- Email: [email protected]
Purinergic modulation of ethanol-induced sleep time in long-sleep and short-sleep mice
Alcohol. Vol. 8 pp. 123-130. o Pergamon Press plc. 1991. Printed in the U.S.A.
07,~8,1q9191 $3.00 + .00
Purinergic Modulation of Ethanol-Induced Sleep Time in Long-Sleep and Short-Sleep Mice TONI NESS SMOLEN AND ANDREW
S M O L E N * i "1
hlstitute for Behavioral Genetics *Universit 3, of Colorado Drug Abuse Research Center and "/School of Pharmacy, Campus Box 447, Universit3, of Colorado, Boulder, CO 80309-0447 R e c e i v e d 26 O c t o b e r 1989; A c c e p t e d 22 O c t o b e r 1990
SMOLEN, T. N. AND A. SMOI'~EN. Purinergic modulation of ethanol-induced sleep rime in long-sleep and short-sleep mice. ALCOHOL 8(2) 123-130, 1991.--The long-sleep (IS) and short-sleep (SS) mice were selectively bred for differences in sensitivity to the depressant effects of ethanol. In addition to their differential sensitivity to ethanol, they are also differentially sensitive to purinergic agonists and antagonists. This suggests that there may be differences in the purinergic systems of these lines of mice which may aid in understanding how they differ in ethanol sensitivity. We have investigated whether these drugs are capable of modifying acute ethanol sensitivity as measured by ethanol-induced loss of the righting response (ethanol sleep time), waking blood and brain ethanol concentrations, and blood ethanol elimination rate. The purinergic agonists cyclohexyladenosine (CHA/, Lphenylisopropyladenosine (PIA), 2-chloroadenosine (CAD), and N-ethylcarboxamidoadenosine (NEC) increased sleep time in both LS and SS mice. however, LS mice were generally more affected than SS. The LS and SS mice were also differentially sensitive to the purinergic antagonists, theophylline and caffeine. Blood and brain ethanol concentration on awakening suggested that CNS sensitivity to acute ethanol administration was altered by pretreatment with agonists but not antagonists. Two agonists, CHA and NEC, significantly lowered ethanol elimination in both lines of mice while PIA, CAD, and the antagonists theophylline, and caffeine were without affect on elimination rate. These data support previous observations that adenosine-mediated systems may be involved in the modulation of ethanol sensitivity. Ethanol Sleep time Waking blood ethanol Long-sleep mice Short-sleep mice
Waking brain ethanol
Purinergic drugs
Adenosine
of ethanol-adenosine interactions focused on the A~ receptor subtype, however, recent studies suggest an important role for the A 2 receptor as well (14-16). Evidence supporting an interaction between ethanol and adenosine receptor systems comes from studies of drugs which are known to bind to specific adenosine receptor sites. For example, methylxanthine compounds, such as caffeine and theophylline, are adenosine receptor antagonists and have been found to modify ethanol responses in mice following both acute and chronic administration of ethanol (2, 3, 1 t). While there is good support for the hypothesis that adenosinemediated systems may be involved in modulating some of the CNS effects of ethanol, this notion is strengthened by comparisons of purinergic drug response in genetic stocks of mice selectively bred for differences in initial sensitivity to ethanol. Proctor and Dunwiddie (20,21) reported that behavioral sensitivity to purinergic drugs paralleled ethanol sensitivity in long-sleep (LS) short-sleep (SS) mice. Using a variety of behavioral tests, these investigators showed that LS mice were more sensitive to the potent A t agonist, L-phenylisopropyladenosine than were SS mice.
THE purine nucleoside adenosine is an important neuromodulator of central nervous system (CNS) excitability (5, 17-19). Adenosine-mediated systems have been postulated to be involved in the expression of some of the CNS effects of ethanol. Although adenosine and ethanol are structurally unrelated, both share sedative, hypothermic, and anticonvulsant properties, and acute administration of either ethanol or purinergic agonists produces similar behavioral and physiological responses (2, 6, 29, 31). In addition to their common properties, several lines of evidence suggest that ethanol may interact with an adenosine receptor system. Adenosine exerts its actions through multiple subclasses of receptors. These have been classified as A~, A 2 and P-site receptors, but multiple conformations may occur within each receptor subtype as well (1, 30, 32) The A~ receptor reduces or inhibits adenylate cyclase activity, whereas the A2 receptor enhances or activates adenylate cyclase activity. The P-site receptor has not been as well characterized as the A~ and A., receptors, but it appears to be located intracellularly and agonist binding is associated with decreased adenylate cyclase activity (12). Early studies
'Requests for reprints should be addressed to Dr. Toni N. Smolen, Institute for Behavioral Genetics, Campus Box 447, University of Colorado, Boulder, CO 80309-0447.
123
124
SMOLEN AND SMOLEN
In addition to a differential response to the adenosine agonist, they also reported that LS mice were more sensitive to the adenosine receptor antagonist theophylline than were SS mice. These data suggested that there may be differences in the purinergic systems of these lines of mice. In order to extend these findings, we have taken an alternate approach, namely, to investigate the effect of adenosine receptor agonists and antagonists on ethanol responses in LS and SS mice. In this study we examined whether purinergic drugs could modulate acute sensitivity to ethanol as measured by ethanol sleep time, waking blood and brain ethanol levels, and ethanol elimination rate. METHOD
Experimental Animals Animals used in this study were male and female LS and SS mice (13), 60 to 80 days of age. Mice were bred at the Institute for Behavioral Genetics, maintained on a 12-hour light cycle, and were allowed free access to food (Wayne Lab Blox) and water. All experiments were begun between 0900 and 1000 hours. All procedures described in this paper were approved by the University of Colorado Animal Care and Use Committee as being consistent with USPHS standards of humane care and treatment of laboratory animals.
Drug Solutions All drugs were administered intraperitoneally. Ethanol was prepared as a 20% (w/v) solution in saline. Doses were varied by adjusting the volume injected, because of the marked concentration-dependent effects of ethanol on sleep time (10). The following purinergic drugs were used in these studies: cyclohexyladenosine (CHA), L-phenylisopropyladenosine (PIA), 2chloroadenosine (CAD), N-ethylcarboxamidoadenosine (NEC), theophylline and caffeine. All were dissolved in saline at a concentration such that the injection of 0.01 ml of solution per gram body weight would result in the dose indicated in the figures. The doses and timing used were chosen from previously published reports and from our own preliminary dose-response and timecourse experiments that examined the behavioral and physiological effects of these agonists and antagonists in LS and SS mice (manuscript submitted).
Determination of Sleep Time, Blood and Brain Ethanol Content, and Ethanol Elimination Rate Purinergic drugs or saline were administered 15 min before ethanol. Sleep times were determined by injecting LS mice with 2.5 g ethanol/kg body weight, and SS mice with 5.0 g ethanol/ kg body weight. These are iso-effective doses resulting in sleep times of approximately 60 min. Following ethanol administration the animal was placed on its back in a V-shaped holder until it was able to right itself 3 times in a 30 s period (26,28). At the time of awakening, a 10 p,I blood sample was collected from the retroorbital sinus (23), and the whole brain was removed for measurement of ethanol concentration. Ethanol concentration was measured enzymatically as detailed previously (26,28). Blood and brain samples were collected from mice that failed to lose the righting response at 30 min post ethanol injection, which corresponds to the end of the ethanol distribution phase (27). These mice were given a sleep time score of zero and were included in the sleep time data analysis following verification that they had received their full dose of ethanol. Their corresponding 30-min blood and brain values were not included in the overall analyses
since these do not reflect waking ethanol levels. Sleep time experiments were terminated after 4 hours because we have observed virtually 100% mortality in agonist pretreated mice that had not regained the righting response by this time. Mice which had not awakened by that time were given a score of 240 min, and blood and brain samples were taken for ethanol determinations. Data from these animals are presented separately but were not included in the figures or in our analyses because they, too, do not reflect true waking times or ethanol levels at awakening. Ethanol elimination rate was measured by taking a 10 p.l blood sample 0.5, I, 2, 3 and 4 hours after ethanol administration. Blood ethanol concentration was determined, and the linear decline in ethanol content was calculated as described previously (25). None of the purinergic drugs, when administered alone at the doses used in the present study, caused the animals to lose the righting response.
Data Analysis Data were analyzed by one- or two-way analysis of variance (ANOVA) as appropriate. The LS and SS mice were given different doses of ethanol which are iso-effective with respect to sleep time, therefore the sleep time data were analyzed as a twoway ANOVA using line and drug as between subjects factors. Since the two lines have significantly different ethanol thresholds for regaining the righting response (261 the waking blood and brain ethanol levels are always significantly different; therefore, these data were analyzed separately for each line using dose as the single between-subjects factor. The data for the purinergic agonists were analyzed separately from the the data for the purinergic antagonists. Following a significant overall effect, differences in individual group means were detected using Duncan's Multiple Range test. A p value of 0.05 was considered significant, and is the only level reported. Data presented in the figures have been collapsed across sex due to lack of significant differences between males and females. RESULTS
Figures 1 through 3 show the effect of pretreatment with purinergic agonists on ethanol-induced sleep time and waking blood and brain ethanol concentrations in LS and SS mice. Pretreatment with the A~ agonist CHA increased ethanol-induced sleep time (Fig. l) and decreased ethanol elimination rate (Fig. 7) in both lines of mice. Two-way ANOVA of the sleep time data showed significant main effects of line, F(1,100)=5.46, p<0.05, and dose, F(3, lO0) = 34.8, p<0.05, and a significant line-by-dose interaction, F(3,100)= 5.18, p<0.05, which resulted from the LS mice being affected to a greater degree by CHA than SS mice. Waking blood and brain ethanol levels were lower in CHA-treated LS mice compared to saline-pretreated controls, F(3,48)=21.3 and 12.6, respectively, p<0.05 for both, which is suggestive of a CHA-mediated alteration in CNS sensitivity to ethanol. In contrast, while CHA-pretreated SS mice had lowered waking blood ethanol levels, the waking brain ethanol levels were also lower (not significantly) from saline-pretreated controls, F(3,52)= 6.77 and 2.30, p<0.05 and NS, respectively. Approximately equal numbers (45%) of LS and SS mice tested at 0.3 and 1 mg/kg CHA did not regain the righting response within the four-hour time limit and were given the maximum sleep time score of 240. Blood and brain ethanol levels for these animals are presented in Table 1 and support the finding of diminished blood ethanol elimination following CHA pretreatment. Pretreatment with another A~ agonist, PIA, caused a dose-de-
PURINERGIC DRUGS AND ETHANOL SENSITIVITY
200 :
~
lLS Mice I - - I SS Mice
125
*
200
_E . . 15o
150
.E
-
-~
mE
o
i :f
4.50
g~
• ~ :r,ool-
i...
=o_.
/n in|
4,5O
e,m~3oo 150 0
0
.1 .3 1.0 Dose of CHA (mg/kg)
FIG. I. Sleep times and waking blood and brain ethanol levels in LS and SS mice following pretreatment with cyclohexyladenosine (CHAI. Values are mean ± SEM of 10-24 mice per group. *Significantly different from respective saline-treated control, p<0.05. **Significantly different from LS mice, p<0.05.
pendent increase in ethanol-induced sleep time, 1::(3,109)= 19.2, p<0.05, and the LS mice were more affected than the SS (Fig. 2). However, PIA was less effective than CHA in differentiating between the LS and SS mice since the main effect of line failed to reach statistical significance. Waking blood and brain ethanol levels were decreased in LS mice again suggesting an alteration in CNS sensitivity to ethanol resulted from pretreatment with the A t agonist. Although PIA-pretreated SS mice slept longer, they awoke at blood and brain ethanol levels which were the same as saline-pretreated controls. Figure 7 shows that PIA was also less potent than CHA in reducing ethanol elimination rate in both lines of mice. Figure 3 shows the results of pretreatment of mice with the mixed A2/A ~ agonist NEC (NEC has more A, than Aj activity whereas the other mixed agonist, CAD has more Al than A 2 activity). NEC is a very potent drug with a steep dose response curve (Fig. 3). A 33-fold lower dose produced ethanol sleep times that were the same as, or greater than, that seen with the more specific A~ agonists CHA and PIA. Doses of NEC above 0.03
II
n n n n H
II
~F
,,+;~
~¢
T*
~
/11 /
150
o
._c o °
*
_a_,
lfi
0
•~ - -
*,
lOO 50
"loo
B B L S Mice I-"-I SS Mice
H
i
H
/I .
iI/1 ilill ~
0
.1 .3 1.0 Dose of PIA (mg/kg)
FIG. 2. Sleep times and waking blood and brain ethanol levels in LS and SS mice following pretreatment with L-phenylisopropyladenosine (PIAI. Values are mean---SEM of 10-24 mice per group. *Significantly different from respective saline-treated control, p<0.05.
mg/kg produced ethanol sleep times in excess of 240 min and virtually 100% mortality. The LS and SS mice did not differ in either their sleep time response to NEC (Fig. 3) or the significant decrease in ethanol elimination rate produced by NEC pretreatment (Fig. 7). Waking blood and brain ethanol concentrations were significantly different from control levels only at the highest dose of NEC (Fig. 3). The mixed AI/A 2 agonist, CAD increased sleep time in LS but not SS mice, with no effect on waking blood or brain ethanol levels (Fig. 6) or ethanol elimination rate in either line of mice (Fig. 7). Figure 4 shows that pretreatment with the purinergic antagonist theophylline resulted in effects opposite to those of the purinergic agonists. Theophylline decreased ethanol sleep time significantly. Although the overall pattern of response was similar in both lines of mice, as was found for each of the agonists, LS mice were more affected by theophylline pretreatment than were SS mice. Two-way ANOVA revealed significant main effects of line, F(I,108)=9.47, p<0.05, and dose, F(3,108)= 10.56, p<0.05, with no significant line by dose interaction 1::(3,108) = 1.96, NS. In spite of the marked effect on sleep time,
126
SMOLEN AND SMOLEN
200
I
~ L S Mice [---]SS Mice
*
100
,
•
Ill
I
~
~@o 'roof
II
•
~E
"="
_
r~
II
II
""
..
r=l ""
%_
II
"
;,
I ~
~E
~ L S Mice I--ISS Mice
I
1
lit
II III Ill ill I
r
i ill"
°
,I iIi
,,,E
~@oo 4s° ~E
0
**
0
-6 " o° ~
I'1
°~'-
I
I
I
I
I
R
I
I
r'-,
I
I
"~'-
0
0
0
.005
.01
.O3
Dose of NEC (mg/kg) FIG. 3. Sleep times and waking blood and brain ethanol levels in LS and SS mice following pretreatment with N-ethylcarboxamidoadenosine(NEC). Values are mean - SEM of 8-24 mice per group. *Significantly different from respective saline-treated control, p<0.05.
theophylline did not alter waking blood or brain ethanol levels (Fig. 4) or ethanol elimination rate (Fig. 7) in either line of mice.
TABLE 1 BLOOD AND BRAIN ETHANOL LEVELS AT 240-MIN POSTETHANOL ADMINISTRATION IN CYCLOHEXYLADENOSINE PRETREATED LS AND SS MICE
Dose (mg/kg)
N
Sleep Time (Minutes)
LS
0 (saline) 0.3 1.0
21 9 10
57 --+ 5 240 240
247 +_. 10 173 --- 20 229 __. 16
250 +- 10 164 -)- 14 252 --- 18
SS
0 (saline) 0.3 1.0
24 8 10
57 --- 6 240 240
488 __- 15 351 _-. 19 451 _ 16
478 __- 19 395 .4- 14 475 - 34
Line
' - o° ' ~
Blood (mg/dl)
Tabled values represent the mean +__ SEM.
Brain ( m g / 1 0 0g)
0
,IiItl 10
20
40
Dose of THE (mg/kg) FIG. 4. Sleep times and waking blood and brain ethanol levels in LS and SS mice following pretreatment with theophylline (THE). Values are mean---SEM of 9-24 mice per group. *Significantly different from respective saline-treated control, p<0.05. **Significantly different from LS mice, p<0.05.
In contrast to theophylline, caffeine pretreatment differentially affected ethanol sleep time in LS and SS mice (Fig. 5). Pretreatment with caffeine decreased sleep time in the LS, but increased sleep time in the SS mice resulting in a significant main effect of line, F(1,131)=43.88, p < 0.05. The marked difference in response of the two lines to caffeine was also evident from the significant line by dose interaction, F( 1,131 ) = 12.15, p < 0 . 0 5 . Since caffeine had opposite effects on the sleep time response of the two lines, analysis of the dose effects essentially canceled each other, resulting in no significant main effect of dose, F(3,131)= 1.66, NS. The waking blood and brain ethanol data did not show a consistent pattern. Waking blood ethanol levels tended to be higher in the LS and lower in the SS mice, and within each line, waking brain ethanol levels were not significantly different from saline-pretreated controls. Although the sleep time data suggest an alteration in the metabolism of ethanol, a high dose of caffeine (40 mg/kg) did not affect ethanol elimination rate in either line of mice (Fig. 7). Figure 6 summarizes the effectiveness of the four agonists and
PURINERGIC DRUGS AND ETHANOL SENSITIVITY
100
127
ILS Mice I---I SS Mice
200
E
a
a
I/-S Mice I - - ] S S Mice
15o Eso
m
-~0 oo o,,-
25
50
°
0
"~ "'E
4.50
~E
i
15o ~=
o
- ~ p ~ 600 o
600
0
"~ ~ 600 ~ ~
•
~
"~,m'~
300
150
i
150 0
0 10 20 40 Dose of CAF (mg/kg)
FIG. 5. Sleep times and waking blood and brain ethanol levels in LS and SS mice follox¢ing pretreatment with caffeine (CAF). Values are mean +--SEM of 11-24 mice per group. *Significantly different from respective saline-treated control, p<0.05. **Significantly different from LS mice. p<0.05.
two antagonists in altering ethanol sleep time, waking blood and waking brain ethanol levels• The magnitude of the effects of the purinergic drugs was most pronounced at the highest dose of agonist and lowest dose of antagonist used. All of the agonists except NEC are compared at a dose of 1 mg/kg (for NEC a dose of 0.03 mg/kg was used), and the antagonists at 10 mg/kg. With the exception of caffeine, the differences between the LS and SS mice appear to be quantitative, rather than qualitative in nature: In general, all of the agonists increased, whereas the antagonists decreased ethanol-induced sleep times, but in each case the LS mice were more affected than the SS mice. Waking blood and brain ethanol levels tend to be more affected in the LS mice as well, but the magnitude of the effect of these purinea-gic drugs on ethanol concentrations was not large. DISCUSSION
The LS and SS mice were selectively bred for maximal differences in CNS sensitivity to the depressant effects of ethanol as
SAL CHA
PIA CAD NEC THE CAF
FIG. 6. Sleep times and waking blood and brain ethanol levels in LS and SS mice. Mice were pretreated with: SAL (saline, 0.01 ml/g body weight), CHA, 1.0 mg/kg; PIA, 1.0 mg/kg: CAD, 1.0 mg/kg: NEC, 0.03 mg/kg; THE, 10 mg/kg; or CAF, 10 mg/kg 15 min prior to receiving 2.5 (LS) or 5.0 (SS) g/kg ethanol. Values are mean--. SEM of 8-24 mice per group. (a) Significantly different from respective saline-treated control, p<0.05. (b) Significantly different from LS mice, p<0.05.
measured by loss of the righting response, or sleep time (13). We have previously shown that these lines of mice display a characteristic threshold value for ethanol which determines the point at which they regain the righting response, irrespective of the ethanol dose given (26). Threshold values for waking blood and brain ethanol concentrations were, respectively: 261 mg/100 ml and 230 mg/100 g (LS mice); and 502 mg/100 ml and 468 mg/100 g (SS mice)• Knowing these threshold values is useful for interpreting acute or chronic drug effects on ethanol-induced sleep time. An altered ethanol sleep time may result from a change in the normal metabolism rate, from an alteration in CNS sensitivity, or both. A drug treatment that causes an animal to awaken at an ethanol concentration significantly lower or higher than the normal threshold value has induced a change in the CNS sensitivity to ethanol regardless of the drug's effect on the rate of ethanol metabolism. In the present study, we have investigated the ability of several adenosine agonists and antagonists to alter CNS sensitivity to ethanol in an attempt to characterize the role of this neuronal
128
SMOLEN AND SMOLEN
-4-.-I
D rY ._..
100
~ LS Mice [ ~ q S S Mice
k-
g,I..I. 75
"4~m
** z SAL
CHA
PIA
CAD NEC
THE
CAF
FIG. 7. Ethanol elimination rates in mice pretreated with adenosine agonists or antagonists. Mice were pretreated with: SAL (saline, 0.01 ml/g body weight), CHA. 1.0 mg/kg; PIA. 1.0 mg/kg; CAD, 1.0 mg/kg; NEC, 0.03 mg/kg; THE, 50 mg/kg; or CAF, 40 mg/kg 15 min prior to receiving 2.5 (LS) or 5.0 (SS) g/kg ethanol. Values are mean -- SEM of 7 mice per group. *Significantly different from respective saline-treated control, p<0.05. • system in the behavioral and physiological responses to ethanol. The results of this study confirm and extend previous reports which have shown that sensitivity to purinergic drugs parallels that of ethanol sensitivity (20,21) and that purinergic drugs are capable of modulating some of the acute effects of ethanol (2, 3, 20, 21). All of the drugs used in this study are thought to exert their actions by combining with one or more of the adenosine receptor subtypes. Although none of these drugs have absolute specificity for a single receptor subtype, CHA and PIA have greater specificity for the adenosine A~ receptor. CAD has approximately equal agonist activity at both A t and A2 receptors, whereas NEC, which has activity at both A_~ and A~ receptors, is commonly regarded as primarily an A2 agonist. The competitive antagonists, theophylline and caffeine, are thought to have nearly equal affinities for both receptor subtypes. In spite of this lack of specificity, theophylline and caffeine are important to study because they are found in abundance in commonly consumed beverages such as tea and coffee which are often used by people in an effort to counteract the effects of alcohol. Proctor and Dunwiddie (21) have reported that behavioral sensitivity to purinergic drugs paralleled ethanol sensitivity in LS and SS mice. They used several behavioral tests, including latency to- and number of escape attempts from an elevated platform, and both light-dark crossings and time spent in the light compartment of a light-dark box, to show that LS mice were more sensitive to PIA than were SS mice. All test responses were found to be decreased by PIA in both lines of mice, but the ECso values were four-fold lower for LS than SS mice, indicating a quantitative rather than qualitative difference in their response to this agonist. Our findings support these results. The LS mice were more affected by A~ agonists than were SS mice. Both LS and SS mice are extremely sensitive to CHA. None of the drugs used in these studies are, by themselves, lethal. However, when CHA is administered in combination with ethanol it is almost always lethal at doses which significantly affect sleep time. In the present study a ceiling of 240 min was placed on the sleep time period since in our preliminary studies we observed 100% mortality among CHApretreated mice that slept longer than four hours after ethanol administration. Romm and Collins (24) have previously shown that ethanol elimination rate changes in direct proportion to body temperature. Thus, the mortality seen following CHA and ethanol
administration could be due to the profound hypothermia CHA causes, followed by a substantial decrease in ethanol elimination rate. A 1.0 mg/kg dose of CHA produces a gradual loss in body temperature of approximately 11°C for LS and 8°C for SS mice over a 3-h period. This results in a nearly halving of the ethanol elimination rate in both lines of mice (Fig. 7). Threshold values for waking blood ethanol were significantly lowered in both lines following agonist pretreatment, however, concomitant brain ethanol values were significantly decreased in the LS line only. The greater variability of the SS brain ethanol values may have contributed to the lack of statistical significance. Given the nearly identical pattern of response, we would conclude that CHA alters CNS sensitivity to ethanol in both lines of mice, but that the major difference between the lines was quantitative in nature since for each drug at virtually every dose used the LS were more affected than the SS. PIA was less potent than CHA in exacerbating ethanol sleep time. Only 6% of PIA-treated, compared to approximately 45% of CHA-treated animals received a sleep time score of 240 min. Although the lines did not differ significantly in their sleep time response, our analyses of blood and brain ethanol levels on awakening suggested that CNS sensitivity to ethanol was altered by PIA in LS mice only. The finding that these mice slept longer at the highest dose of PIA (1 mg/kg) but awoke at their threshold values for blood and brain ethanol concentration is puzzling since this dose of PIA, unlike CHA, has no effect on ethanol elimination rate. It is possible that, at high doses, the sleep time response is due to some other physiological action of the drug. The At/A,- agonist CAD was also less potent in modulating sleep time response than CHA. None of the CAD-treated mice received a sleep time score of 240 min, and only LS mice were affected by this agonist. Since the animals awoke at their characteristic threshold values for blood and brain ethanol levels, CNS sensitivity to ethanol was not altered by CAD. Due to a lack of specific adenosine A2 receptor agonists and antagonists, there have been few in vivo studies of this receptor and its potential role in modulating ethanol actions. Although NEC also suffers from this lack of specificity, it is, nevertheless, one of the most selective of the few A2 agonists commercially available. Recent cell culture studies suggest that A2-dependent responses may be important in both the acute and chronic effects of ethanol (1416). It has been shown that acute exposure to ethanol increases the extracellular concentration of adenosine, which then activates A 2 receptors resulting in an increase in intracellular cAMP levels. It has been suggested that chronic ethanol exposure results in an heterologous desensitization of receptor-stimulated cAMP production which is caused by the extracellular accumulation of adenosine (14-16). At the present time the role this would play in the physiological response to acute ethanol administration (e.g., sleep time) is not known, however, it further implicates adenosine receptor systems in the actions of ethanol. In our studies we found that NEC was particularly potent at increasing ethanol sleep time. A dose of 0.03 mg/kg increased sleep time to near maximum (240 min). As with CHA, this could be related to the profound hypothermia caused by NEC when given alone or when coadministered with ethanol. This results in a markedly diminished elimination rate for ethanol (Fig. 7). Since NEC is not a pure A 2 agonist, it is unclear whether the effects of this drug are due to actions at the A 2 or Aj site, or both. Waking blood and brain ethanol concentrations were diminished in both lines of mice. The pattern of response leads us to conclude that CNS sensitivity to ethanol is altered in both lines of mice, but that the response is more robust in LS than SS mice. In addition to their differential responsiveness to adenosine agonists, Proctor and Dunwiddie reported that LS mice were more
PURINERGIC DRUGS AND ETHANOL SENSITIVITY
129
sensitive to the adenosine receptor antagonist theophylline compared to SS mice (20). In the present study we found that theophylline was far more effective in decreasing sleep time in LS (especially the 10 and 20 mg/kg doses) than in SS mice (lowest dose only). This decrease in ethanol-induced sleep time by theophylline also agrees with the work of Dar et al. (2) who reported that theophylline pretreatment delayed the onset and decreased the duration of ethanol-induced sleep time in ICR mice. In the present study, both lines of mice awoke at their characteristic threshold values for blood and brain ethanol concentrations suggesting there was no alteration in CNS sensitivity to ethanol. That theophylline pretreatment resulted in a decrease in sleep time without a change in waking ethanol levels, suggests an alteration in ethanol elimination rate. However, we found that ethanol elimination rate was not altered by theophylline pretreatments of 10 (not shown) or 50 mg/kg (Fig. 7). The ethanol elimination rates presented in Fig. 7 were based on blood samples taken at 0.5, 1, 2, 3, and 4 h following ethanol injection. Thus, the first blood sample was collected at 30 min, which is beyond the waking time of the LS (15-29 min, Fig. 4) and nearly that of SS mice (36-52 min). It is possible that shorter sampling times may have revealed an alteration in either the time to peak blood ethanol concentration or elimination rate which were not evident using our standard protocol. A striking difference between LS an'd SS sleep time response was seen following caffeine pretreatment. The decrease in ethanol sleep time in LS mice is nearly identical to that seen in theophylline-treated mice. However, SS mice responded to caffeine pretreatment with an increase in ethanol sleep time. The fact that waking brain ethanol values of caffeine-treated mice were not significantly different from controls, coupled with the lack of a consistent pattern of response among the blood values, suggests
that caffeine does not alter CNS sensitivity to ethanol in either line. Similarly, caffeine does not appear to alter ethanol elimination rate in either line of mice although, as with theophylline, sampling blood at earlier time points may be more informative. It is unclear what factors beyond an alteration in CNS sensitivity or a change in the rate of ethanol metabolism might contribute to the sleep time response seen following treatment with these antagonists. However, theophylline and caffeine are known to have effects independent of their actions at adenosine receptors which may play a role in their ability to modulate ethanol sleep time (7,8). In summary, the results of the present study support the hypothesis that adenosine systems may be involved in modulating some of the effects of ethanol. The mechanism by which this modulation occurs is unclear, but it is not likely that ethanol combines directly with an adenosine receptor. Ethanol may perturb the membrane microenvironment surrounding the receptors, or may exert its effects through an adenosine-mediated second messenger system as has been recently proposed by Nagy and coworkers (15,161. With respect to the LS and SS lines of mice, there may be differences in the numbers of adenosine receptors themselves, as has been suggested by Fredholm and co-workers (9). Whether ethanol sensitivity can ultimately be associated with such receptor differences remains to be determined. ACKNOWLEDGEMENTS The authors wish to thank Lorie Lockwood, Christine Ross and Dawn Waterman for their excellent technical assistance. This work was supported by grants AA 07127 (T.N.S.I from the National Institute on Alcohol Abuse and Alcoholism and HD 21709 (A.S.) from the National Institute of Child Health and Human Development.
REFERENCES 1. Bruns, R. F.; Daly, J. W.; Snyder, S. H. Adenosine receptors in brain membranes: Binding of N°-cyclohexyl[3H]adenosine and 1,3diethyl-8-[3H]phenylxanthine. Proc. Natl. Acad. Sci. USA 77:55475551; 1980. 2. Dar, M. S.; Mustafa, S. J.; Wooles, W. R. Possible role of adenosine in the.CNS effects of ethanol. Life Sci. 33:1363-1374; 1983. 3. Dar, M. S.; Jones, M.; Close, G.; Mustafa, S. J.; Wooles, W. R. Behavioral interactions of ethanol and methylxanthines. Psychopharmacology (Berlin) 91:1-4; 1987. 4. Dunwiddie, T. V. The physiological role of adenosine in the central nervous system. Int. Rev. Neurobiol. 27:63-139; 1985. 5. Dunwiddie, T. V.; Basile, A. S.; Palmer, M. R. Electrophysiological responses to adenosine analogs in rat hippocampus and cerebellum: Evidence for mediation by adenosine receptors of the A~ subtype. Life Sci. 34:37-47; 1983. 6. Dunwiddie, T. V.; Worth, T. Sedative and anticonvulsant effects of adenosine analogs in mouse and rat. J. Pharmacol. Exp. Ther. 220: 70-76; 1982. 7. Fain, J. N.; Pointer, R. H.; Ward, W. F. Effects of adenosine nucleosides on adenylate cyclase, phosphodiesterase, cyclic adenosine monophosphate accumulation, and lipolysis in fat cells. J. Biol. Chem. 247:6866-6872; 1972. 8. Fredholm, B. B. On the mechanism of action of theophylline and caffeine. Acta Med. Scand. 217:149-153; 1985. 9. Fredholm, B. B.; Zahniser, N. R.; Weiner, G. R.; Proctor, W. R.; Dunwiddie, T. V. Behavioral sensitivity to PIA in selectively bred mice is related to a number of A a adenosine receptors but not to cyclic AMP accumulation in brain slices. Eur. J. Pharmacol. 111:133136; 1985. 10. Gilliam, D. M.; Collins, A. C. Concentration-dependent effects of ethanol in Long-Sleep and Short-Sleep mice. Alcohol.: Clin. Exp. Res. 7:337-342; 1983. I 1. Hilakivi, L. A.; Durcan, M. J.; Lister, R. G. Effects of caffeine on social behavior, exploration and locomotor activity: Interactions with ethanol. Life Sci. 8:543-553; 1989.
12. Londos, C.; Wolff, J. Two distinct adenosine-sensitive sites on adenylate cyclase. Proc. Natl. Acad. Sci. USA 74:5482-5486; 1977. 13. McClearn. G. E.; Kakihana, R. Selective breeding for ethanol sensitivity: SS and LS mice. In: McClearn, G. E.; Deitrich, R. A.; Erwin, V. G., eds. The development of animal models as pharmacogenetic tools. DHHS Publication No. (ADM 81-1133). Washington, DC: U.S. Government Printing Office; 1981:147-159. 14. Mochly-Rosen, D.; Chang, F.-H.; Cheevers, L.; Kim, M.; Diamond, I.; Gordon, A. S. Chronic ethanol causes heterologous desensitization of receptors by reducing alpha~ messenger RNA. Nature 333: 848-850; 1988. 15. Nagy, L. E.; Diamond, I.; Gordon, A. S. Cultured lymphocytes from alcoholic subjects have altered cAMP signal transduction. Proc. Natl. Acad. Sci. USA 85:6973--6976; 1988. 16. Nagy, L. E.; Diamond, I.; Collier, K.; Lopez, L.; Ullman, B.; Gordon, A. S. Adenosine is required for ethanol-induced heterologous desensitization. Mol. Pharmacol. 36:744-748; 1989. 17. Phillis, J. W.; Edstrom, J. P.; Kostopoulos, G. K.; Kirkpatrick, J. R. Effects of adenosine and adenine nucleotides on synaptic transmission in the cerebral cortex. Can. J. Physiol. Pharmacol. 57:12891312; 1979. 18. Phillis, J. W.; Wu, P. H. The role of adenosine and its nucleotides in central synaptic transmission. Prog. Neurobiol. 16:187-239; 1981. 19. Phillis, J. W.; Wu, P. H. Adenosine mediates sedative action of various centrally active drugs. Med. Hypotheses 9:361-367; 1982. 20. Proctor, W. R.; Baker, R.; Dunwiddie, T. V. Differential CNS sensitivity to PIA and theophylline in Long-Sleep and Short-Sleep mice. Alcohol 2:287-291; 1985. 21. Proctor, W. R.; Dunwiddie, T. V. Behavioral sensitivity to puriner-, gic drugs parallels ethanol sensitivity in selectively bred mice. Science 224:519-521 ; 1984. 22. Ribeiro, J. A.; Sa-Almeida, A. M.; Namorado, J. M. Adenosine and adenosine triphosphate decrease aSca uptake by synaptosomes stimulated by potassium. Biochem. Pharmacol. 28:1297-1300; 1979. 23. Riley, V. Adaptation of orbital bleeding technic to rapid serial blood
130
studies. Proc. Soc. Exp. Biol. Med. 104:751-754; 1960. 24. Romm, E.; Collins. A. C. Body temperature influences on ethanol elimination rate. Alcohol 4:189-198; 1987. 25. Smolen, A,: Marks, M. J.; Smolen, T. N.: Collins. A. C. Dose and route of administration alter the relative elimination of ethanol by LS and SS mice. Alcohol.: Clin. Exp. Res. 10:198-204; 1986. 26. Smolen, A,; Smolen. T. N. Demonstration of a threshold concentration for ethanol at the time of regaining the righting response in Long-Sleep and Shorl-Sleep mice. Alcohol Drug Res. 7:279-283; 1987. 27. Smolen, T. N.; Smolen. A. Blood and brain ethanol concentrations during absorption and distribution in long-sleep and short-sleep mice. Alcohol 6:33-38; 1989. 28. Smolen, A.; Smolen. T. N.; van de Kamp, J. L. Sensitivity of in-
SMOLEN AND SMOLEN
bred and selectively bred mice to ethanol. Alcohol 4:57~2; 1987. 29. Snyder, S. H.; Katims, J. J.; Annau, Z.; Bruns, R. F.; Daly, J. W. Adenosine receptors and behavioral actions of methylxanthines. Proc. Natl. Acad. Sci. USA. 78:3260-3264; 1981. 30. Van Calker, D.: Muller, M.; Hamprecht, B. Adenosine regulates via two different types of receptors, the accumulation of cyclic AMP in cultured brain cells. J. Neurochem. 33:999-1005; 1979. 31. Wager-Srdar, S. A.; Oken, M. M.; Morley. J. E.; Levine. A. S. Thermoregulatory effects of purines and caffeine. Life Sci. 33:24312438; 1983. 32. Wu, P. H.: Phillis, J. W.; Balls, K.; Rinaldi, B. Specific binding of 2-[3H]chloroadenosine to rat brain cortical membranes. Can. J. Physiol. Phannacol. 58:576-583; 1980.
07,~8,1q9191 $3.00 + .00
Purinergic Modulation of Ethanol-Induced Sleep Time in Long-Sleep and Short-Sleep Mice TONI NESS SMOLEN AND ANDREW
S M O L E N * i "1
hlstitute for Behavioral Genetics *Universit 3, of Colorado Drug Abuse Research Center and "/School of Pharmacy, Campus Box 447, Universit3, of Colorado, Boulder, CO 80309-0447 R e c e i v e d 26 O c t o b e r 1989; A c c e p t e d 22 O c t o b e r 1990
SMOLEN, T. N. AND A. SMOI'~EN. Purinergic modulation of ethanol-induced sleep rime in long-sleep and short-sleep mice. ALCOHOL 8(2) 123-130, 1991.--The long-sleep (IS) and short-sleep (SS) mice were selectively bred for differences in sensitivity to the depressant effects of ethanol. In addition to their differential sensitivity to ethanol, they are also differentially sensitive to purinergic agonists and antagonists. This suggests that there may be differences in the purinergic systems of these lines of mice which may aid in understanding how they differ in ethanol sensitivity. We have investigated whether these drugs are capable of modifying acute ethanol sensitivity as measured by ethanol-induced loss of the righting response (ethanol sleep time), waking blood and brain ethanol concentrations, and blood ethanol elimination rate. The purinergic agonists cyclohexyladenosine (CHA/, Lphenylisopropyladenosine (PIA), 2-chloroadenosine (CAD), and N-ethylcarboxamidoadenosine (NEC) increased sleep time in both LS and SS mice. however, LS mice were generally more affected than SS. The LS and SS mice were also differentially sensitive to the purinergic antagonists, theophylline and caffeine. Blood and brain ethanol concentration on awakening suggested that CNS sensitivity to acute ethanol administration was altered by pretreatment with agonists but not antagonists. Two agonists, CHA and NEC, significantly lowered ethanol elimination in both lines of mice while PIA, CAD, and the antagonists theophylline, and caffeine were without affect on elimination rate. These data support previous observations that adenosine-mediated systems may be involved in the modulation of ethanol sensitivity. Ethanol Sleep time Waking blood ethanol Long-sleep mice Short-sleep mice
Waking brain ethanol
Purinergic drugs
Adenosine
of ethanol-adenosine interactions focused on the A~ receptor subtype, however, recent studies suggest an important role for the A 2 receptor as well (14-16). Evidence supporting an interaction between ethanol and adenosine receptor systems comes from studies of drugs which are known to bind to specific adenosine receptor sites. For example, methylxanthine compounds, such as caffeine and theophylline, are adenosine receptor antagonists and have been found to modify ethanol responses in mice following both acute and chronic administration of ethanol (2, 3, 1 t). While there is good support for the hypothesis that adenosinemediated systems may be involved in modulating some of the CNS effects of ethanol, this notion is strengthened by comparisons of purinergic drug response in genetic stocks of mice selectively bred for differences in initial sensitivity to ethanol. Proctor and Dunwiddie (20,21) reported that behavioral sensitivity to purinergic drugs paralleled ethanol sensitivity in long-sleep (LS) short-sleep (SS) mice. Using a variety of behavioral tests, these investigators showed that LS mice were more sensitive to the potent A t agonist, L-phenylisopropyladenosine than were SS mice.
THE purine nucleoside adenosine is an important neuromodulator of central nervous system (CNS) excitability (5, 17-19). Adenosine-mediated systems have been postulated to be involved in the expression of some of the CNS effects of ethanol. Although adenosine and ethanol are structurally unrelated, both share sedative, hypothermic, and anticonvulsant properties, and acute administration of either ethanol or purinergic agonists produces similar behavioral and physiological responses (2, 6, 29, 31). In addition to their common properties, several lines of evidence suggest that ethanol may interact with an adenosine receptor system. Adenosine exerts its actions through multiple subclasses of receptors. These have been classified as A~, A 2 and P-site receptors, but multiple conformations may occur within each receptor subtype as well (1, 30, 32) The A~ receptor reduces or inhibits adenylate cyclase activity, whereas the A2 receptor enhances or activates adenylate cyclase activity. The P-site receptor has not been as well characterized as the A~ and A., receptors, but it appears to be located intracellularly and agonist binding is associated with decreased adenylate cyclase activity (12). Early studies
'Requests for reprints should be addressed to Dr. Toni N. Smolen, Institute for Behavioral Genetics, Campus Box 447, University of Colorado, Boulder, CO 80309-0447.
123
124
SMOLEN AND SMOLEN
In addition to a differential response to the adenosine agonist, they also reported that LS mice were more sensitive to the adenosine receptor antagonist theophylline than were SS mice. These data suggested that there may be differences in the purinergic systems of these lines of mice. In order to extend these findings, we have taken an alternate approach, namely, to investigate the effect of adenosine receptor agonists and antagonists on ethanol responses in LS and SS mice. In this study we examined whether purinergic drugs could modulate acute sensitivity to ethanol as measured by ethanol sleep time, waking blood and brain ethanol levels, and ethanol elimination rate. METHOD
Experimental Animals Animals used in this study were male and female LS and SS mice (13), 60 to 80 days of age. Mice were bred at the Institute for Behavioral Genetics, maintained on a 12-hour light cycle, and were allowed free access to food (Wayne Lab Blox) and water. All experiments were begun between 0900 and 1000 hours. All procedures described in this paper were approved by the University of Colorado Animal Care and Use Committee as being consistent with USPHS standards of humane care and treatment of laboratory animals.
Drug Solutions All drugs were administered intraperitoneally. Ethanol was prepared as a 20% (w/v) solution in saline. Doses were varied by adjusting the volume injected, because of the marked concentration-dependent effects of ethanol on sleep time (10). The following purinergic drugs were used in these studies: cyclohexyladenosine (CHA), L-phenylisopropyladenosine (PIA), 2chloroadenosine (CAD), N-ethylcarboxamidoadenosine (NEC), theophylline and caffeine. All were dissolved in saline at a concentration such that the injection of 0.01 ml of solution per gram body weight would result in the dose indicated in the figures. The doses and timing used were chosen from previously published reports and from our own preliminary dose-response and timecourse experiments that examined the behavioral and physiological effects of these agonists and antagonists in LS and SS mice (manuscript submitted).
Determination of Sleep Time, Blood and Brain Ethanol Content, and Ethanol Elimination Rate Purinergic drugs or saline were administered 15 min before ethanol. Sleep times were determined by injecting LS mice with 2.5 g ethanol/kg body weight, and SS mice with 5.0 g ethanol/ kg body weight. These are iso-effective doses resulting in sleep times of approximately 60 min. Following ethanol administration the animal was placed on its back in a V-shaped holder until it was able to right itself 3 times in a 30 s period (26,28). At the time of awakening, a 10 p,I blood sample was collected from the retroorbital sinus (23), and the whole brain was removed for measurement of ethanol concentration. Ethanol concentration was measured enzymatically as detailed previously (26,28). Blood and brain samples were collected from mice that failed to lose the righting response at 30 min post ethanol injection, which corresponds to the end of the ethanol distribution phase (27). These mice were given a sleep time score of zero and were included in the sleep time data analysis following verification that they had received their full dose of ethanol. Their corresponding 30-min blood and brain values were not included in the overall analyses
since these do not reflect waking ethanol levels. Sleep time experiments were terminated after 4 hours because we have observed virtually 100% mortality in agonist pretreated mice that had not regained the righting response by this time. Mice which had not awakened by that time were given a score of 240 min, and blood and brain samples were taken for ethanol determinations. Data from these animals are presented separately but were not included in the figures or in our analyses because they, too, do not reflect true waking times or ethanol levels at awakening. Ethanol elimination rate was measured by taking a 10 p.l blood sample 0.5, I, 2, 3 and 4 hours after ethanol administration. Blood ethanol concentration was determined, and the linear decline in ethanol content was calculated as described previously (25). None of the purinergic drugs, when administered alone at the doses used in the present study, caused the animals to lose the righting response.
Data Analysis Data were analyzed by one- or two-way analysis of variance (ANOVA) as appropriate. The LS and SS mice were given different doses of ethanol which are iso-effective with respect to sleep time, therefore the sleep time data were analyzed as a twoway ANOVA using line and drug as between subjects factors. Since the two lines have significantly different ethanol thresholds for regaining the righting response (261 the waking blood and brain ethanol levels are always significantly different; therefore, these data were analyzed separately for each line using dose as the single between-subjects factor. The data for the purinergic agonists were analyzed separately from the the data for the purinergic antagonists. Following a significant overall effect, differences in individual group means were detected using Duncan's Multiple Range test. A p value of 0.05 was considered significant, and is the only level reported. Data presented in the figures have been collapsed across sex due to lack of significant differences between males and females. RESULTS
Figures 1 through 3 show the effect of pretreatment with purinergic agonists on ethanol-induced sleep time and waking blood and brain ethanol concentrations in LS and SS mice. Pretreatment with the A~ agonist CHA increased ethanol-induced sleep time (Fig. l) and decreased ethanol elimination rate (Fig. 7) in both lines of mice. Two-way ANOVA of the sleep time data showed significant main effects of line, F(1,100)=5.46, p<0.05, and dose, F(3, lO0) = 34.8, p<0.05, and a significant line-by-dose interaction, F(3,100)= 5.18, p<0.05, which resulted from the LS mice being affected to a greater degree by CHA than SS mice. Waking blood and brain ethanol levels were lower in CHA-treated LS mice compared to saline-pretreated controls, F(3,48)=21.3 and 12.6, respectively, p<0.05 for both, which is suggestive of a CHA-mediated alteration in CNS sensitivity to ethanol. In contrast, while CHA-pretreated SS mice had lowered waking blood ethanol levels, the waking brain ethanol levels were also lower (not significantly) from saline-pretreated controls, F(3,52)= 6.77 and 2.30, p<0.05 and NS, respectively. Approximately equal numbers (45%) of LS and SS mice tested at 0.3 and 1 mg/kg CHA did not regain the righting response within the four-hour time limit and were given the maximum sleep time score of 240. Blood and brain ethanol levels for these animals are presented in Table 1 and support the finding of diminished blood ethanol elimination following CHA pretreatment. Pretreatment with another A~ agonist, PIA, caused a dose-de-
PURINERGIC DRUGS AND ETHANOL SENSITIVITY
200 :
~
lLS Mice I - - I SS Mice
125
*
200
_E . . 15o
150
.E
-
-~
mE
o
i :f
4.50
g~
• ~ :r,ool-
i...
=o_.
/n in|
4,5O
e,m~3oo 150 0
0
.1 .3 1.0 Dose of CHA (mg/kg)
FIG. I. Sleep times and waking blood and brain ethanol levels in LS and SS mice following pretreatment with cyclohexyladenosine (CHAI. Values are mean ± SEM of 10-24 mice per group. *Significantly different from respective saline-treated control, p<0.05. **Significantly different from LS mice, p<0.05.
pendent increase in ethanol-induced sleep time, 1::(3,109)= 19.2, p<0.05, and the LS mice were more affected than the SS (Fig. 2). However, PIA was less effective than CHA in differentiating between the LS and SS mice since the main effect of line failed to reach statistical significance. Waking blood and brain ethanol levels were decreased in LS mice again suggesting an alteration in CNS sensitivity to ethanol resulted from pretreatment with the A t agonist. Although PIA-pretreated SS mice slept longer, they awoke at blood and brain ethanol levels which were the same as saline-pretreated controls. Figure 7 shows that PIA was also less potent than CHA in reducing ethanol elimination rate in both lines of mice. Figure 3 shows the results of pretreatment of mice with the mixed A2/A ~ agonist NEC (NEC has more A, than Aj activity whereas the other mixed agonist, CAD has more Al than A 2 activity). NEC is a very potent drug with a steep dose response curve (Fig. 3). A 33-fold lower dose produced ethanol sleep times that were the same as, or greater than, that seen with the more specific A~ agonists CHA and PIA. Doses of NEC above 0.03
II
n n n n H
II
~F
,,+;~
~¢
T*
~
/11 /
150
o
._c o °
*
_a_,
lfi
0
•~ - -
*,
lOO 50
"loo
B B L S Mice I-"-I SS Mice
H
i
H
/I .
iI/1 ilill ~
0
.1 .3 1.0 Dose of PIA (mg/kg)
FIG. 2. Sleep times and waking blood and brain ethanol levels in LS and SS mice following pretreatment with L-phenylisopropyladenosine (PIAI. Values are mean---SEM of 10-24 mice per group. *Significantly different from respective saline-treated control, p<0.05.
mg/kg produced ethanol sleep times in excess of 240 min and virtually 100% mortality. The LS and SS mice did not differ in either their sleep time response to NEC (Fig. 3) or the significant decrease in ethanol elimination rate produced by NEC pretreatment (Fig. 7). Waking blood and brain ethanol concentrations were significantly different from control levels only at the highest dose of NEC (Fig. 3). The mixed AI/A 2 agonist, CAD increased sleep time in LS but not SS mice, with no effect on waking blood or brain ethanol levels (Fig. 6) or ethanol elimination rate in either line of mice (Fig. 7). Figure 4 shows that pretreatment with the purinergic antagonist theophylline resulted in effects opposite to those of the purinergic agonists. Theophylline decreased ethanol sleep time significantly. Although the overall pattern of response was similar in both lines of mice, as was found for each of the agonists, LS mice were more affected by theophylline pretreatment than were SS mice. Two-way ANOVA revealed significant main effects of line, F(I,108)=9.47, p<0.05, and dose, F(3,108)= 10.56, p<0.05, with no significant line by dose interaction 1::(3,108) = 1.96, NS. In spite of the marked effect on sleep time,
126
SMOLEN AND SMOLEN
200
I
~ L S Mice [---]SS Mice
*
100
,
•
Ill
I
~
~@o 'roof
II
•
~E
"="
_
r~
II
II
""
..
r=l ""
%_
II
"
;,
I ~
~E
~ L S Mice I--ISS Mice
I
1
lit
II III Ill ill I
r
i ill"
°
,I iIi
,,,E
~@oo 4s° ~E
0
**
0
-6 " o° ~
I'1
°~'-
I
I
I
I
I
R
I
I
r'-,
I
I
"~'-
0
0
0
.005
.01
.O3
Dose of NEC (mg/kg) FIG. 3. Sleep times and waking blood and brain ethanol levels in LS and SS mice following pretreatment with N-ethylcarboxamidoadenosine(NEC). Values are mean - SEM of 8-24 mice per group. *Significantly different from respective saline-treated control, p<0.05.
theophylline did not alter waking blood or brain ethanol levels (Fig. 4) or ethanol elimination rate (Fig. 7) in either line of mice.
TABLE 1 BLOOD AND BRAIN ETHANOL LEVELS AT 240-MIN POSTETHANOL ADMINISTRATION IN CYCLOHEXYLADENOSINE PRETREATED LS AND SS MICE
Dose (mg/kg)
N
Sleep Time (Minutes)
LS
0 (saline) 0.3 1.0
21 9 10
57 --+ 5 240 240
247 +_. 10 173 --- 20 229 __. 16
250 +- 10 164 -)- 14 252 --- 18
SS
0 (saline) 0.3 1.0
24 8 10
57 --- 6 240 240
488 __- 15 351 _-. 19 451 _ 16
478 __- 19 395 .4- 14 475 - 34
Line
' - o° ' ~
Blood (mg/dl)
Tabled values represent the mean +__ SEM.
Brain ( m g / 1 0 0g)
0
,IiItl 10
20
40
Dose of THE (mg/kg) FIG. 4. Sleep times and waking blood and brain ethanol levels in LS and SS mice following pretreatment with theophylline (THE). Values are mean---SEM of 9-24 mice per group. *Significantly different from respective saline-treated control, p<0.05. **Significantly different from LS mice, p<0.05.
In contrast to theophylline, caffeine pretreatment differentially affected ethanol sleep time in LS and SS mice (Fig. 5). Pretreatment with caffeine decreased sleep time in the LS, but increased sleep time in the SS mice resulting in a significant main effect of line, F(1,131)=43.88, p < 0.05. The marked difference in response of the two lines to caffeine was also evident from the significant line by dose interaction, F( 1,131 ) = 12.15, p < 0 . 0 5 . Since caffeine had opposite effects on the sleep time response of the two lines, analysis of the dose effects essentially canceled each other, resulting in no significant main effect of dose, F(3,131)= 1.66, NS. The waking blood and brain ethanol data did not show a consistent pattern. Waking blood ethanol levels tended to be higher in the LS and lower in the SS mice, and within each line, waking brain ethanol levels were not significantly different from saline-pretreated controls. Although the sleep time data suggest an alteration in the metabolism of ethanol, a high dose of caffeine (40 mg/kg) did not affect ethanol elimination rate in either line of mice (Fig. 7). Figure 6 summarizes the effectiveness of the four agonists and
PURINERGIC DRUGS AND ETHANOL SENSITIVITY
100
127
ILS Mice I---I SS Mice
200
E
a
a
I/-S Mice I - - ] S S Mice
15o Eso
m
-~0 oo o,,-
25
50
°
0
"~ "'E
4.50
~E
i
15o ~=
o
- ~ p ~ 600 o
600
0
"~ ~ 600 ~ ~
•
~
"~,m'~
300
150
i
150 0
0 10 20 40 Dose of CAF (mg/kg)
FIG. 5. Sleep times and waking blood and brain ethanol levels in LS and SS mice follox¢ing pretreatment with caffeine (CAF). Values are mean +--SEM of 11-24 mice per group. *Significantly different from respective saline-treated control, p<0.05. **Significantly different from LS mice. p<0.05.
two antagonists in altering ethanol sleep time, waking blood and waking brain ethanol levels• The magnitude of the effects of the purinergic drugs was most pronounced at the highest dose of agonist and lowest dose of antagonist used. All of the agonists except NEC are compared at a dose of 1 mg/kg (for NEC a dose of 0.03 mg/kg was used), and the antagonists at 10 mg/kg. With the exception of caffeine, the differences between the LS and SS mice appear to be quantitative, rather than qualitative in nature: In general, all of the agonists increased, whereas the antagonists decreased ethanol-induced sleep times, but in each case the LS mice were more affected than the SS mice. Waking blood and brain ethanol levels tend to be more affected in the LS mice as well, but the magnitude of the effect of these purinea-gic drugs on ethanol concentrations was not large. DISCUSSION
The LS and SS mice were selectively bred for maximal differences in CNS sensitivity to the depressant effects of ethanol as
SAL CHA
PIA CAD NEC THE CAF
FIG. 6. Sleep times and waking blood and brain ethanol levels in LS and SS mice. Mice were pretreated with: SAL (saline, 0.01 ml/g body weight), CHA, 1.0 mg/kg; PIA, 1.0 mg/kg: CAD, 1.0 mg/kg: NEC, 0.03 mg/kg; THE, 10 mg/kg; or CAF, 10 mg/kg 15 min prior to receiving 2.5 (LS) or 5.0 (SS) g/kg ethanol. Values are mean--. SEM of 8-24 mice per group. (a) Significantly different from respective saline-treated control, p<0.05. (b) Significantly different from LS mice, p<0.05.
measured by loss of the righting response, or sleep time (13). We have previously shown that these lines of mice display a characteristic threshold value for ethanol which determines the point at which they regain the righting response, irrespective of the ethanol dose given (26). Threshold values for waking blood and brain ethanol concentrations were, respectively: 261 mg/100 ml and 230 mg/100 g (LS mice); and 502 mg/100 ml and 468 mg/100 g (SS mice)• Knowing these threshold values is useful for interpreting acute or chronic drug effects on ethanol-induced sleep time. An altered ethanol sleep time may result from a change in the normal metabolism rate, from an alteration in CNS sensitivity, or both. A drug treatment that causes an animal to awaken at an ethanol concentration significantly lower or higher than the normal threshold value has induced a change in the CNS sensitivity to ethanol regardless of the drug's effect on the rate of ethanol metabolism. In the present study, we have investigated the ability of several adenosine agonists and antagonists to alter CNS sensitivity to ethanol in an attempt to characterize the role of this neuronal
128
SMOLEN AND SMOLEN
-4-.-I
D rY ._..
100
~ LS Mice [ ~ q S S Mice
k-
g,I..I. 75
"4~m
** z SAL
CHA
PIA
CAD NEC
THE
CAF
FIG. 7. Ethanol elimination rates in mice pretreated with adenosine agonists or antagonists. Mice were pretreated with: SAL (saline, 0.01 ml/g body weight), CHA. 1.0 mg/kg; PIA. 1.0 mg/kg; CAD, 1.0 mg/kg; NEC, 0.03 mg/kg; THE, 50 mg/kg; or CAF, 40 mg/kg 15 min prior to receiving 2.5 (LS) or 5.0 (SS) g/kg ethanol. Values are mean -- SEM of 7 mice per group. *Significantly different from respective saline-treated control, p<0.05. • system in the behavioral and physiological responses to ethanol. The results of this study confirm and extend previous reports which have shown that sensitivity to purinergic drugs parallels that of ethanol sensitivity (20,21) and that purinergic drugs are capable of modulating some of the acute effects of ethanol (2, 3, 20, 21). All of the drugs used in this study are thought to exert their actions by combining with one or more of the adenosine receptor subtypes. Although none of these drugs have absolute specificity for a single receptor subtype, CHA and PIA have greater specificity for the adenosine A~ receptor. CAD has approximately equal agonist activity at both A t and A2 receptors, whereas NEC, which has activity at both A_~ and A~ receptors, is commonly regarded as primarily an A2 agonist. The competitive antagonists, theophylline and caffeine, are thought to have nearly equal affinities for both receptor subtypes. In spite of this lack of specificity, theophylline and caffeine are important to study because they are found in abundance in commonly consumed beverages such as tea and coffee which are often used by people in an effort to counteract the effects of alcohol. Proctor and Dunwiddie (21) have reported that behavioral sensitivity to purinergic drugs paralleled ethanol sensitivity in LS and SS mice. They used several behavioral tests, including latency to- and number of escape attempts from an elevated platform, and both light-dark crossings and time spent in the light compartment of a light-dark box, to show that LS mice were more sensitive to PIA than were SS mice. All test responses were found to be decreased by PIA in both lines of mice, but the ECso values were four-fold lower for LS than SS mice, indicating a quantitative rather than qualitative difference in their response to this agonist. Our findings support these results. The LS mice were more affected by A~ agonists than were SS mice. Both LS and SS mice are extremely sensitive to CHA. None of the drugs used in these studies are, by themselves, lethal. However, when CHA is administered in combination with ethanol it is almost always lethal at doses which significantly affect sleep time. In the present study a ceiling of 240 min was placed on the sleep time period since in our preliminary studies we observed 100% mortality among CHApretreated mice that slept longer than four hours after ethanol administration. Romm and Collins (24) have previously shown that ethanol elimination rate changes in direct proportion to body temperature. Thus, the mortality seen following CHA and ethanol
administration could be due to the profound hypothermia CHA causes, followed by a substantial decrease in ethanol elimination rate. A 1.0 mg/kg dose of CHA produces a gradual loss in body temperature of approximately 11°C for LS and 8°C for SS mice over a 3-h period. This results in a nearly halving of the ethanol elimination rate in both lines of mice (Fig. 7). Threshold values for waking blood ethanol were significantly lowered in both lines following agonist pretreatment, however, concomitant brain ethanol values were significantly decreased in the LS line only. The greater variability of the SS brain ethanol values may have contributed to the lack of statistical significance. Given the nearly identical pattern of response, we would conclude that CHA alters CNS sensitivity to ethanol in both lines of mice, but that the major difference between the lines was quantitative in nature since for each drug at virtually every dose used the LS were more affected than the SS. PIA was less potent than CHA in exacerbating ethanol sleep time. Only 6% of PIA-treated, compared to approximately 45% of CHA-treated animals received a sleep time score of 240 min. Although the lines did not differ significantly in their sleep time response, our analyses of blood and brain ethanol levels on awakening suggested that CNS sensitivity to ethanol was altered by PIA in LS mice only. The finding that these mice slept longer at the highest dose of PIA (1 mg/kg) but awoke at their threshold values for blood and brain ethanol concentration is puzzling since this dose of PIA, unlike CHA, has no effect on ethanol elimination rate. It is possible that, at high doses, the sleep time response is due to some other physiological action of the drug. The At/A,- agonist CAD was also less potent in modulating sleep time response than CHA. None of the CAD-treated mice received a sleep time score of 240 min, and only LS mice were affected by this agonist. Since the animals awoke at their characteristic threshold values for blood and brain ethanol levels, CNS sensitivity to ethanol was not altered by CAD. Due to a lack of specific adenosine A2 receptor agonists and antagonists, there have been few in vivo studies of this receptor and its potential role in modulating ethanol actions. Although NEC also suffers from this lack of specificity, it is, nevertheless, one of the most selective of the few A2 agonists commercially available. Recent cell culture studies suggest that A2-dependent responses may be important in both the acute and chronic effects of ethanol (1416). It has been shown that acute exposure to ethanol increases the extracellular concentration of adenosine, which then activates A 2 receptors resulting in an increase in intracellular cAMP levels. It has been suggested that chronic ethanol exposure results in an heterologous desensitization of receptor-stimulated cAMP production which is caused by the extracellular accumulation of adenosine (14-16). At the present time the role this would play in the physiological response to acute ethanol administration (e.g., sleep time) is not known, however, it further implicates adenosine receptor systems in the actions of ethanol. In our studies we found that NEC was particularly potent at increasing ethanol sleep time. A dose of 0.03 mg/kg increased sleep time to near maximum (240 min). As with CHA, this could be related to the profound hypothermia caused by NEC when given alone or when coadministered with ethanol. This results in a markedly diminished elimination rate for ethanol (Fig. 7). Since NEC is not a pure A 2 agonist, it is unclear whether the effects of this drug are due to actions at the A 2 or Aj site, or both. Waking blood and brain ethanol concentrations were diminished in both lines of mice. The pattern of response leads us to conclude that CNS sensitivity to ethanol is altered in both lines of mice, but that the response is more robust in LS than SS mice. In addition to their differential responsiveness to adenosine agonists, Proctor and Dunwiddie reported that LS mice were more
PURINERGIC DRUGS AND ETHANOL SENSITIVITY
129
sensitive to the adenosine receptor antagonist theophylline compared to SS mice (20). In the present study we found that theophylline was far more effective in decreasing sleep time in LS (especially the 10 and 20 mg/kg doses) than in SS mice (lowest dose only). This decrease in ethanol-induced sleep time by theophylline also agrees with the work of Dar et al. (2) who reported that theophylline pretreatment delayed the onset and decreased the duration of ethanol-induced sleep time in ICR mice. In the present study, both lines of mice awoke at their characteristic threshold values for blood and brain ethanol concentrations suggesting there was no alteration in CNS sensitivity to ethanol. That theophylline pretreatment resulted in a decrease in sleep time without a change in waking ethanol levels, suggests an alteration in ethanol elimination rate. However, we found that ethanol elimination rate was not altered by theophylline pretreatments of 10 (not shown) or 50 mg/kg (Fig. 7). The ethanol elimination rates presented in Fig. 7 were based on blood samples taken at 0.5, 1, 2, 3, and 4 h following ethanol injection. Thus, the first blood sample was collected at 30 min, which is beyond the waking time of the LS (15-29 min, Fig. 4) and nearly that of SS mice (36-52 min). It is possible that shorter sampling times may have revealed an alteration in either the time to peak blood ethanol concentration or elimination rate which were not evident using our standard protocol. A striking difference between LS an'd SS sleep time response was seen following caffeine pretreatment. The decrease in ethanol sleep time in LS mice is nearly identical to that seen in theophylline-treated mice. However, SS mice responded to caffeine pretreatment with an increase in ethanol sleep time. The fact that waking brain ethanol values of caffeine-treated mice were not significantly different from controls, coupled with the lack of a consistent pattern of response among the blood values, suggests
that caffeine does not alter CNS sensitivity to ethanol in either line. Similarly, caffeine does not appear to alter ethanol elimination rate in either line of mice although, as with theophylline, sampling blood at earlier time points may be more informative. It is unclear what factors beyond an alteration in CNS sensitivity or a change in the rate of ethanol metabolism might contribute to the sleep time response seen following treatment with these antagonists. However, theophylline and caffeine are known to have effects independent of their actions at adenosine receptors which may play a role in their ability to modulate ethanol sleep time (7,8). In summary, the results of the present study support the hypothesis that adenosine systems may be involved in modulating some of the effects of ethanol. The mechanism by which this modulation occurs is unclear, but it is not likely that ethanol combines directly with an adenosine receptor. Ethanol may perturb the membrane microenvironment surrounding the receptors, or may exert its effects through an adenosine-mediated second messenger system as has been recently proposed by Nagy and coworkers (15,161. With respect to the LS and SS lines of mice, there may be differences in the numbers of adenosine receptors themselves, as has been suggested by Fredholm and co-workers (9). Whether ethanol sensitivity can ultimately be associated with such receptor differences remains to be determined. ACKNOWLEDGEMENTS The authors wish to thank Lorie Lockwood, Christine Ross and Dawn Waterman for their excellent technical assistance. This work was supported by grants AA 07127 (T.N.S.I from the National Institute on Alcohol Abuse and Alcoholism and HD 21709 (A.S.) from the National Institute of Child Health and Human Development.
REFERENCES 1. Bruns, R. F.; Daly, J. W.; Snyder, S. H. Adenosine receptors in brain membranes: Binding of N°-cyclohexyl[3H]adenosine and 1,3diethyl-8-[3H]phenylxanthine. Proc. Natl. Acad. Sci. USA 77:55475551; 1980. 2. Dar, M. S.; Mustafa, S. J.; Wooles, W. R. Possible role of adenosine in the.CNS effects of ethanol. Life Sci. 33:1363-1374; 1983. 3. Dar, M. S.; Jones, M.; Close, G.; Mustafa, S. J.; Wooles, W. R. Behavioral interactions of ethanol and methylxanthines. Psychopharmacology (Berlin) 91:1-4; 1987. 4. Dunwiddie, T. V. The physiological role of adenosine in the central nervous system. Int. Rev. Neurobiol. 27:63-139; 1985. 5. Dunwiddie, T. V.; Basile, A. S.; Palmer, M. R. Electrophysiological responses to adenosine analogs in rat hippocampus and cerebellum: Evidence for mediation by adenosine receptors of the A~ subtype. Life Sci. 34:37-47; 1983. 6. Dunwiddie, T. V.; Worth, T. Sedative and anticonvulsant effects of adenosine analogs in mouse and rat. J. Pharmacol. Exp. Ther. 220: 70-76; 1982. 7. Fain, J. N.; Pointer, R. H.; Ward, W. F. Effects of adenosine nucleosides on adenylate cyclase, phosphodiesterase, cyclic adenosine monophosphate accumulation, and lipolysis in fat cells. J. Biol. Chem. 247:6866-6872; 1972. 8. Fredholm, B. B. On the mechanism of action of theophylline and caffeine. Acta Med. Scand. 217:149-153; 1985. 9. Fredholm, B. B.; Zahniser, N. R.; Weiner, G. R.; Proctor, W. R.; Dunwiddie, T. V. Behavioral sensitivity to PIA in selectively bred mice is related to a number of A a adenosine receptors but not to cyclic AMP accumulation in brain slices. Eur. J. Pharmacol. 111:133136; 1985. 10. Gilliam, D. M.; Collins, A. C. Concentration-dependent effects of ethanol in Long-Sleep and Short-Sleep mice. Alcohol.: Clin. Exp. Res. 7:337-342; 1983. I 1. Hilakivi, L. A.; Durcan, M. J.; Lister, R. G. Effects of caffeine on social behavior, exploration and locomotor activity: Interactions with ethanol. Life Sci. 8:543-553; 1989.
12. Londos, C.; Wolff, J. Two distinct adenosine-sensitive sites on adenylate cyclase. Proc. Natl. Acad. Sci. USA 74:5482-5486; 1977. 13. McClearn. G. E.; Kakihana, R. Selective breeding for ethanol sensitivity: SS and LS mice. In: McClearn, G. E.; Deitrich, R. A.; Erwin, V. G., eds. The development of animal models as pharmacogenetic tools. DHHS Publication No. (ADM 81-1133). Washington, DC: U.S. Government Printing Office; 1981:147-159. 14. Mochly-Rosen, D.; Chang, F.-H.; Cheevers, L.; Kim, M.; Diamond, I.; Gordon, A. S. Chronic ethanol causes heterologous desensitization of receptors by reducing alpha~ messenger RNA. Nature 333: 848-850; 1988. 15. Nagy, L. E.; Diamond, I.; Gordon, A. S. Cultured lymphocytes from alcoholic subjects have altered cAMP signal transduction. Proc. Natl. Acad. Sci. USA 85:6973--6976; 1988. 16. Nagy, L. E.; Diamond, I.; Collier, K.; Lopez, L.; Ullman, B.; Gordon, A. S. Adenosine is required for ethanol-induced heterologous desensitization. Mol. Pharmacol. 36:744-748; 1989. 17. Phillis, J. W.; Edstrom, J. P.; Kostopoulos, G. K.; Kirkpatrick, J. R. Effects of adenosine and adenine nucleotides on synaptic transmission in the cerebral cortex. Can. J. Physiol. Pharmacol. 57:12891312; 1979. 18. Phillis, J. W.; Wu, P. H. The role of adenosine and its nucleotides in central synaptic transmission. Prog. Neurobiol. 16:187-239; 1981. 19. Phillis, J. W.; Wu, P. H. Adenosine mediates sedative action of various centrally active drugs. Med. Hypotheses 9:361-367; 1982. 20. Proctor, W. R.; Baker, R.; Dunwiddie, T. V. Differential CNS sensitivity to PIA and theophylline in Long-Sleep and Short-Sleep mice. Alcohol 2:287-291; 1985. 21. Proctor, W. R.; Dunwiddie, T. V. Behavioral sensitivity to puriner-, gic drugs parallels ethanol sensitivity in selectively bred mice. Science 224:519-521 ; 1984. 22. Ribeiro, J. A.; Sa-Almeida, A. M.; Namorado, J. M. Adenosine and adenosine triphosphate decrease aSca uptake by synaptosomes stimulated by potassium. Biochem. Pharmacol. 28:1297-1300; 1979. 23. Riley, V. Adaptation of orbital bleeding technic to rapid serial blood
130
studies. Proc. Soc. Exp. Biol. Med. 104:751-754; 1960. 24. Romm, E.; Collins. A. C. Body temperature influences on ethanol elimination rate. Alcohol 4:189-198; 1987. 25. Smolen, A,: Marks, M. J.; Smolen, T. N.: Collins. A. C. Dose and route of administration alter the relative elimination of ethanol by LS and SS mice. Alcohol.: Clin. Exp. Res. 10:198-204; 1986. 26. Smolen, A,; Smolen. T. N. Demonstration of a threshold concentration for ethanol at the time of regaining the righting response in Long-Sleep and Shorl-Sleep mice. Alcohol Drug Res. 7:279-283; 1987. 27. Smolen, T. N.; Smolen. A. Blood and brain ethanol concentrations during absorption and distribution in long-sleep and short-sleep mice. Alcohol 6:33-38; 1989. 28. Smolen, A.; Smolen. T. N.; van de Kamp, J. L. Sensitivity of in-
SMOLEN AND SMOLEN
bred and selectively bred mice to ethanol. Alcohol 4:57~2; 1987. 29. Snyder, S. H.; Katims, J. J.; Annau, Z.; Bruns, R. F.; Daly, J. W. Adenosine receptors and behavioral actions of methylxanthines. Proc. Natl. Acad. Sci. USA. 78:3260-3264; 1981. 30. Van Calker, D.: Muller, M.; Hamprecht, B. Adenosine regulates via two different types of receptors, the accumulation of cyclic AMP in cultured brain cells. J. Neurochem. 33:999-1005; 1979. 31. Wager-Srdar, S. A.; Oken, M. M.; Morley. J. E.; Levine. A. S. Thermoregulatory effects of purines and caffeine. Life Sci. 33:24312438; 1983. 32. Wu, P. H.: Phillis, J. W.; Balls, K.; Rinaldi, B. Specific binding of 2-[3H]chloroadenosine to rat brain cortical membranes. Can. J. Physiol. Phannacol. 58:576-583; 1980.