- Email: [email protected]
Correlation of sedative effects with brain levels of barbiturates in LS and SS mice
Alcohol. Vol. 8, pp. 461--466. © Pergamon Press plc. 1991. Printed in the U.S.A.
07,41-8329191 $3.00 + .00
Correlation of Sedative Effects With Brain Levels of Barbiturates in LS and SS Mice C H A R L E S C. D U N C A N A N D J A M E S A. R U T H j
School o f Pharmacy, University o f Colorado, Boulder, CO 80309-0297 R e c e i v e d 19 F e b r u a r y 1991; A c c e p t e d 10 M a y 1991 DUNCAN, C. C. AND J. A. RUTH. Correlation of sedative effects with brain levels of barbiturates in LS and SS mice. ALCOHOL 8(6) 461-466, 1991.--Long-sleep (LS) and short-sleep (SS) mice, genetically selected for their differential CNS sensitivity to ethanol, have also been shown to differ in their response to other sedative-hypnotics, including the barbiturates. We have applied a gas-chromatographic method of analysis of brain barbiturate concentrations following IP administration of either the water-soluble barbiturate dieth3tlbarbital (DB) or the lipid-soluble barbiturate secobarbital (SB). Brain barbiturate levels were assessed at loss of fighting response, and at regaining fighting response (waking). In addition, latency to loss of righting response and duration of loss of fighting response were measured following IP barbiturate administration. We have observed a differential sensitivity of LS and SS mice to the sedative effects of DB, with LS mice having greater sensitivity compared to SS. This differential sensitivity to DB, as measured by a lower concentration of DB which caused loss of fighting in LS, was accompanied by an equal rate of water-soluble barbiturate brain distribution and elimination in the two lines. With the lipid-soluble barbiturate SB, LS and SS mice did not differ in brain SB concentration at loss of fighting response or at waking. However, sleep time was much longer in SS mice than LS due to slower brain clearance of the barbiturate in SS. Therefore, duration of loss of fighting (sleep time) did not adequately reflect central sensitivity to the lipid-soluble barbiturate. These data suggest the importance of quantifying brain concentrations at loss of fighting reflex when assessing central sensitivity to sedative-hypnotic agents. Furthermore, it appears important to consider differences in physico-chemical properties of a particular pharmacological agent before generalizing findings to an entire class of such agents. LS and SS mice Diethylbarbital Secobarbital Differential central sensitivity to water-soluble barbiturates Sedatives Barbiturate elimination rates GLC analysis of brain barbiturate concentrations
(7) study. Alpern and Mclntyre (3) have attributed the discrepancies in findings on pentobarbital-induced sleep time to methodological differences in the studies using PB, such as not controlling for differences in response of the two sexes and not controlling circadian effects on ethanol-induced sedation. However, these results can be explained by the findings of O'Connor et al. (I 6) who demonstrated that PB effects correlated with differential brain clearance rates of [~4C]-pentobarbital. These authors found that SS mice sleep longer than LS and have equal brain PB concentrations at waking. However, SS mice have a slower rate of lipid-soluble barbiturate (PB) brain elimination than LS (16). These discrepancies may also be aggravated by the use of sleep time measurement as the bioassay of CNS sensitivity to these agents. Sleep time, hypothermia and withdrawal seizure sensitivity are all in vivo measures of perturbations in numerous interacting molecular and physiological mechanisms that have been employed as an index of differential central sensitivity to EtOH (19). That these measures are inherently variable can be ascribed to their strong dependence on laboratory environmental conditions, such as ambient temperature and activity noise. Thus, while brain clearance of lipid-soluble barbiturates has been studied, to date no reports have followed the course of brain concentration of the water-soluble barbiturates. Addition-
LONG-SLEEP (LS) and short-sleep (SS) mice have been genetically selected for their differential acute CNS sensitivity to ethanol (12). Numerous studies [e.g., (6)] have shown that these lines are also differentially responsive to other sedative-hypnotic agents including the lipid-soluble barbiturates (8, 16, 21), and the water-soluble barbiturates (8). However, these conclusions have been based in part on a physiological measure of central sensitivity, namely the duration of loss of righting response (anesthesia) induced by the sedative. That this measure is inherently variable is illustrated by the comparison of data using the lipidsoluble barbiturate pentobarbital (PB) as the pharmacologic agent, and sleep time (duration of loss of righting) as the physiological measure of CNS sensitivity. In 1976, Erwin et al. (7) published the seminal paper in this field reporting no difference between LS and SS in central sensitivity to the soporific effects of PB as measured by duration of loss of righting. In this study (7), brain barbiturate concentration was not measured. Moreover, Siemens and Chan (21) have reported that SS sleep longer than LS mice, and were thus more sensitive than LS mice to the effects of this lipid-soluble barbiturate, as sleep time was classically used to define sedative (ethanol) sensitivity in the selection of these mice lines. However, these investigators (21) also reported that the lines did not differ in waking brain barbiturate concentrations following a slightly lower dose of PB than that used in Erwin's 1Requests for reprints should be addressed to Dr. J. A. Ruth.
461
462
DUNCAN AND RUTH
ally, quantification of the amount of water-soluble barbiturate in the brain at loss of righting response, which is a convenient parameter for assessing CNS sensitivity to the sedation-producing effects of sedatives, has not been undertaken, even though the LS and SS mice have been shown to be differentially sensitive to the soporific effects of the water-soluble barbiturates (8, 13, 15). The purpose of the current study was to evaluate and compare the duration of loss of fighting response, and the brain barbiturate concentrations at loss of righting as measures of central sensitivity to both water-soluble and lipid-soluble barbiturates. In this study, we have correlated the brain barbiturate levels with physiological response to both a water-soluble barbiturate, diethylbarbituric acid (DB), and to a relatively lipid-soluble barbiturate, secobarbital [SB, 5-allyl-5-(l-methylbutyl)barbituric acid] in LS and SS mice. In addition, we have studied the time course of DB brain distribution and elimination. This report confirms and extends the findings of O'Connor et al. (16), and reports a simple, generally applicable technique for determining brain barbiturate levels. METHOD
Anbnal Experimental LS and SS male mice of the 47th generation were obtained from the Institute for Behavioral Genetics (IBG) at the University of Colorado-Boulder. All animals were 60-90 days of age at the time of testing. The mice were housed with five like-sex litter mates on a 12-hour light-dark cycle, and received food (Wayne Rodent Blox) and water ad lib. Experimentation occurred between 9 a.m. and noon in a quiet, lighted room with ambient air temperature of 22-24°C. DB and SB (sodium salts) were obtained from Sigma Chemical Co., St. Louis, MO. To assess whether LS and SS mice differed in their central sensitivity to the water-soluble and lipid-soluble barbiturates, we analyzed the brain barbiturate concentration of DB and SB, which are characterized by an octanol/water partition coefficient (log P) of 0.65-0.77 and 1.97, respectively (11), at loss of righting response. We also analyzed the brain barbiturate concentration of DB and SB at regaining righting response. In addition, to correlate these biochemical parameters with the classically used measure of central sensitivity to sedatives, we measured the latency to loss of righting response and the duration of the loss of righting response following IP barbiturate. To study the time course of brain distribution of water-soluble barbiturates in these lines, we analyzed the brain barbiturate concentration of DB at various times following IP barbiturate administration. The sodium salts of DB and SB were prepared in 15 mg/ml (3,09 mM) and 7.5 mg/ml (1.95 mM) solutions, respectively of 50% distilled water/50% propylene glycol. The drugs were administered IP in a volume of 0.01 ml/g body weight, resulting in a dosage of 150 mg/kg DB and 75 mg/kg SB. Following administration of a hypnotic dose of barbiturate, loss of righting response was assessed as the inability of the mouse to right itself from a supine position in a V-shaped acrylic trough for 30 seconds. At this time, the mice of the " l o s s " experimental group were sacrificed by cervical dislocation, their brains rapidly removed, weighed and frozen at - 7 0 ° C until brain barbiturate analysis. Alternatively, mice of the "full-sleep" experimental groups were allowed to remain sleeping until waking, at which time the mice were treated as above for the " l o s s " group. To determine the duration of a full sleep cycle (DLRR), the LLRR value was subtracted from the time value obtained at regaining of the righting response in the "'full-sleep" experimental group.
Regaining fighting response ("full-sleep," waking) was assessed as the ability of the mouse to right itself from supine twice within 30 seconds. Additionally, the time-course of DB brain distribution was studied using tissue obtained from mice sacrificed at the time (minutes) given (Fig. 2) following IP administration of the barbiturate. This tissue was prepared as above for the "'loss" experimental group. Between line data analysis of latency to loss of righting response (LLRR), sleep times (DLRR), and brain barbiturate concentrations of the " l o s s " and "regain" experimental groups was perfornled using the unpaired Student's t-test with p<0.05 being required for significance of experimental effect.
Barbiturate Analysis Whole brain barbiturate concentration was assessed by comparing the relative peak areas of the administered barbiturate with that of an internal standard in gas-liquid chromatographic (GLC) analysis of an extract of brain tissue. This method is a modification of current GLC analysis procedures for barbituric acid analogues (4, 9, 10). Thus brains frozen for I week were thawed and homogenized in a hand-held ground-glass homogenizer (Kontes. size 23) in 1.0 ml 0.05 mg/ml aqueous internal standard solution with 3.0 ml 0.1 N perchloric acid. When brain DB concentration was analyzed, SB was the internal standard and when brain SB concentration was analyzed, DB served as the internal standard. The tissue homogenate was then centrifuged for 10 minutes at 12,350×g. Barbiturate was then extracted from the supernatant using preconditioned C-18 solid phase extraction columns (SPE) (J. T. Baker, #7020-03, 3 ml volume). Elution was effected with 1.0 ml acetone. To the eluent, 1.0 ml of 4.0 mg/l methyl-laurate (in chloroform) was added and this solution was evaporated in a ReactiVap (Pierce Chemical Company, #18780) under N 2 stream to dryness. Extraction efficiency using this procedure was approximately 90% for both barbiturates. The barbiturate residue was then dissolved and briefly mixed by vortexing in 25 ILl 0.2 M trimethylanilinium hydroxide methanol solution (TMAH) (Pierce Chemical Company) and 224 ILl methanol. This solution was injected onto the GLC column in a volume of 3 ILl using a Hamilton syringe. The GC (Hewlett-Packard #5730A) was equipped with a flameionization detector and a Recordall (Fisher, Series 5,000) chart recorder. The column conditions were: 3% OV-17 on Chromasorb W (length 7 feet, i.d. 2 mm), beginning temperature 140 ° programmed to 200 ° at 8°/minute, injection port temperature 300 °, which promotes the in-port derivatization reaction. RESULTS
The in vivo response of LS and SS mice following DB administration is shown in Table 1. These data consist of the LS and SS latency to loss of righting response (LLRR) and the duration of loss of righting response (DLRR). Thus a 150 mg/kg IP dose of DB resulted in a significantly shorter LLRR in LS than in SS mice ( L S = I I . 4 _ 1.1 minutes, S S = 2 3 . 1 _ 1.9 minutes). Additionally, this barbiturate induced a significantly longer DLRR in LS (296.0__.63.0 min) mice as compared to that of SS (20.0---5.2) mice (Student's t-test, p<0.05). The in vivo effect of SB is illustrated in Table 2, The SB data consist of the LS and SS latency to loss of righting response (LLRR) and the duration of loss of righting response (DLRR). Thus a 75 mg/kg IP dose of this barbiturate resulted in an equal LLRR ( L S = 3 . 0 ± 0 . 2 min, SS=3.5---0.5 min). However, in the physiological measure of CNS sensitivity, SB induced a significantly longer DLRR in SS (284.0---28.3 min) mice as corn-
BARBITURATE EFFECTS IN LS AND SS MICE
463
TABLE 1
Line LS SS
•
2.s
DIETHYLBARBITAL (DB) PHYSIOLOGICAL DATA*
LLRR
DLRR
11.4 -+- l.l.'t 23.1 z 1.9
296.0 ~ 63.0"I" 20.0-+ 5.2
Loss
2.0 JO
ixl o
1.5
¢-
~ Each data point represents the mean -+ S.E.M. of 4-6 determinations performed on 2 or 3 different days. *LLRR = latency to loss of righting response, DLRR = duration of loss of righting response, both measured in minutes. "tSignificantly different from the corresponding SS value, p<0.05. unpaired Student's t-test.
' ~
1.0 0.5 0.0
kS
SS Line
pared to that of LS (198.0--17.6 min) mice (Student's t-test, p<0.05). Figure 1 illustrates the brain DB concentration at loss of fighting response and at regaining the righting response following a 150 mg/kg IP dose of DB. That LS mice have greater CNS sensitivity to DB than SS mice is illustrated by a significantly lower brain concentration at loss of righting response in LS ( 0 . 9 2 - + 0 . 0 9 × 1 0 - 3 I-tg D B / m g brainl c o m p a r e d to SS (1.44+-0.05 × 10-3 ixg DB/mg brain) (Student's t-test, p<0.05). However, the DB brain concentration at which LS and SS regain the righting response is the same in the two lines (LS: 1.79+-0.20x 10 -3 la,g DB/mg brain, SS: 1.88+-0.14x 10 - 3 ~g DB/mg brain). Figure 2 illustrates the time course of DB brain distribution in LS and SS mice. This figure illustrates that the DB brain distribution rate is the same in the two lines, especially during the period of distribution leading to loss of the righting response. Therefore, LS mice, which have a shorter LLRR and a longer DLRR induced by DB, but have an equal rate of DB distribution to that of SS mice, thus have greater central sensitivity to DB as measured by a lower DB brain concentration at loss of the righting response compared to that in SS mice. In addition, the brain DB elimination rates do not differ in these lines during the time course of a mean duration of loss of fighting. Figure 3 illustrates the brain SB concentration at loss of righting response and at regaining the righting response following a 75 mg/kg IP dose of SB. That LS and SS mice do not have significantly different CNS sensitivity to SB is illustrated by equal brain concentration at loss of righting response LS ( 4 . 0 9 + - 0 . 3 3 x 1 0 -'~ I.,tg S B / m g brain) c o m p a r e d to SS (4.32+-0.02× l0 - 4 ~g SB/mg brain). These data (Fig. 3) indicate that the SB brain elimination rate is different in the two lines [LS, "regain" (198 minutes following LLRR)= 1.6 +- 0.29 ×
FIG. I. Brain diethylbarbital (DB) concentration (× 10 -3. measured in Ixg DB/mg brain wet weight) at loss of fighting response (loss) and at regaining the righting response (regain) following a 150 mg/kg IP dose of DB. Each data point represents the mean ± S.E.M. of 5-10 determinations performed on 2 or 3 different days. Note that LS Loss= I 1.4-- I.I and Regain=296.0--63.0 minutes; whereas SS Loss= 23. I _+ 1.9 and Regain = 20.5---5.2 minutes. *Significantly different from the corresponding SS value, p<0.05, unpaired Student's t-test.
10 - 4 p~g SB/mg brain: SS " r e g a i n " (284 minutes following L L R R ) = 2 . 1 8 - + 0 . 3 0 × 10 -8 p,g SB/mg brain]. Thus, SS mice, which have a longer DLRR induced by SB, have a slower rate of SB clearance than do LS mice. Other studies have demonstrated that the LS and SS mice differ in lipid-soluble barbiturate clearance rates ( 16,21 ). Considering that administration route and composition of the administration vehicle, and thus activation of adrenocortical response systems, may play a role in differential responses of the LS and SS mice (25), we employed a 50% distilled water/50% propylene glycol injection vehicle in which the barbiturates were administered. We employed this vehicle to decrease the discom-
LS Loss ~
Line
LLRR
DLRR
1.5"
1.00.5-
0.0
• I
20 LS
3.0 -+- 0.2
198.0 -+- 17.6"t
SS
3.5 - 0.5
284.0 +-- 28.3
Each data point represents the mean _ S.E.M. of 4-6 determinations performed on 3 or 4 different days. *LLRR = latency to loss of fighting response. DLRR = duration of loss of fighting response, both measured in minutes. 1"Significantly different from the corresponding SS value, p<0.05, unpaired Student's t-test.
LS R e g a l n ~
Regain
2.0-
a o ¢
TABLE 2
SS
2.5-
CO
SECOBARBITAL (SB) PHYSIOLOGICALDATA*
SS Loss
3.0-
•
I
40
"
I
60
•
I
•
I
•
I
•
I
"
l
'
$s I
•
I
•
I
80 1 0 0 1 2 0 1 4 0 1 6 0 1 8 0 2 0 0 2 2 0 Minutes
FIG. 2. Brain diethylbarbital (DB) concentration in LS and SS mice (measured in p,g DB/mg brain wet weight) as a function of time following a 150 mg/kg IP dose of DB. For minutes 0-30, samples were collected every 5 minutes; 30-60, every l0 minutes; and 60-120, every 30 minutes. Each data point represents the mean+-S.E.M, of 4-6 determinations performed in 3 different experiments.
464
DUNCAN AND RUTH
5.0 4.5 4.0 .Q 3.5 .~3.o al 2.5 c
m 2.o
'o
1.0
" 0.5 0.0 LS
SS
Line FIG. 3. Brain secobarbital (SBI concentration ( x l0 -4, measured in p~g SB/mg brain wet weight) at loss of righting response (Iossl and at regaining the righting response (regain) following a 75 mg/kg IP dose of SB. Each data point represents the mean - S.E.M. of 5-8 determinations performed in 3 different experiments. Note that LS Loss = 3.0__.0.2. and Regain=198.0_ 17.6 minutes, whereas SS Loss=3.5---0.5 and Regain = 284.0 ___28.3 minutes.
fort and possible confounding changes in neurochemistry associated with IP administration, and to enhance the chemical stability of the barbiturate injection solution. Previously, Osterlind et al. (18) have reported that similar propylene glycol doses had no effect on the naive animal EEG, but did lower the threshold dose for perturbation of the EEG recorded in the presence of hexobarbital. Therefore, the possibly confounding effect of the 50% distilled water/50% propylene glycol injection vehicle on barbiturateinduced sleep time was assessed by studying the effects of administration of this solution alone to these mice. This dose of propylene glycol (approximately 4.0 g/kg) was ineffective in producing significant hypothermia (0.5 ± 0. I°C loss) measured at 15 minutes postinjection, or in inducing loss of righting response in both lines. DISCUSSION
The purpose of the current study was to evaluate and compare A) duration of loss of righting response, and B) brain barbiturate concentrations at loss of righting as measures of central sensitivity to both water-soluble and lipid-soluble barbiturates. The in vivo measures which have been employed as an index of differential sensitivity to sedative/hypnotics are considered to be inherently variable (19). Considering that this inherent variability is probably a function of the complex physiological sequelae following EtOH exposure which results in these in vivo responses, this study does not attempt to control for all the possible variables, but does suggest that loss of righting response, when combined with quantification of brain concentrations of the pharmacological agent, does provide an accurate index of differential central sensitivity. The results of examining both waterand lipid-soluble barbiturates under the same laboratory conditions verify that LS mice have greater CNS sensitivity to the water-soluble barbiturate DB than do SS mice as illustrated by a significantly lower brain concentration at loss of righting response, a shorter LLRR, as well as a significantly longer duration of sedation induced by a 150 mg/kg IP dose of DB. That a shorter LLRR and a longer DLRR is induced in LS following IP DB is due to differences in neuronal sensitivity to the sedative effects of DB and is not due to differences in brain distribution or elimination is shown in Fig. 2, where these brain barbiturate
concentration time-course functions are the same in the two lines. Thus the conclusion that LS have greater CNS sensitivity to the water soluble barbiturate DB than SS is supported by both brain concentration of the barbiturate at loss of righting and by the DLRR. These data support the suggestions of Marley et al. (15) that water-soluble barbiturates may have a common mechanism of CNS depressant action with ethanol and other sedativehypnotic agents that have similar physico-chemical properties including the octanol/water partition coefficient, and that these agents may thus be of value in elucidating the mechanisms by which ethanol elicits these depressant effects. The observation that brain DB concentration in LS and SS at waking do not differ between these lines (Fig. I), may reflect a time-dependent redistribution of the drug into brain compartments not involved in the pharmacological effect. That the distribution of barbiturates is not simply a function of log P has been recently elucidated (22), which suggests that the particular functionality of the barbiturate may be an important consideration for pharmacokinetic and (possibly) pharmacological studies on the effects of these agents. Furthermore, the greater brain concentration of DB at waking than at the loss of righting in both lines suggests that the development of acute tolerance to the effects of the water-soluble barbiturate occurs. Indeed, Tabakoff et al. (231 have studied the phenomenon of acute tolerance to the sedative effects of EtOH in the LS and SS mice, and these investigators reported that the lines did not exhibit acute tolerance following IP EtOH. However, even though the current study was performed by quantifying brain barbiturate concentrations, a measure which necessitates the use of two sets of animals to generate the loss and waking data, the development of within-session tolerance to the effects of barbiturates may further confound the determination of relative central sensitivities when using brain concentration at the return of the righting response as an index. Thus it may be more accurate (pharmacologically) as well as being more convenient to use barbiturate concentration at loss of righting response rather than at regaining righting response as a measure of central sensitivity. Based solely upon duration of loss of righting response, one would conclude that LS and SS mice differ in sensitivity to the soporific effects of SB with SS mice having greater sensitivity. However, in contrast to the observations made with the watersoluble barbiturate, this conclusion is inconsistent with the SB biochemical data which demonstrate no between-line differences exist in CNS sensitivity to SB as measured by brain concentrations at loss of righting response in LS and SS mice. Because these two measures of central sensitivity are divergent, disagreements have arisen as to whether, in fact, the LS and SS lines differ in CNS sensitivity to these sedative hypnotics. However, our data suggest that brain clearance of SB in SS mice is much slower than in LS. as has been previously observed for clearance of the lipid-soluble barbiturate pentobarbital by O'Connor et al. (16). These investigators have attributed this differential rate of lipid-soluble barbiturate clearance from brain to the significantly greater body fat content of the LS mouse, therefore facilitating ready redistribution of the barbiturate from the circulation of LS mice. These differential elimination rates account for the differences in sleep time in the two lines. Thus, considering this pharmacokinetic difference, central sensitivity to lipidsoluble barbiturates can only be accurately assessed with brain concentration of the sedative agent producing loss of righting response. Furthermore, the brain concentrations of SB are on the same order as those reported for the relatively lipid-soluble barbiturate pentobarbital (16). Interestingly, the sedative-producing concentrations of DB were an order of magnitude greater than those of SB. Clearly, this reflects differences in the physico-chemical
BARBITURATE EFFECTS IN LS AND SS MICE
465
properties of the two barbiturates, properties which undoubtedly influence the partitioning of the agents to their respective sites of action. Support for this hypothesis is provided by reports in which the effect of various barbiturates on 3,-aminobutyric acid (GABA) binding was studied (17,24). These investigations have demonstrated that phenobarbital does not enhance GABAergic binding, whereas barbiturates characterized as being relatively more lipid-soluble do have an effect on binding to chloride-ionophore/GABA receptor complex (17,24). Numerous investigations have reported line differences between the LS and SS mice in GABAergic system activity or regulation thereof (14), as well as differences in in vitro and in vivo binding assessments of the chloride-ionophore/GABA receptor complex [e.g., (I)]. However, between-line differences in other neurochemical systems exist and barbiturates may also interact with these systems to induce sedation with a mechanism which is a function of, or independent of log P. In summary, this study supports the hypothesis that LS and SS mice, which have been selected for differential CNS sensitivity to ethanol, also have divergent central sensitivity to the water-soluble barbiturates represented in this study by DB but not the lipid-soluble barbiturates (e.g., SB). These conclusions are consistent with other studies which have employed numerous central depressants (2, 7, 15). Therefore, as Marley et al. (15) previously concluded, this differential sensitivity of LS and SS mice, as measured by divergent sleep time, to depressants depends largely on the specific physico-chemical properties of the depressant, and not the general structural category of the pharmacologic agent (ethanol vs. other depressants).
The method used for the measurement of brain barbiturate concentrations employs techniques that are elegant and generally applicable to the barbiturates. Using this GC method combining solid phase extraction techniques and a facile derivatization reaction, the full range of lipid-soluble, water-soluble, and epileptogenic barbiturates can be assayed from the brain. The sensitivity and specificity of this technique depends on the chromatographic behavior of the drug molecules on a gas-liquid column and is thus highly sensitive and specific. In addition, while performing this assay in greater than 100 experiments, we have observed a high degree of reproducibility with small variability as measured by standard drug peak area analysis (unpublished observations). Numerous analytical techniques, including liquid-scintillation spectrometry (12), UV-spectrophotometric (5) and GC methods (4, I 0), have been employed in the quantification of barbiturates from various biological samples. However, the advantage of this assay is that it has a high degree of selectivity and is generally more facile because of the solid phase extraction combined with gas-chromatographic techniques. Thus, as is possible with radiolabelled agents or in assays depending on chromophore behavior above background, the metabolites do not confound the concentration measurements derived with this technique. ACKNOWLEDGEMENTS This work was supported by USPHS grant AA-03527. Partial support was also provided by research awards from Sigma Xi and the Upjohn Company (C.C.D.). The authors also wish to thank Dr. A. C. Collins for valuable discussions.
REFERENCES 1. Allan, A. M.; Harris, R. A. Gamma-aminobutyric acid and alcohol actions: Neurochemical studies of long sleep and short sleep mice. Life Sci. 39:2005-2015; 1986. 2. Alpern, H. P.; Mclntyre, T. D. Evidence that the selectively bred long- and short-sleep mouse lines display common narcotic reactions to many depressants. Psychopbarmacology (Berlin) 85:456--459; 1985. 3. Alpern, H. P.; Mclntyre, T. D. Sedative-hypnotic anomalies related to dose of pentobarbital in long-sleep and short-sleep selectivelybred mice. Pharmacol. Biochem. Behav. 25:333-336; 1986. 4. Brochmann-Hanssen, E.; Oke, T. O. Gas chromatography of barbiturates, phenolic alkaloids, and xanthine bases: Flash-heater methylation by means of trimethylanilinium hydroxide. J. Pharm. Sci. 58(3):370-371; 1969. 5. Broughton, P. M. G. A rapid ultraviolet spectrophotometric method for the detection, estimation and identification of barbiturates in biological material. Biochem. J. 63:207-213; 1956. 6. Collins, A. C. A review of research using short-sleep and long-sleep mice. In: McClearn, G. E.; Dietrich, R. A.; Erwin, V. G., eds. Development of animal models as pharmacogenetic tools. DHHS publication #81-1133. Washington, DC: US Government Printing Office; 1981:161-170. 7. Erwin, V. G.; Heston, W. D. W.; McClearn, G. E. Effect of hypnotics on mice genetically selected for sensitivity to ethanol. Pharmacol. Biochem. Behav. 4:679-683; 1976. 8. Howerton, T. C.; O'Conner, M. F.; Collins, A. C. Lipid solubility is correlated with hypnotic and hypothermic responses of long-sleep (LS) and short-sleep (SS) mice to various depressant drugs. J. Pharmacol. Exp. Ther. 227:389-393; 1983. 9. Kananen, G.; Osiewicz, R.; Sunshine, I. Barbiturate analysis--A current assessment. J. Chromatogr. Sci. 10:283-287; 1972. 10. Kim, C.; Berg, R.; Khanna, J. M. Rapid and sensitive gas chromatographic method for detection of barbital and pentobarbital in blood using flash-heater methytation and nitrogen specific detection. Chromatographia 23:269-272; 1987. II. Leo, A.; Hansch, C.; Elkins, D. Partition coefficients and their uses. Chem. Rev. 71:525-615; 1971. 12. McClearn, G. E.; Kakihana, R. Selective breeding for ethanol sen-
13.
14. 15. 16. 17. 18. 19. 20.
21. 22. 23.
sitivity: Short-sleep and long-sleep mice. In: McClearn, G. E.; Dietrich, R. A.; Erwin, V. G., eds. Development of animal models as pharmacogenetic tools. DHHS publication #81-1133. Washington, DC: U.S. Government Printing Office; 1981:147-159. Mclntyre, T. D.; Alpern, H. P. Thiopental, phenobarbital, and chlordiazepoxide induce the same differences in narcotic reaction as ethanol in long-sleep and short-sleep selectively bred mice. Pharmacol. Biochem. Behav. 24:895-898; 1986. Marley, R. J.; Freund, R. K.; Wehner, J. M. Differential response to flurazepam in long-sleep and short-sleep mice. Pharmacol. Bitchem. Behav. 31:453--458; 1988. Marley, R. J.; Miner, L. L.; Wehner, J. M.; Collins, A. C. Differential effects of central nervous system depressants in long-sleep and short-sleep mice. J. Pharmacol. Exp. Ther. 238:1028-1033; 1986. O'Connor, M. F.; Howerton, T. C.; Collins, A. C. Effects of pentobarbital in mice selected for differential sensitivity to ethanol. Pharmacol. Biochem. Behav. 17:245-248; 1982. Olsen, R. W.; Snowman, A. M. Chloride dependent enhancement by barbiturates of gamma-amino butyric acid receptor binding. J. Neurosci. 2:1812-1823; 1982. Osterlind, A.; Akesson, A.; Wahlstrom, G. Interactions between 1,2-propanediol (propylene glycol) and hexobarbital. Acta Pharmacol. Toxicol. 45:245-248; 1979. Phillips, T. J.; Feller, D. J.; Crabbe, J. C. Selected lines, alcohol and behavior. Experientia 45:805-827; 1989. San-Marina, A.; Khanna, J. M.; Kalant, H. Relationship between initial sensitivity, acute tolerance and chronic tolerance to ethanol in a heterogeneous population of Swiss mice. Psychopharmacology (Berlin) 99:450--457; 1989. Siemens, A. J.; Chan, A. W. K. Differential effects of pentobarbital and ethanol in mice. Life Sci. 19:581-590; 1976. Steiner, S. H.; Moor, M. J.; Bickel, M. H. Kinetics of distribution and adipose tissue storage as a function of lipophilicity and chemical structure. I. Barbiturates. Drug Metab. Dispos. 19:8-14; 1991. Tabakoff, B.; Ritzmann, R. F.; Raju, T. S.; Deitrich, R. A. Characterization of acute and chronic tolerance in mice selected for inherent differences in sensitivity to ethanol. Alcohol.: Clin. Exp.
466
Res. 4:70-73; 1980. 24. Whittle, S. R.; Turner, A. J. Differential effects of sedative and anticonvulsant barbiturates on specific [3H]GABA binding to membrane preparations from rat brain cortex. Biochem. Pharmacol. 31:
DUNCAN AND RUTH
2891-2895; 1982. 25. Zgombick, J, M,; Erwin, V, G. Central mechanisms of ethanol-induced adrenocortical response in selectively bred lines of mice. Neuroendocrinology 46:324--332: 1987.
07,41-8329191 $3.00 + .00
Correlation of Sedative Effects With Brain Levels of Barbiturates in LS and SS Mice C H A R L E S C. D U N C A N A N D J A M E S A. R U T H j
School o f Pharmacy, University o f Colorado, Boulder, CO 80309-0297 R e c e i v e d 19 F e b r u a r y 1991; A c c e p t e d 10 M a y 1991 DUNCAN, C. C. AND J. A. RUTH. Correlation of sedative effects with brain levels of barbiturates in LS and SS mice. ALCOHOL 8(6) 461-466, 1991.--Long-sleep (LS) and short-sleep (SS) mice, genetically selected for their differential CNS sensitivity to ethanol, have also been shown to differ in their response to other sedative-hypnotics, including the barbiturates. We have applied a gas-chromatographic method of analysis of brain barbiturate concentrations following IP administration of either the water-soluble barbiturate dieth3tlbarbital (DB) or the lipid-soluble barbiturate secobarbital (SB). Brain barbiturate levels were assessed at loss of fighting response, and at regaining fighting response (waking). In addition, latency to loss of righting response and duration of loss of fighting response were measured following IP barbiturate administration. We have observed a differential sensitivity of LS and SS mice to the sedative effects of DB, with LS mice having greater sensitivity compared to SS. This differential sensitivity to DB, as measured by a lower concentration of DB which caused loss of fighting in LS, was accompanied by an equal rate of water-soluble barbiturate brain distribution and elimination in the two lines. With the lipid-soluble barbiturate SB, LS and SS mice did not differ in brain SB concentration at loss of fighting response or at waking. However, sleep time was much longer in SS mice than LS due to slower brain clearance of the barbiturate in SS. Therefore, duration of loss of fighting (sleep time) did not adequately reflect central sensitivity to the lipid-soluble barbiturate. These data suggest the importance of quantifying brain concentrations at loss of fighting reflex when assessing central sensitivity to sedative-hypnotic agents. Furthermore, it appears important to consider differences in physico-chemical properties of a particular pharmacological agent before generalizing findings to an entire class of such agents. LS and SS mice Diethylbarbital Secobarbital Differential central sensitivity to water-soluble barbiturates Sedatives Barbiturate elimination rates GLC analysis of brain barbiturate concentrations
(7) study. Alpern and Mclntyre (3) have attributed the discrepancies in findings on pentobarbital-induced sleep time to methodological differences in the studies using PB, such as not controlling for differences in response of the two sexes and not controlling circadian effects on ethanol-induced sedation. However, these results can be explained by the findings of O'Connor et al. (I 6) who demonstrated that PB effects correlated with differential brain clearance rates of [~4C]-pentobarbital. These authors found that SS mice sleep longer than LS and have equal brain PB concentrations at waking. However, SS mice have a slower rate of lipid-soluble barbiturate (PB) brain elimination than LS (16). These discrepancies may also be aggravated by the use of sleep time measurement as the bioassay of CNS sensitivity to these agents. Sleep time, hypothermia and withdrawal seizure sensitivity are all in vivo measures of perturbations in numerous interacting molecular and physiological mechanisms that have been employed as an index of differential central sensitivity to EtOH (19). That these measures are inherently variable can be ascribed to their strong dependence on laboratory environmental conditions, such as ambient temperature and activity noise. Thus, while brain clearance of lipid-soluble barbiturates has been studied, to date no reports have followed the course of brain concentration of the water-soluble barbiturates. Addition-
LONG-SLEEP (LS) and short-sleep (SS) mice have been genetically selected for their differential acute CNS sensitivity to ethanol (12). Numerous studies [e.g., (6)] have shown that these lines are also differentially responsive to other sedative-hypnotic agents including the lipid-soluble barbiturates (8, 16, 21), and the water-soluble barbiturates (8). However, these conclusions have been based in part on a physiological measure of central sensitivity, namely the duration of loss of righting response (anesthesia) induced by the sedative. That this measure is inherently variable is illustrated by the comparison of data using the lipidsoluble barbiturate pentobarbital (PB) as the pharmacologic agent, and sleep time (duration of loss of righting) as the physiological measure of CNS sensitivity. In 1976, Erwin et al. (7) published the seminal paper in this field reporting no difference between LS and SS in central sensitivity to the soporific effects of PB as measured by duration of loss of righting. In this study (7), brain barbiturate concentration was not measured. Moreover, Siemens and Chan (21) have reported that SS sleep longer than LS mice, and were thus more sensitive than LS mice to the effects of this lipid-soluble barbiturate, as sleep time was classically used to define sedative (ethanol) sensitivity in the selection of these mice lines. However, these investigators (21) also reported that the lines did not differ in waking brain barbiturate concentrations following a slightly lower dose of PB than that used in Erwin's 1Requests for reprints should be addressed to Dr. J. A. Ruth.
461
462
DUNCAN AND RUTH
ally, quantification of the amount of water-soluble barbiturate in the brain at loss of righting response, which is a convenient parameter for assessing CNS sensitivity to the sedation-producing effects of sedatives, has not been undertaken, even though the LS and SS mice have been shown to be differentially sensitive to the soporific effects of the water-soluble barbiturates (8, 13, 15). The purpose of the current study was to evaluate and compare the duration of loss of fighting response, and the brain barbiturate concentrations at loss of righting as measures of central sensitivity to both water-soluble and lipid-soluble barbiturates. In this study, we have correlated the brain barbiturate levels with physiological response to both a water-soluble barbiturate, diethylbarbituric acid (DB), and to a relatively lipid-soluble barbiturate, secobarbital [SB, 5-allyl-5-(l-methylbutyl)barbituric acid] in LS and SS mice. In addition, we have studied the time course of DB brain distribution and elimination. This report confirms and extends the findings of O'Connor et al. (16), and reports a simple, generally applicable technique for determining brain barbiturate levels. METHOD
Anbnal Experimental LS and SS male mice of the 47th generation were obtained from the Institute for Behavioral Genetics (IBG) at the University of Colorado-Boulder. All animals were 60-90 days of age at the time of testing. The mice were housed with five like-sex litter mates on a 12-hour light-dark cycle, and received food (Wayne Rodent Blox) and water ad lib. Experimentation occurred between 9 a.m. and noon in a quiet, lighted room with ambient air temperature of 22-24°C. DB and SB (sodium salts) were obtained from Sigma Chemical Co., St. Louis, MO. To assess whether LS and SS mice differed in their central sensitivity to the water-soluble and lipid-soluble barbiturates, we analyzed the brain barbiturate concentration of DB and SB, which are characterized by an octanol/water partition coefficient (log P) of 0.65-0.77 and 1.97, respectively (11), at loss of righting response. We also analyzed the brain barbiturate concentration of DB and SB at regaining righting response. In addition, to correlate these biochemical parameters with the classically used measure of central sensitivity to sedatives, we measured the latency to loss of righting response and the duration of the loss of righting response following IP barbiturate. To study the time course of brain distribution of water-soluble barbiturates in these lines, we analyzed the brain barbiturate concentration of DB at various times following IP barbiturate administration. The sodium salts of DB and SB were prepared in 15 mg/ml (3,09 mM) and 7.5 mg/ml (1.95 mM) solutions, respectively of 50% distilled water/50% propylene glycol. The drugs were administered IP in a volume of 0.01 ml/g body weight, resulting in a dosage of 150 mg/kg DB and 75 mg/kg SB. Following administration of a hypnotic dose of barbiturate, loss of righting response was assessed as the inability of the mouse to right itself from a supine position in a V-shaped acrylic trough for 30 seconds. At this time, the mice of the " l o s s " experimental group were sacrificed by cervical dislocation, their brains rapidly removed, weighed and frozen at - 7 0 ° C until brain barbiturate analysis. Alternatively, mice of the "full-sleep" experimental groups were allowed to remain sleeping until waking, at which time the mice were treated as above for the " l o s s " group. To determine the duration of a full sleep cycle (DLRR), the LLRR value was subtracted from the time value obtained at regaining of the righting response in the "'full-sleep" experimental group.
Regaining fighting response ("full-sleep," waking) was assessed as the ability of the mouse to right itself from supine twice within 30 seconds. Additionally, the time-course of DB brain distribution was studied using tissue obtained from mice sacrificed at the time (minutes) given (Fig. 2) following IP administration of the barbiturate. This tissue was prepared as above for the "'loss" experimental group. Between line data analysis of latency to loss of righting response (LLRR), sleep times (DLRR), and brain barbiturate concentrations of the " l o s s " and "regain" experimental groups was perfornled using the unpaired Student's t-test with p<0.05 being required for significance of experimental effect.
Barbiturate Analysis Whole brain barbiturate concentration was assessed by comparing the relative peak areas of the administered barbiturate with that of an internal standard in gas-liquid chromatographic (GLC) analysis of an extract of brain tissue. This method is a modification of current GLC analysis procedures for barbituric acid analogues (4, 9, 10). Thus brains frozen for I week were thawed and homogenized in a hand-held ground-glass homogenizer (Kontes. size 23) in 1.0 ml 0.05 mg/ml aqueous internal standard solution with 3.0 ml 0.1 N perchloric acid. When brain DB concentration was analyzed, SB was the internal standard and when brain SB concentration was analyzed, DB served as the internal standard. The tissue homogenate was then centrifuged for 10 minutes at 12,350×g. Barbiturate was then extracted from the supernatant using preconditioned C-18 solid phase extraction columns (SPE) (J. T. Baker, #7020-03, 3 ml volume). Elution was effected with 1.0 ml acetone. To the eluent, 1.0 ml of 4.0 mg/l methyl-laurate (in chloroform) was added and this solution was evaporated in a ReactiVap (Pierce Chemical Company, #18780) under N 2 stream to dryness. Extraction efficiency using this procedure was approximately 90% for both barbiturates. The barbiturate residue was then dissolved and briefly mixed by vortexing in 25 ILl 0.2 M trimethylanilinium hydroxide methanol solution (TMAH) (Pierce Chemical Company) and 224 ILl methanol. This solution was injected onto the GLC column in a volume of 3 ILl using a Hamilton syringe. The GC (Hewlett-Packard #5730A) was equipped with a flameionization detector and a Recordall (Fisher, Series 5,000) chart recorder. The column conditions were: 3% OV-17 on Chromasorb W (length 7 feet, i.d. 2 mm), beginning temperature 140 ° programmed to 200 ° at 8°/minute, injection port temperature 300 °, which promotes the in-port derivatization reaction. RESULTS
The in vivo response of LS and SS mice following DB administration is shown in Table 1. These data consist of the LS and SS latency to loss of righting response (LLRR) and the duration of loss of righting response (DLRR). Thus a 150 mg/kg IP dose of DB resulted in a significantly shorter LLRR in LS than in SS mice ( L S = I I . 4 _ 1.1 minutes, S S = 2 3 . 1 _ 1.9 minutes). Additionally, this barbiturate induced a significantly longer DLRR in LS (296.0__.63.0 min) mice as compared to that of SS (20.0---5.2) mice (Student's t-test, p<0.05). The in vivo effect of SB is illustrated in Table 2, The SB data consist of the LS and SS latency to loss of righting response (LLRR) and the duration of loss of righting response (DLRR). Thus a 75 mg/kg IP dose of this barbiturate resulted in an equal LLRR ( L S = 3 . 0 ± 0 . 2 min, SS=3.5---0.5 min). However, in the physiological measure of CNS sensitivity, SB induced a significantly longer DLRR in SS (284.0---28.3 min) mice as corn-
BARBITURATE EFFECTS IN LS AND SS MICE
463
TABLE 1
Line LS SS
•
2.s
DIETHYLBARBITAL (DB) PHYSIOLOGICAL DATA*
LLRR
DLRR
11.4 -+- l.l.'t 23.1 z 1.9
296.0 ~ 63.0"I" 20.0-+ 5.2
Loss
2.0 JO
ixl o
1.5
¢-
~ Each data point represents the mean -+ S.E.M. of 4-6 determinations performed on 2 or 3 different days. *LLRR = latency to loss of righting response, DLRR = duration of loss of righting response, both measured in minutes. "tSignificantly different from the corresponding SS value, p<0.05. unpaired Student's t-test.
' ~
1.0 0.5 0.0
kS
SS Line
pared to that of LS (198.0--17.6 min) mice (Student's t-test, p<0.05). Figure 1 illustrates the brain DB concentration at loss of fighting response and at regaining the righting response following a 150 mg/kg IP dose of DB. That LS mice have greater CNS sensitivity to DB than SS mice is illustrated by a significantly lower brain concentration at loss of righting response in LS ( 0 . 9 2 - + 0 . 0 9 × 1 0 - 3 I-tg D B / m g brainl c o m p a r e d to SS (1.44+-0.05 × 10-3 ixg DB/mg brain) (Student's t-test, p<0.05). However, the DB brain concentration at which LS and SS regain the righting response is the same in the two lines (LS: 1.79+-0.20x 10 -3 la,g DB/mg brain, SS: 1.88+-0.14x 10 - 3 ~g DB/mg brain). Figure 2 illustrates the time course of DB brain distribution in LS and SS mice. This figure illustrates that the DB brain distribution rate is the same in the two lines, especially during the period of distribution leading to loss of the righting response. Therefore, LS mice, which have a shorter LLRR and a longer DLRR induced by DB, but have an equal rate of DB distribution to that of SS mice, thus have greater central sensitivity to DB as measured by a lower DB brain concentration at loss of the righting response compared to that in SS mice. In addition, the brain DB elimination rates do not differ in these lines during the time course of a mean duration of loss of fighting. Figure 3 illustrates the brain SB concentration at loss of righting response and at regaining the righting response following a 75 mg/kg IP dose of SB. That LS and SS mice do not have significantly different CNS sensitivity to SB is illustrated by equal brain concentration at loss of righting response LS ( 4 . 0 9 + - 0 . 3 3 x 1 0 -'~ I.,tg S B / m g brain) c o m p a r e d to SS (4.32+-0.02× l0 - 4 ~g SB/mg brain). These data (Fig. 3) indicate that the SB brain elimination rate is different in the two lines [LS, "regain" (198 minutes following LLRR)= 1.6 +- 0.29 ×
FIG. I. Brain diethylbarbital (DB) concentration (× 10 -3. measured in Ixg DB/mg brain wet weight) at loss of fighting response (loss) and at regaining the righting response (regain) following a 150 mg/kg IP dose of DB. Each data point represents the mean ± S.E.M. of 5-10 determinations performed on 2 or 3 different days. Note that LS Loss= I 1.4-- I.I and Regain=296.0--63.0 minutes; whereas SS Loss= 23. I _+ 1.9 and Regain = 20.5---5.2 minutes. *Significantly different from the corresponding SS value, p<0.05, unpaired Student's t-test.
10 - 4 p~g SB/mg brain: SS " r e g a i n " (284 minutes following L L R R ) = 2 . 1 8 - + 0 . 3 0 × 10 -8 p,g SB/mg brain]. Thus, SS mice, which have a longer DLRR induced by SB, have a slower rate of SB clearance than do LS mice. Other studies have demonstrated that the LS and SS mice differ in lipid-soluble barbiturate clearance rates ( 16,21 ). Considering that administration route and composition of the administration vehicle, and thus activation of adrenocortical response systems, may play a role in differential responses of the LS and SS mice (25), we employed a 50% distilled water/50% propylene glycol injection vehicle in which the barbiturates were administered. We employed this vehicle to decrease the discom-
LS Loss ~
Line
LLRR
DLRR
1.5"
1.00.5-
0.0
• I
20 LS
3.0 -+- 0.2
198.0 -+- 17.6"t
SS
3.5 - 0.5
284.0 +-- 28.3
Each data point represents the mean _ S.E.M. of 4-6 determinations performed on 3 or 4 different days. *LLRR = latency to loss of fighting response. DLRR = duration of loss of fighting response, both measured in minutes. 1"Significantly different from the corresponding SS value, p<0.05, unpaired Student's t-test.
LS R e g a l n ~
Regain
2.0-
a o ¢
TABLE 2
SS
2.5-
CO
SECOBARBITAL (SB) PHYSIOLOGICALDATA*
SS Loss
3.0-
•
I
40
"
I
60
•
I
•
I
•
I
•
I
"
l
'
$s I
•
I
•
I
80 1 0 0 1 2 0 1 4 0 1 6 0 1 8 0 2 0 0 2 2 0 Minutes
FIG. 2. Brain diethylbarbital (DB) concentration in LS and SS mice (measured in p,g DB/mg brain wet weight) as a function of time following a 150 mg/kg IP dose of DB. For minutes 0-30, samples were collected every 5 minutes; 30-60, every l0 minutes; and 60-120, every 30 minutes. Each data point represents the mean+-S.E.M, of 4-6 determinations performed in 3 different experiments.
464
DUNCAN AND RUTH
5.0 4.5 4.0 .Q 3.5 .~3.o al 2.5 c
m 2.o
'o
1.0
" 0.5 0.0 LS
SS
Line FIG. 3. Brain secobarbital (SBI concentration ( x l0 -4, measured in p~g SB/mg brain wet weight) at loss of righting response (Iossl and at regaining the righting response (regain) following a 75 mg/kg IP dose of SB. Each data point represents the mean - S.E.M. of 5-8 determinations performed in 3 different experiments. Note that LS Loss = 3.0__.0.2. and Regain=198.0_ 17.6 minutes, whereas SS Loss=3.5---0.5 and Regain = 284.0 ___28.3 minutes.
fort and possible confounding changes in neurochemistry associated with IP administration, and to enhance the chemical stability of the barbiturate injection solution. Previously, Osterlind et al. (18) have reported that similar propylene glycol doses had no effect on the naive animal EEG, but did lower the threshold dose for perturbation of the EEG recorded in the presence of hexobarbital. Therefore, the possibly confounding effect of the 50% distilled water/50% propylene glycol injection vehicle on barbiturateinduced sleep time was assessed by studying the effects of administration of this solution alone to these mice. This dose of propylene glycol (approximately 4.0 g/kg) was ineffective in producing significant hypothermia (0.5 ± 0. I°C loss) measured at 15 minutes postinjection, or in inducing loss of righting response in both lines. DISCUSSION
The purpose of the current study was to evaluate and compare A) duration of loss of righting response, and B) brain barbiturate concentrations at loss of righting as measures of central sensitivity to both water-soluble and lipid-soluble barbiturates. The in vivo measures which have been employed as an index of differential sensitivity to sedative/hypnotics are considered to be inherently variable (19). Considering that this inherent variability is probably a function of the complex physiological sequelae following EtOH exposure which results in these in vivo responses, this study does not attempt to control for all the possible variables, but does suggest that loss of righting response, when combined with quantification of brain concentrations of the pharmacological agent, does provide an accurate index of differential central sensitivity. The results of examining both waterand lipid-soluble barbiturates under the same laboratory conditions verify that LS mice have greater CNS sensitivity to the water-soluble barbiturate DB than do SS mice as illustrated by a significantly lower brain concentration at loss of righting response, a shorter LLRR, as well as a significantly longer duration of sedation induced by a 150 mg/kg IP dose of DB. That a shorter LLRR and a longer DLRR is induced in LS following IP DB is due to differences in neuronal sensitivity to the sedative effects of DB and is not due to differences in brain distribution or elimination is shown in Fig. 2, where these brain barbiturate
concentration time-course functions are the same in the two lines. Thus the conclusion that LS have greater CNS sensitivity to the water soluble barbiturate DB than SS is supported by both brain concentration of the barbiturate at loss of righting and by the DLRR. These data support the suggestions of Marley et al. (15) that water-soluble barbiturates may have a common mechanism of CNS depressant action with ethanol and other sedativehypnotic agents that have similar physico-chemical properties including the octanol/water partition coefficient, and that these agents may thus be of value in elucidating the mechanisms by which ethanol elicits these depressant effects. The observation that brain DB concentration in LS and SS at waking do not differ between these lines (Fig. I), may reflect a time-dependent redistribution of the drug into brain compartments not involved in the pharmacological effect. That the distribution of barbiturates is not simply a function of log P has been recently elucidated (22), which suggests that the particular functionality of the barbiturate may be an important consideration for pharmacokinetic and (possibly) pharmacological studies on the effects of these agents. Furthermore, the greater brain concentration of DB at waking than at the loss of righting in both lines suggests that the development of acute tolerance to the effects of the water-soluble barbiturate occurs. Indeed, Tabakoff et al. (231 have studied the phenomenon of acute tolerance to the sedative effects of EtOH in the LS and SS mice, and these investigators reported that the lines did not exhibit acute tolerance following IP EtOH. However, even though the current study was performed by quantifying brain barbiturate concentrations, a measure which necessitates the use of two sets of animals to generate the loss and waking data, the development of within-session tolerance to the effects of barbiturates may further confound the determination of relative central sensitivities when using brain concentration at the return of the righting response as an index. Thus it may be more accurate (pharmacologically) as well as being more convenient to use barbiturate concentration at loss of righting response rather than at regaining righting response as a measure of central sensitivity. Based solely upon duration of loss of righting response, one would conclude that LS and SS mice differ in sensitivity to the soporific effects of SB with SS mice having greater sensitivity. However, in contrast to the observations made with the watersoluble barbiturate, this conclusion is inconsistent with the SB biochemical data which demonstrate no between-line differences exist in CNS sensitivity to SB as measured by brain concentrations at loss of righting response in LS and SS mice. Because these two measures of central sensitivity are divergent, disagreements have arisen as to whether, in fact, the LS and SS lines differ in CNS sensitivity to these sedative hypnotics. However, our data suggest that brain clearance of SB in SS mice is much slower than in LS. as has been previously observed for clearance of the lipid-soluble barbiturate pentobarbital by O'Connor et al. (16). These investigators have attributed this differential rate of lipid-soluble barbiturate clearance from brain to the significantly greater body fat content of the LS mouse, therefore facilitating ready redistribution of the barbiturate from the circulation of LS mice. These differential elimination rates account for the differences in sleep time in the two lines. Thus, considering this pharmacokinetic difference, central sensitivity to lipidsoluble barbiturates can only be accurately assessed with brain concentration of the sedative agent producing loss of righting response. Furthermore, the brain concentrations of SB are on the same order as those reported for the relatively lipid-soluble barbiturate pentobarbital (16). Interestingly, the sedative-producing concentrations of DB were an order of magnitude greater than those of SB. Clearly, this reflects differences in the physico-chemical
BARBITURATE EFFECTS IN LS AND SS MICE
465
properties of the two barbiturates, properties which undoubtedly influence the partitioning of the agents to their respective sites of action. Support for this hypothesis is provided by reports in which the effect of various barbiturates on 3,-aminobutyric acid (GABA) binding was studied (17,24). These investigations have demonstrated that phenobarbital does not enhance GABAergic binding, whereas barbiturates characterized as being relatively more lipid-soluble do have an effect on binding to chloride-ionophore/GABA receptor complex (17,24). Numerous investigations have reported line differences between the LS and SS mice in GABAergic system activity or regulation thereof (14), as well as differences in in vitro and in vivo binding assessments of the chloride-ionophore/GABA receptor complex [e.g., (I)]. However, between-line differences in other neurochemical systems exist and barbiturates may also interact with these systems to induce sedation with a mechanism which is a function of, or independent of log P. In summary, this study supports the hypothesis that LS and SS mice, which have been selected for differential CNS sensitivity to ethanol, also have divergent central sensitivity to the water-soluble barbiturates represented in this study by DB but not the lipid-soluble barbiturates (e.g., SB). These conclusions are consistent with other studies which have employed numerous central depressants (2, 7, 15). Therefore, as Marley et al. (15) previously concluded, this differential sensitivity of LS and SS mice, as measured by divergent sleep time, to depressants depends largely on the specific physico-chemical properties of the depressant, and not the general structural category of the pharmacologic agent (ethanol vs. other depressants).
The method used for the measurement of brain barbiturate concentrations employs techniques that are elegant and generally applicable to the barbiturates. Using this GC method combining solid phase extraction techniques and a facile derivatization reaction, the full range of lipid-soluble, water-soluble, and epileptogenic barbiturates can be assayed from the brain. The sensitivity and specificity of this technique depends on the chromatographic behavior of the drug molecules on a gas-liquid column and is thus highly sensitive and specific. In addition, while performing this assay in greater than 100 experiments, we have observed a high degree of reproducibility with small variability as measured by standard drug peak area analysis (unpublished observations). Numerous analytical techniques, including liquid-scintillation spectrometry (12), UV-spectrophotometric (5) and GC methods (4, I 0), have been employed in the quantification of barbiturates from various biological samples. However, the advantage of this assay is that it has a high degree of selectivity and is generally more facile because of the solid phase extraction combined with gas-chromatographic techniques. Thus, as is possible with radiolabelled agents or in assays depending on chromophore behavior above background, the metabolites do not confound the concentration measurements derived with this technique. ACKNOWLEDGEMENTS This work was supported by USPHS grant AA-03527. Partial support was also provided by research awards from Sigma Xi and the Upjohn Company (C.C.D.). The authors also wish to thank Dr. A. C. Collins for valuable discussions.
REFERENCES 1. Allan, A. M.; Harris, R. A. Gamma-aminobutyric acid and alcohol actions: Neurochemical studies of long sleep and short sleep mice. Life Sci. 39:2005-2015; 1986. 2. Alpern, H. P.; Mclntyre, T. D. Evidence that the selectively bred long- and short-sleep mouse lines display common narcotic reactions to many depressants. Psychopbarmacology (Berlin) 85:456--459; 1985. 3. Alpern, H. P.; Mclntyre, T. D. Sedative-hypnotic anomalies related to dose of pentobarbital in long-sleep and short-sleep selectivelybred mice. Pharmacol. Biochem. Behav. 25:333-336; 1986. 4. Brochmann-Hanssen, E.; Oke, T. O. Gas chromatography of barbiturates, phenolic alkaloids, and xanthine bases: Flash-heater methylation by means of trimethylanilinium hydroxide. J. Pharm. Sci. 58(3):370-371; 1969. 5. Broughton, P. M. G. A rapid ultraviolet spectrophotometric method for the detection, estimation and identification of barbiturates in biological material. Biochem. J. 63:207-213; 1956. 6. Collins, A. C. A review of research using short-sleep and long-sleep mice. In: McClearn, G. E.; Dietrich, R. A.; Erwin, V. G., eds. Development of animal models as pharmacogenetic tools. DHHS publication #81-1133. Washington, DC: US Government Printing Office; 1981:161-170. 7. Erwin, V. G.; Heston, W. D. W.; McClearn, G. E. Effect of hypnotics on mice genetically selected for sensitivity to ethanol. Pharmacol. Biochem. Behav. 4:679-683; 1976. 8. Howerton, T. C.; O'Conner, M. F.; Collins, A. C. Lipid solubility is correlated with hypnotic and hypothermic responses of long-sleep (LS) and short-sleep (SS) mice to various depressant drugs. J. Pharmacol. Exp. Ther. 227:389-393; 1983. 9. Kananen, G.; Osiewicz, R.; Sunshine, I. Barbiturate analysis--A current assessment. J. Chromatogr. Sci. 10:283-287; 1972. 10. Kim, C.; Berg, R.; Khanna, J. M. Rapid and sensitive gas chromatographic method for detection of barbital and pentobarbital in blood using flash-heater methytation and nitrogen specific detection. Chromatographia 23:269-272; 1987. II. Leo, A.; Hansch, C.; Elkins, D. Partition coefficients and their uses. Chem. Rev. 71:525-615; 1971. 12. McClearn, G. E.; Kakihana, R. Selective breeding for ethanol sen-
13.
14. 15. 16. 17. 18. 19. 20.
21. 22. 23.
sitivity: Short-sleep and long-sleep mice. In: McClearn, G. E.; Dietrich, R. A.; Erwin, V. G., eds. Development of animal models as pharmacogenetic tools. DHHS publication #81-1133. Washington, DC: U.S. Government Printing Office; 1981:147-159. Mclntyre, T. D.; Alpern, H. P. Thiopental, phenobarbital, and chlordiazepoxide induce the same differences in narcotic reaction as ethanol in long-sleep and short-sleep selectively bred mice. Pharmacol. Biochem. Behav. 24:895-898; 1986. Marley, R. J.; Freund, R. K.; Wehner, J. M. Differential response to flurazepam in long-sleep and short-sleep mice. Pharmacol. Bitchem. Behav. 31:453--458; 1988. Marley, R. J.; Miner, L. L.; Wehner, J. M.; Collins, A. C. Differential effects of central nervous system depressants in long-sleep and short-sleep mice. J. Pharmacol. Exp. Ther. 238:1028-1033; 1986. O'Connor, M. F.; Howerton, T. C.; Collins, A. C. Effects of pentobarbital in mice selected for differential sensitivity to ethanol. Pharmacol. Biochem. Behav. 17:245-248; 1982. Olsen, R. W.; Snowman, A. M. Chloride dependent enhancement by barbiturates of gamma-amino butyric acid receptor binding. J. Neurosci. 2:1812-1823; 1982. Osterlind, A.; Akesson, A.; Wahlstrom, G. Interactions between 1,2-propanediol (propylene glycol) and hexobarbital. Acta Pharmacol. Toxicol. 45:245-248; 1979. Phillips, T. J.; Feller, D. J.; Crabbe, J. C. Selected lines, alcohol and behavior. Experientia 45:805-827; 1989. San-Marina, A.; Khanna, J. M.; Kalant, H. Relationship between initial sensitivity, acute tolerance and chronic tolerance to ethanol in a heterogeneous population of Swiss mice. Psychopharmacology (Berlin) 99:450--457; 1989. Siemens, A. J.; Chan, A. W. K. Differential effects of pentobarbital and ethanol in mice. Life Sci. 19:581-590; 1976. Steiner, S. H.; Moor, M. J.; Bickel, M. H. Kinetics of distribution and adipose tissue storage as a function of lipophilicity and chemical structure. I. Barbiturates. Drug Metab. Dispos. 19:8-14; 1991. Tabakoff, B.; Ritzmann, R. F.; Raju, T. S.; Deitrich, R. A. Characterization of acute and chronic tolerance in mice selected for inherent differences in sensitivity to ethanol. Alcohol.: Clin. Exp.
466
Res. 4:70-73; 1980. 24. Whittle, S. R.; Turner, A. J. Differential effects of sedative and anticonvulsant barbiturates on specific [3H]GABA binding to membrane preparations from rat brain cortex. Biochem. Pharmacol. 31:
DUNCAN AND RUTH
2891-2895; 1982. 25. Zgombick, J, M,; Erwin, V, G. Central mechanisms of ethanol-induced adrenocortical response in selectively bred lines of mice. Neuroendocrinology 46:324--332: 1987.