Summary. The interactions between ethanol and antidepressant drugs (both tri- cyclics and newer non-tricyclics) were studied in mice. The ability oi: these.
Jo~of
Neur~ ~ s ~ n
j. Neural Transmission 48, 223--240 (1980)
9 by Springer-Verlag 1980
Antidepressant Drugs and Ethanol: Behavioral and Pharmacokinetic Interactions in Mice J. M. Cott and S.-O. Ogren Astra L~.kemedel AB, Research and Development Laboratories, S6dert~lje, Sweden
With 2 Figures Received October 26, 1979
Summary The interactions between ethanol and antidepressant drugs (both tricyclics and newer non-tricyclics) were studied in mice. The ability oi: these drugs to enhance the sedative effects of ethanol at two different doses (3.2 and 4.0 g/kg) was measured. The percentage of mice losing the rightingreflex was used for the lower dose, and the duration of ethanol-induced sleep wag used at the higher dose. The relative order of potency was amitriptyline ~ imipramine ~ maprotiline = mianserin ~ desipramine chlorimipramine ~ iprindole > alaproclate > norzimelidine ~ zimetidine. Amitriptyline (60 mg/kg) caused death in all mice when combined with 4.0 g/kg ethanol. Clinically established antidepressants which enhanced ethanol sedation only at doses considerably above therapeutic l e v e l ~ e r e zimelidine and iprindole. The relative potency of the antidepressants " / >to enhance ethanol sedation is correlated with their inherent sedative/properties which are in turn related to their ability to block central 5-HT, a-NA, muscarinic and Hi-receptors. Amitriptyline (20mg/kg) was found to increase ethanol plasma levels to 202, 167 and 132 % of control values at 30, 60 and 90 rain aflcer ethanol administration, respectively. Desipramine, mianserin and alaproclate also increased ethanol plasma levels initially, but to a lesser extent. These findings suggest that in addition to their sedative effect, several antidepressants, particularly amitriptyline, are likely to interact with ethanol by increasing its concentration in plasma.
Introduction While tricyclic antidepressants may enhance the central depressant effects and toxicity of ethanol and other sedatives, in depth studies 15 Journal of Neural Transmission 48/4
0300-9564/80/0048/0223/$ 03.60
224
J. M. Cott and S.-O.Ogren:
on this subject are lacking. Amitriptyline has been shown to enhance the detrimental effects of ethanol on various learning and motor skills (Patrnan et al., 1969; Seppiil~i, 1977; Liljequist et al., 1978), while nortriptyline and mianserin enhance ethanol sedation in man (Seppiil~i, 1977; Liljequist et al., 1974). In animal studies, amitriptyline appears to be the most sedative, desipramine the least, with other tricyclics ranging in between (Milner, 1968; Allen et aI., 1971; HalliwelI et al., 1964; Suchorosky and Pegrassi, 1969; Voith and Herr, 1969). However, no attempts were made in these studies to determine the mechanism of the interaction with ethanol, which might be due either to pharmacokinetic alterations or additive sedative effects in the central nervous system. The only studies which included blood ethanol measurements were clinical studies in which only low doses of antidepressants and ethanol were given orally. These reports showed no change in blood ethanol levels after amitriptyline (Landauer et al., 1969; Seppiil~i et aI., 1975), nortriptyline (Hughes and Forney, 1963) or doxepine (Sepp~ilii et al., 1975). Another study (Hall et al., 1976), reported a significantly lower ethanol level a~er desipramine pretreatment. The purpose of the present study was first to determine the relative effects of a wide range of antidepressants, both the older tricyclics and some newer, structurally different types (Fig. 1), on ethanolinduced sedation and sleep in mice. In addition, plasma ethanol concentrations were determined for a few representative antidepressants given before ethanol treatment. To avoid the problem of different rates of gastro-intestinal absorption, all drugs were given intraperitoneally (i.p.). The Fotency of the various drugs to enhance the effect of ethanol was compared with their reported monoamine neuron reuptake inhibiting properties--ranging from the specific uptake inhibition of noradrenaline (NA) by desipramine (Ross and Renyi, 1975 a, 1975 b) and maprotiline (Maitre, 1975) to the specific uptake inhibition of 5-hydroxytryptamine (5-HT) by the bicyclic compounds, zimelidine and its active metabolite, norzimelidine (Ross et al., 1976; Ross and Renyi, 1977; Carlsson and Lindqvist, 1978) and the amino acid ester, alaproclate (Lindberg et aI., 1978; Carlsson and Lindqvist, 1978). Relative amine receptor blocking properties in the brain and the sedative effects of the tricyclic antidepressants have previously been compared. However, these reports (see Discussion) are somewhat unconvincing due to the restricted types and numbers of antidepressants used. Thus, the potency of all the present compounds to enhance the effects of ethanol was used as a measure of sedation, and correlated with their affinities for several different receptors.
Ethanol Potentiation by Antidepressant Drugs
II
(CH2)3 N(CH3)2
CH(CH~)2N(CH~)2
Imipramine
Amitriptyline
(CH2)3NHCH3
~-NCH3
Maprotiline
Mianserin
iI
i
(CHa)3N(CH.~)2
CHOH2N(CH3)2
Iprindole C!
225
Zimelidine CH3 0 CH~ O C H NH2
!"c
CH3 CH3
Alaproclate (GEA 654)
Fig. 1. Structural configurations of the antidepressants used in the present studies
Methods Behavioral Studies Male albino mice (NMRI) weighing 18 to 36 g were used in all studies. The mice were housed 25 per cage for at least 2 days before use on a 12-hour light-dark cycle and fed ad libitum. Experiments were conducted between 8 a.m. and 3 p.m. at an ambient temperature of 21 ~ to 23 ~ All mice were used only once. In the first experiment, a threshold dose (3.2 g/kg) of ethanol was given to mice 30 min al~ various antidepressant drug treatments, and the percentage of mice which lost the righting-reflex for at least one rain was determined. In other experiments, a "sleeping" dose of ethanol (4.0 g/kg) was given to mice 15--30 min after various antidepressant drug treatments, and the length of the loss of righting-reflex (sleep) was measured. Since ethanol-induced sleep duration is known to vary considerably, each experimental group was compared only with its own matched control group. In a third group of experiments, mice were treated with 10mg/kg twice daily for 4 days with amitriptyline, zimelidine, alaproclate or saline, and on the fiith day they were given a final dose of the antidepressant or saline 30 to 60 rain before ethanol (4 g/kg) and "sleep" duration was measured as before. I5"
226
J.M. Cott and S.-O. 121gren:
Determination of Plasma Ethanol Levels Mice were pretreated with 20 mg/kg amitriptyline, zimelidine, alaproclate, desipramine or mianserin 30 to 60 min before 5 g/kg ethanol. The animals were decapitated 30, 60 or 90 min aPcer ethanol administration and blood samples were collected in glass tubes containing 50 I.U. of heparin. The tubes were then centrifuged and the separated plasma stored at - 4 ~ for determination of ethanol concentration by the micro distillation technique (Buijten et al., 1977).
Drugs Drugs used were maprotiline' HC1, imipramine. HC1, desipramineHC1, and chlorimipramine. HC1 (Ciba-Geigy); amitriptyline- HC1 (Lundbeck) ; mianserin 9HC1 (Organon); iprindole 9HC1 (Wyeth) ; and zimelidine 2 HC1, norzimelidine 92 HC1 and alaproclate 9HC1 (Astra). All drugs were administered i.p., dissolved in saline solution, in a volume of 10 ml/kg body weight except for ethanol which was given as 16, 20 or 25 % wt./vol, solution in a final volume of 20 ml/kg. All drugs weights refer to the weight of the salt. The highest dose of any drug used was 60 mg/kg, since higher doses of most of the compounds produced signs of behavioral toxicity such as ataxia and hindlimb abduction.
Statistics For experiments involving the threshold ethanol dose, Fisher's Exact Probability test (one-tailed) was used. For the sleeping-time experiments a one-tailed Dunnett's t-test (Dunnett, 1955) was used. The ethanol plasma levels were converted to per cent of control for each of 5 separate experiments, and the totals were analysed with the two-tailed MannWhitney U-test. Results
Behavioral Studies The lower dose of ethanol (3.2 g/kg) was at the threshold for producing loss of righting-reflex, averaging one or two out of 10 mice. All antidepressants tested, except zimelidine, produced a significant increase in the number of mice losing the righting-reflex. The relative order of potency was amitriptyline >_ imlpramine maprotiline ~ mianserin _> desipramine . ~ chlorimipramine iprindole ~ alaproclate ~ norzimelidine (Table 1). The higher dose of ethanol (4.0 g/kg) induced sleep in all 330 mice but one, which was then excluded. Pretreatment with the antidepressants resulted in dose-related increases in sleep duration
Ethanol Potentiation by Antidepressant Drugs Table 1.
227
Effect of acute treatment with antidepressant drugs on the induction of sleep in mice by a threshold dose of ethanol
Drug Amitriptyline
Imipramine
Dose (mg/kg) 0 1.25 2.5 5 0 0.5 1.0
Maprotiline
Mianserin
Desipramine
Chlorimipramine
Iprindole
Alaproclate
Norzimelidine
Zimelidine
2.5 5 10 0 2 5 15 0 3 10 2O 40 0 10 2O 40 0 10 20 3O 40 0 10 2O 3O 40 0 2O 40 60 0 10 30 6O 0 2O 3O 4O 6O
Proportion sleeping
P value
4/20 5/10 8/10 10/10 3/20 0/10 4/10 9/10 9/10 10/10 3/20 3/10 7/10 9/10 2/20 4/10 4/10 6/10 10/10 4/20 5/10 7/I0 8/10 0/10 4/10 8/10 9/10 10/10 0/10 1/10 4/10 7/10 8/10 1/10 4/10 8/10 10/10 2/20 0/10 4/10 4/10 2/20 2/10 2/10 2/10 1/10
-NS < .001 < .001 -NS NS < .005 < .OO5 < .001 -NS < .01 < .OO5 -NS NS < .O5 < .001 -NS < .01 < .001 -< .05 < .001 < .001 < .001 -NS < .O5 < .OO5 < .001 -NS < .005 < .001 -NS < .05 < .O5 -NS NS NS NS
All drugs were given i.p. 30 min before ethanol (3.2 g/kg i.p.). The proportion of mice sleeping a~er the drug combination was compared with the appropriate group receiving ethanol alone according to Fisher's Exact Probability test (onetailed). NS = not significant (P ~> .05).
228
J . M . C o t t a n d S.-O. O g r e n :
(Table 2). The order of potency was amitriptyline > mianserin maprotiline > desipramine > alaproclate > iprindole > zimelidine which is similar to the order with the lower dose of ethanol (Table 3). All mice which received 60 mg/kg amitriptyline before ethanol died (Table 2). T a b l e 2.
Effect of acute treatment with antidepressant drugs on ethanolinduced sleep duration in mice Dose (mg/kg)
Sleep d u r a t i o n (rain + S.E.M.)
N
P value ( D u n n e t t ' s t-test)
0 1 3 10 30 60
27.0 + 3.2 5 0 . 0 + 6.7 5 1 . 8 + 5.0 5 8 . 6 + 4.5 120.5 + 11.3 All dead 15--120 min
20 10 10 10 10 10
-< < < < --
Maprotiline
0 3 10 30 60
37.7 + 52.2+ 78.9 + 129.1+ 244.8+
5.3 7.5 10.0 5.8 9.0
10 10 10 9~ 10
-NS < 0.005 < 0.001 < 0.001
Mianserin
0 3 10 30 60
18.6 + 39.5+ 66.5+ 92.4+ 155.4+
5.8 5.9 8.4 4.7 14.2
9 10 10 10 9a
-< < <
0.05).
Ethanol Potentiation by Antidepressant Drugs
229
Table 3. Comparative potency of antidepressants to enhance ethanol sedation and relieve clinical depression
Drug
3.2 g/kg ethanol EDs0 + S.E.M. (mg/kg i.p.)
4.0 g/kg ethanol ED100 (mg/kg i.p.)
Therapeutic dose range+ (mg/kg/day)
Amitriptyline Imipramine Maprotiline Mianserin Desipramine Chlorimipramine Iprindole Alaproclate Norzimelidine Zimelidine
1.3 + 0.3 i .6 + 0.4 3.3 + 1.0 7.6 + 2.9 9.5 + 5.9 11.9 + 2.4 22.7 + 3.2 23.4 + 4.9 > 60 > 60
2.3 ND 6.5 3.6 28 ND > 60 36 ND > 60
1.1--4.3 1.1--4.3 1.1--4.3 0.5--I .7 1.1--2.9 1.1--2.0 0.6--2.6 * * 2.0--3.0
ED~0 was determined by probit analysis and indicates the dose of antidepressant necessary to induce loss of righting-reflex in 50~ of the animals. ED100 was determined by linear regression analysis and indicates the dose of antidepressant necessary to prolong ethanol sleep by 100 ~ ND = not determined. + Based on average manufacturers recommended dose for a 70 kg adult. * Not yet established antidepressant.
Repeated pretreatment with amitriptyline (10 mg/kg, twice daily for 4 days) with a final dose 60 min before 4 g/kg ethanol, increased sleep duration by about the same amount as acute pretreatment (55.2_+4.5 min compared to saline 31.6_+4.7 min), which suggests little tolerance development to the effects of amitriptyline. Similar repeated treatments with alaproclate (10mg/kg) and zimelidine (10 mg/kg) did not affect sleep duration (data not shown), indicating that subchronic treatment with these drugs does not alter the effects of ethanol. Biochemical Studies The effects of 20 mg/kg of five different types of antidepressants on plasma concentrations of ethanol 30, 60 and 90 min after 5 g/kg ethanol administration is shown in Fig. 2. Amitriptyline increased ethanol concentrations to 202+4.9, 167+4.9 and 132_+5.4 percent of control values at 30, 60 and 90 min, respectively. Mianserin caused significant increases at 30 and 60min, while alaproclate and desipramine increased ethanol concentrations at 30 min only. Zimelidine had no significant effect at any time period.
J. M. Cott and S.-O. Ogren:
230
--] Saline control (S)
[ ~ Amitriptyline(Am) ~
Alaproclate (AI)
] Mianserine(M) ] Desipramin(D) e [ ] Zimelidine(Z) E m
S Am AI
M
30min
D
Z
S Am Ai
M
60 min
D
Z
S Am AI 90
M
D
Z
min
Fig. 2. Effects of acute pretreatment with antidepressant drugs on ethanol plasma
levels in mice. All antidepressants were given i.p. at 20 mg/kg 60 rain before ethanol (5 g/kg) except alaproclate which was given 1 min before ethanol (due to its rapid onset and short duration). Ethanol was measured by the micro distillation method. Control columns represent the mean + S.E.M. of 10 mice. Drug columns represent the means + S.E.M. of 6 mice. Statistical comparisons were made according to Mann-Whitney U-test (two-tailed). * = p < .02, ** = p < .002
Discussion The present findings emphasize the m a r k e d synergism of the sedation, and in the case of amitriptyline the lethality, of ethanolantidepressant combinations. The potency of this interaction varies considerably with different structural types of antidepressants and appears to be related to their inherent sedative effect, as well as to changes in ethanol pharmacokinetics. Previous attempts to relate antidepressant-induced sedation to their structure and to their effects on various neurotransmitter mechanisms have been limited to the tricyclic compounds. Within this structural group, several generalizations have been proposed. 1. The tertiary amines such as imipramine, amitriptyline and chlorimipramine are more potent inhibitors of 5 - h y d r o x y t r y p t a m i n e ( 5 - H T ) reuptake into neurons and are more often used in agitated
Ethanol Potentiation by Antidepressant Drugs
231
depression where they have a sedative effect; the secondary amines such as desipramine and protriptyline are more potent inhibitors of neuronal noradrenaline (NA) reuptake and have a more prominent psychomotor-activating effect which may be beneficial in retarded depression (see Carlsson, 1976). 2. There is a rough positive correlation between the sedative effects and the potency of a-adrenergic blockade by the tricyclics (Brodie et al., 1961; Van Zwieten, 1975; U'Prichard et al., 1977; Smith and Hauser, 1979; Hall and Ogren, submitted for publication). 3. A positive correlation also exists between anticholinergic activity and sedation (Vaillant, 1969; Bene~ovc~ and N~hunek, 1971; BlackwelI et al., 1972; Snyder and Yamamura, 1977; Hall and Ogren, submitted for publication). These relationships hold fairly well within the older tricyclic group, but several of the newer drugs with different molecular structures (Fig. 1) have quite different neuropharmacological actions and different clinical side-effects. Zimelidine, a bicyclic compound, which has recently been shown to be an effective antidepressant (Cox et al., 1978; dberg and Holmberg, 1979; Coppen et al., 1979), and its main metabolite, norzimelidine, are both selective and potent inhibitors of 5-HT uptake (Ross et al., 1976; Ross and Renyi, 1977). Zimelidine, at doses markedly inhibiting 5-HT uptake, does not cause significant sedation in mice (Ross et al., 1976) and none of the aforementioned clinical trials have reported sedative side-effects--thus differentiating zimelidine from the tertiary amine tricyclics. Alaproclate is even more selective than zimelidine in 5-HT uptake inhibition (Lindberg et aI., 1978; Carlsson and Lindqvist, 1978) and causes no sedation at doses which inhibit uptake in animals (unpublished observations). Furthermore, neither zimelidine (Hall and Ogren, submitted for publication; Cox et al., 1978; Xberg and Holmberg, t979) nor alaproclate (Hall and Ogren, submitted for publication) have significant anticholinergic or adrenolytic properties, which is in keeping with the second and third hypotheses above. The tetracyclic compound, mianserin, also has a different pharmacological profile from the tricyclics. Its effects on monoamine metabolism are quite different (see Brogden et al., 1978), in fact it causes a potent blockade of central postsynaptic 5-HT receptors (Maj et aL, 1978; Ogren et aI., 1979). It also appears to be an a-adrenergic receptor antagonist (Van Zwieten, 1975). It is strongly sedative but is not a potent anticholinergic (Brogden et al., 1978). The tetracyclic compound, maprotiline, is a potent and selective NA uptake inhibitor, but unlike the tricyclics which preferentially
232
J.M. Cott and S.-O. Cigren.
inhibit NA uptake, it is quite sedative (Maitre, 1975). There is little information concerning its effects on postsynaptic amine receptors, but clinical trials indicate a substantial incidence of dry mouth, visual disturbances and other symptoms associated with anticholinergic pr.operties (Jukes, 1975). Iprindole is also quite different from the imipramine-like tricyclics in that it has no significant effect on monoamine uptake (Ross et al., 1971) or turnover (Rosloff and Davis, 1978; Carlsson and Lindqvist, 1978). There is conflicting data regarding its effects on adrenergic receptor mechanisms (Vetulani and Sulser, 1975; Palmer, 1976; Van Zwieten, 1977; Jones, 1978). This controversy may be partially explained by Bevan et al. (1975), who showed biphasic, dosedependent potentiation and antagonism of the response to electrophoretic application of NA, dopamine, 5-HT and acetylcholine on single neurons. Thus, it is unknown whether potentiation or antagonism predominates after exogenous administration of therapeutic doses of iprindole. Neither sedation nor anticholinergic effects have been reported during clinical use, however (Cliff, 1970; Anonymous, 1971). In addition, iprindole is not a 5-HT antagonist in vivo or in vitro (Ogren et al., 1979). In summary, the 3 previously mentioned correlations of antidepressant properties with sedative effects are not significant when the structurally new antidepressants are included. Sedation and the relative potency of 5-HT over NA uptake inhibition appears to correlate only with the dibenzazepine and the cycloheptadiene tricyclics. When correlating anticholinergic effects with sedation, maprotiline, iprindole, and zimelidine appear to fit, but mianserin does not. Adrenergic antagonism may be a good predictor of sedation, but sufficient evidence is lacking due to the absence of selective in vivo tests for NA receptor activity. Table 4 summarizes the relationships between the various pharmacological properties of the drugs used in the present studies. In order to obtain a more quantitative estimation of the affinity of the antidepressants for brain amine-receptors, other data from this laboratory (Hall and Ogren, submitted for publication) were examined. Relative potencies of these receptor affinities are shown to the right in Table 4. The receptor ligands were tritiated QNB, WB 4101, d-LSD and mepyramine for the muscarinic, a-antagonist, 5-HT and histamine (H 0 receptors, respectively. Receptors were prepared and extracted from rat brain homogenates with a millipore filter or with a semi-automatic filtration method developed here (Hall and Thor, 1979). Correlation of ethanol potentiation with the binding affinities for
233
Ethanol Potentiation by Antidepressant Drugs
+ ++++
+
e'~
+ v
+++++++000 +
:0
+ +
-t- + ~: + o + o o o o + ++
o
+ +
+
o
+ + + + + +0000
~
+ + + + + +++++~000
+ ++ ++++
~o
2
+
++++++o,q~o
0
0
b ~
a Z
0
+ +++
+
F:::~- ~ + + + 0 + + o ~ : o
f R
b
,-0
~.~
_•
+++ ++
~ § ooo+
o ,+ + +
~-~-~ ='~= . -~~o ~ ~ .~~ $l~.,--n.~ ~ ~ "~"~o 8~ . ~~_ ~ ~ ~ ~'~
~.~
~"~'~
~ ~.
i? N
II
~ =
0 g
~
+++0++0000
.= ~ . ~ o
0
N
2~
+ ~+
o e~
234
J. M. Cott and S.-O. Ogren:
the drugs by the Spearman Rank test gave the correlations shown in Table 5. Antidepressant-induced sedation probably involves a combination of factors, in which the action on several different amine receptors plays a primary role. This is evidenced by the significant correlation of ethanol potentiation with all four receptors when the ten drugs from Table 3 are compared. Table 5. Summary of rank order correlations of 3.2 g/kg ethanol potentiation with in vitro receptor binding potencies of various antidepressants~ Radioligand displacement Drug
3H-WB 4101
8H-QNB
8H-d-LSD
3H-Mepyramine
All antidepressants from T a b l e 3
r s = .70 n = 10 p