nM [3H]SR95531 was measured in the absence of pentobarbital or the presence of 0.1 to 5 mM sodium pentobarbital. The binding in the presence of pentobarbital was divided by the binding in the absence of pentobarbital and expressed as a percentage of control. The IC50 for inhibition of [3H]SR95531 binding by pentobarbital was 6.5 mM + 0.56 (N = 8). The calculated Ki for inhibition of binding (19) was 3.40 mM. 0-025
m
A
0.025
0.020
0.020
0.015
0.015
0.010
0.010
B
0.005
1
1
2
Bound ( ~ / . ~
2
3
Bound C,~/,,~)
Figure 2. Scatchard plots of cortical [3H]SR95531 binding in the presence and absence of 5 mM sodium pentobarbital Representative Scatchard plots of [3H]SR95531 binding to cortical membranes in the presence and absence of 5 mM p entobarbital were generated using concentrations of 6 to 400 nM of ['~H]SR95531. A: binding in the absence of pentobarbital. B: binding in the presence of 5 mM sodium pentobarbital.
Vol. 50, NO. 22, 1992
Barbiturates and [3H]SR95531 Binding
1707
Table 3. Effect of in vitro pentobarbital on [3H]SR95531 binding
(N) Control
KDH
Bmax H
KDL
BmaxL
(nM)
(pmol/mg)
(nM)
(pmol/mg)
(6)
5.16 +
0.75
0.59 +
0.07
285 + 4 6
3.32 +0.43
5 mM (6) pentobarbital
7.07 +
1.26
0.27 +
0.05 *
228 + 4 6
3.22 +0.54
*
P< 0.01 vs control, t-test
used, sensitivity to TBPS-induced seizures was examined. The degree of increased seizure sensitivity in the pentobarbital withdrawn rats (18%) was similar to the increase in the number of [35S]TBPS binding sites observed in the frontal cortex (19%). Since TBPS blocks GABA-activated but not glycine-activated chloride channels (20), the changes in binding and convulsant sensitivity are likely to be due to changes in the properties of the GABAA receptor complex. Pentobarbital tolerance increased the K D of the low affinity [3H]SR95531 binding in the frontal cortex (Table 2). Removal of the pellets for 24 hours to induce withdrawal in the dependent group reversed the effect on the low affinity site KD and caused a decrease in the number of low affinity binding sites. The fact that pentobarbital tQlerance and withdrawal produced different changes in the properties of cortical [oH]SR95531 binding suggests that there are multiple ways of regulating GABAA receptor activity. Two mechanisms of regulating the properties of the GABA receptor have been proposed. First, it is well documented that isotypes of the subunits of the GABAA receptor affect the properties of the complex. Changes in subunit composition would therefore have profound effect on receptor properties. The lack of an effect of pentobarbital tolerance and withdrawal on the binding of [3H]SR95531 in the cerebellum may be due to a preponderance of receptors with a subunit composition which makes them refractory to modulation by barbiturates. Second, many of the GABAA receptor subunits multiple consensus sequences for phosphorylation by protein kinases (21). Manipulation of protein phosphoryiation has been observed to alter GABAA receptor activity in cultured cells (22). Similar receptor regulation may take place ~n vivo. Either of these mechanism, or both, may be involved in causing the multiple changes in receptor properties observed in our experiments. Although the IC~o of the in vitro inhibition of [3H]SR95531 binding by pentobarbital seems high, it is important for three reasons. First, it shows that within the therapeutic concentration range the binding of this antagonist is insensitive to modulation by barbiturates. This suggests that the changes in binding properties observed were due to fundamental changes in the GABAA receptor, not residual pentobarbital in the membranes. Second, while the in vitro effects of pentobarbital are on the HIGH affinity site, the changes in receptor properties caused by pentobarbital tolerance and dependence were in the LOW affinity binding site, suggesting that the changes were independent of allosteric modulation. Third, [3H]SR95531 is a relatively new compound and the interactions of this ligand with modulators of the GABAA receptor complex have not been fully characterized. Our data show that when modulation by barbiturates can be produced, the number of high affinity binding sites is affected and the low affinity sites are unchanged. Since it had been previously shown that barbiturates decrease the affinity of [3H]bicuculline
1708
Barbiturates and [3H]SR95531 Binding
Vol. 50, No. 22, 1992
methchloride binding (23), this suggests that there are differences in the binding properties of these two antagonists which merit further study. In summary, withdrawal from pentobarbital tolerance causes an increase in TBPS-induced seizure sensitivity which correlates with an increase in cortical [35S]TBPS binding. Pentobarbital tolerance and dependence caused brain regionspecific changes in the binding of the GABA receptor antagonist, [3H]SR95531. The cortex showed an increase in KD of the low affinity binding site with pentobarbital tolerance. The change in K D was reversed by withdrawal, and caused a decrease in the number of binding sites. While the binding in the frontal cortex was affected by pentobarbital tolerance and dependence, the cerebellum was unaffected by pentobarbital treatment. While the low affinity antagonist binding site was affected by ~nvivo treatment with pentobarbital, the high affinity site was affected by pentobarbital in vitro. These observations suggest that pentobarbital tolerance and withdrawal cause changes in the properties of the GABAA receptor antagonist binding site which are different from those caused by in vitro exposure to the drug. The similarities of the changes observed to those seen in GABA agonist binding (9) are even more intriguing in light of the differences in the allosteric effects of pentobarbital on [3H]SR95531 binding in vitro. Binding studies using the ligand SR95531 may therefore give new insights into the properties of the GABAA receptor complex. This research was supported by grant DA-04480 from NIDA. References
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21.
S. TZENG and I.K. HO. Biochem. Pharmacol. 26699-704 (1977). R.W. OLSEN. J. Neurochem. 371-13 (1981). S.P. SlVAM, T. NABESHIMA and I.K. HO. J. Neurosci. Res. 737-47 (1982). H. MOHLER, T. OKADA and S.J. ENNA. Brain Res. 1,56391-395 (1978). Y. TATSUOKA, Y. KATO, K. YOSHIDA and H. IMURA. Neurosci. Lett. 46255260 (1984). S. LILJEQUIST and B. TABAKOFF. Alcohol 2215-220 (1985). B.R. SONAWANE, S.J. YAFFE and B.H. SHAPIRO. Life Sci. 2"71335-1338 (1980). B.A. FLINT and I.K. HO. Eur. J. Pharmaco165 355-363 (1980). P.A. SAUNDERS, Y. ITO, M.L. BAKER, A.S. HUME and I.K. HO. Pharm. Biochem. Behav. 37343-348 (1990). Y. ITO, P.A. SAUNDERS, D.K. LIM and I.K. HO. J. Neurochem. 521093-1098 (1989). C. VAN RENTERGHEM, G. BILBE, S. MOSS, T.G. SMART, A. CONSTANTI, D.A. BROWN and E.A. BARNARD. Mol. Brain Res. 221-31 (1987). R.B. RAFFA and F. PORRECA. Life Sci. 44245-258 (1989). H. MOHLER and T. OKADA. Mol. Pharmacol. 1_44256-265 (1977). M. HEAULME, J.P. CHAMBON, R. LEYRIS, C.G. WERMUTH and K. BlZlERE. J. Neurochem. 4__881677-1686(1987). T.P. BURCH, R. THYAGARAJAN and M.K. TICKU. Mol. Pharmacol. 2352-59 (1983). J. GLOWlNSKI and L.I. IVERSEN. J. Neurochem. 13655-699 (1966). O.H. LOWRY, N.J. ROSEBROUGH, A.L. FARR and R.J. RANDALL. J. Biol. Chem. 193265-275 (1951). G.A. MCPHERSON. Compu. Prog. in Biomed. 1_7107-114 (1983). Y.-C. CHENG and W.H. PRUSOFF. Biochem. Pharmacol. 2__223099-3108 (1973). A. RIENITZ, C.-M. BECKER, H. BETZ and B. SCHMI'IH. Neurosci. Lett. 769195 (1987). R.W. OLSEN and A.J. TOBIN. FASEB J. 41469-1480 (1990).
Vol. 50, No. 22, 1992
22. 23.
Barbiturates
and [3H]SR95531 Binding
1709
Q.X. CHEN, A. STELZER, A.R. KAY and R.K.S. WONG. J. Physiol 420 207-221 (1990). E.H.F. WONG, A.M. SNOWMAN, L.M.F. LEEB-LUNDBERG, and R.W. OLSEN Europ. J. Pharmacol. 102 205-212 (1984).
